I created this analysis stub; a fake game to post comments for
Twixtbot vs. The World.
The game is unrated, and the victory or loss will be recorded under my fake ID "MCx".
Awaiting commencement.
On 2019-12-24 at 14:27, MisterCat said:
my hyperlink brings you to the game at LG. You need to get here by using the 'Twixt Commentator' button. I tried.
Now that you are HERE, you see how this goes - post comments below, and once the game starts, I will use the comments to decide on moves. I have not even decided on whether to go FIRST or SECOND, and it gives you that option when you make a challenge.
Comment?
On 2019-12-24 at 17:26, twixter said:
I would advise you make TB move first. Theoretically we will be assured of at least a draw, if there is a first move that leads to a draw with best play. If there is no such move, then we have a win. All that remains are details.
I hope Alan H. can help!
On 2019-12-24 at 19:13, Alan Hensel said:
Hm, that's one approach, but I think it makes the most sense close to perfect play. I'm guessing TB is not really that close to perfect play.
If you let TB go first, it'll probably play on row 5, 6, 7, or 8, and that is unfamiliar territory right away for a human. If you go first and try to lead with a move that far from the edge, TB tends to make a tactical battle out of it, which is also unfamiliar territory for us humans. If you go first and pick something like d3 that seems even-handed for humans, TB seems to think that's way too weak and knows how to crush it.
Of these bad options, I personally think our best strategy is to go first, and choose a first move that TB would play, and keep playing it at its own game - throw its own kind of tactical opening moves back in its face. If you try to make it play the way a human plays, things seem to deteriorate quickly. We have a pretty good idea what first moves TB swaps, so If you want to effectively go second, you can just choose one that we know TB swaps.
Here are my TwixtBot first move stats for reference.
TB likes to lead with h8, but swaps it. It also likes j6, which it doesn't swap. It likes h6, which looks like it's probably in the no-swap zone, but we can't be sure until somebody tries it. Or, if you're really feeling adventurous, we haven't seen any TB game start on i7 at all.
I'd like to recommend a pre-first-move movie: "Miracle" ;-) Go Team!
On 2019-12-24 at 21:36, MisterCat said:
So far, David suggests letting the Bot go first, so we can swap if desired. Alan disagrees, suggesting that we go first and choose one of the Bot's favorite moves.
I am purposefully trying to stay out of this, as much as I can. I hope some more comments chime in, and we can agree - at least on the first move!
I wish this game to begin by New Year's Eve, at the latest. I will start it earlier if there appears to be more of a consensus. (meow)
On 2019-12-24 at 21:39, MisterCat said:
1. h6. This is NOT the official game move, but at least the board won't look empty! (meow)
On 2019-12-25 at 18:51, twixter said:
Alan, are any bot v. bot games included in your opening stats? I hope Mr. Jordan will allow us (meaning you, Alan) access to as much of an encyclopedia of Twixt openings as possible. At least, we have the games he posted here.
On 2019-12-25 at 19:30, Alan Hensel said:
The games posted here are included.
On 2019-12-27 at 19:34, bob440 said:
well, I'll jump in and opine that twixter's suggestion seems to leave a few salient details yet to be supplied.
Alan sounds much more confident, but I, for one, haven't the foggiest notion as to how to play TB "at its own game."
[ sigh ] none of that leads to a suggestion for a 1st move.
On 2019-12-27 at 23:58, MisterCat said:
I've been thinking about what Alan and David said. My inclination is to go with David, and give move 1 to the Bot. We have seen Twixtbot start with plenty of LOUSY LOOKING MOVES (and still win, of course), so we give it a chance to make a lousy move for us to play against. On the other hand, if the bot makes what appears to be a good move, such as analyzed by Alan, then we swap.
So I say pass. (mc)
On 2019-12-29 at 00:46, bob440 said:
don't remember where in the LG forum I read it, but I thought Bony said that TB doesn't have an opening book
wouldn't that mean that it evaluates all of its opening moves as roughly 50-50 to result in a win (whether or not its opponent swaps)?
so, are we overthinking the 1st move stragedy? (yes, I know that's not a word)
On 2019-12-29 at 18:27, MisterCat said:
We have TIME to overthink the 1st move; after that, we will be under more pressure. But New Year's Eve approaches - David is now leaning towards using a move suggested by Alan; I still say that by passing, we have a chance that Bot will play a weak 1st move (no opening book, as noted by Bob) - something we can take advantage of. And if the Bot plays any of the better 1st moves (as shown by Alan's charts, for example), then we swap. I need somebody to persuade me otherwise, else I'll be doing this.
On 2019-12-30 at 06:19, spd_iv said:
hi all,
what about C8, this move looks quite strong, not as good as C12 but still offers many possibilities (good connection to the top by C4). I think the TEAM should make a first move which TB did not play very often. I am not good at the beginning and I will try to visit this site from time to time to make some comments.
In my opinion moves like H8, J6, H6, I7 looks like to be easily swapped. If TB doesn't swap J6 maybe we should try this one. Disadventage of this move is that TB played it many times and knows how to play against it.
On 2019-12-30 at 21:45, Alan Hensel said:
TwixtBot reliably swaps or doesn't swap moves based on the thousands of games it has played against itself, so I think if we start with a move that TwixtBot swaps, we should have a plan for move 3. We have no public data for C8, but looks like it's in TwixtBot's swap zone. |1.c82.swap3.???
TwixtBot seems to believe that certain first moves are worth attacking directly. For example, it appears that TwixtBot thinks a good response to |1.f8 is 2.f4. Would you feel comfortable playing 2.f4? Would you know how to follow through with that? What if the best response to |1.c8 is actually something weird like that, and anything we're familiar with is something TwixtBot easily turns into an advantage for itself?
I'm not hopeful that basing decisions on what we think TwixtBot is familiar with is useful. TwixtBot has played thousands of times against itself starting on every legal spot on the board. It should know how to play against any first move, as comfortably as putting on an old shoe. Our own familiarity is a factor.
Yes, we know TB doesn't swap J6. If we start there, we'll get 3rd peg. I feel that's easier to analyze than 2nd peg. If spd_iv is ok with |1.j6, I'm comfortable with that.
However, as I said in the LG forums, I think we should let the highest ranking Twixt player in the room make final decisions. At the moment, that is spd_iv.
On 2019-12-30 at 22:52, Alan Hensel said:
A little research on |1.j6: TwixtBot's response has been 2.L13, 3 times out of 3, shown here from 3 different angles: j6, o19 swap, j19 swap.
When faced with |1.j62.L13, TwixtBot responded with 3.p13, as seen here.
Games 2054062 and 2082323 also involve 1.j6 and TwixtBot, but are not as informative.
On 2019-12-31 at 00:02, MisterCat said:
I'm with spd_iv's c8 to open. A look at Alan's first move stats show c8 to be one of the most successful winning moves, for HUMANS. A looks at Alan's Twixtbot stats show that it has NEVER played this, so doesn't like it for some reason. Therefore, we have a good chance to not get swapped (and if we are, oh well).
As the master of ceremonies here, I hope NEVER to be put in a position where I have to make the final, faulty choice of a move. I am really hoping the discussions will swirl around agreed moves. But if I MUST, I will read the discussion and try to go with a 'majority rule' - not strictly going with the highest rated player. Three 2200's will overrule one 2400, etc., Remember, practically everybody has a losing record against Twixtbot now, and I hope that we can prevail by working as a TEAM.
Bony Jordan HAS given approval for this experiment, and he is welcome to contribute his own, HUMAN analysis.
Tomorrow is New Year's Eve, everybody. Happy New Year, and the game WILL commence. (meow)
On 2019-12-31 at 05:04, BonyJordan said:
My human analysis is that I am really bad at this game and when I read commentary and analyses it's mostly over my head.
On 2019-12-31 at 06:48, spd_iv said:
today I will meet my brother, I will try to convince him to make some posts here.
from the other hand I will always swap |1.j6 - I do not remember any of my game starts with j6 but my intuitions tells me this move is very strong: |1.j62.l133.h134.h105.l116.m11 (game nr 2075839) what would TB play after 7.f10?
I think both j6 and c8 looks fine for the first move and I agree to use one ot them.
On 2019-12-31 at 07:09, spd_iv said:
Can someone tell me if TB will swap |1.b6? this seems to be bad move but on "TB first move stats" I see that TB swapped it.
On 2019-12-31 at 11:42, nie_wiesz said:
I would suggest starting the game with a move on the boundary between "Swap" and "No Swap" zone. So I agree to Alan that |1.c8 is too strong and is likely to be swapped by TB.
I have no idea why TB swaps |1.b6 - seems much weaker than |1.d6 (which TB do not swap). I think that all the advantages |1.b6 has are included in |1.d6...
So I would recommend to play: 1.b62.swap3.j9 (here 3.j9 can be changed to an adjacent field) OR 1.d6 and we'll see what is the TB response
On 2019-12-31 at 13:47, nie_wiesz said:
Case: |1.b6 |1.b62.swap3.j9 can be countered by 4.j9 so I would recommend to leave left top corner battle for later and play 3rd move in right bottom corner: |1.b62.swap3.q15 ...
Case: |1.d6 -- I would recommend this one (I'm curious what would be the TB response)
Looks like folks are rejecting PASS, and 1. c8 has been trashed by Maceij. Alan's above analysis of 1. j6 seems ok; The bot loves playing it (for some reason), but doesn't swap it. Between 1. b6 and 1. d6, it appears that the bot will grab our 1. b6 if we play it, but not 1. d6 (according to Alan's chart). My feeling is: why hand Twixtbot one of it's favorite moves on a silver platter? If we are going to move first, let's hopefully make a move that we can keep. Bot has only played 1. d6 twice, but has played 1. j6 seven times, so perhaps 1. d6 offers a better chance since less data is available.
Noting Alan's above comments on j6, it appears that if we do that, the Bot will move right in the middle; that seems scary to me. c8 IS a good move, but Maceij is just guessing, since it's never been played; b6 doesn't look like a good move to me. I'd rather play d6 personally. Unfortunately, spd_iv is not too keen on d6, but maybe has been persuaded.
I'm just trying to make sense and summarize the comments I'm reading above, and maybe not doing too good a job at it. I'm going to get the game underway NOW. That gives us 240 hours to move, plus vacation days, and I sure hope we don't blow it all on move #1!
On 2019-12-31 at 14:11, twixter said:
Woah hold on! There is no pass move. Who starts the game is determined by the game parameters. I still think you should tell the bot to move first.
On 2019-12-31 at 14:23, twixter said:
Most players seem to agree on whether or not to swap most of tb’s favorite first moves. Therefore I still advise the bot should move first. If you start the game now mc, then you are the one who decides who moves first. Once the game has begun, with us to move first, we cannot “pass”
We could of course resign and start again if we choose to.
On 2019-12-31 at 14:29, MisterCat said:
I call it a 'pass' move, meaning to 'pass on the right to make the first move'. I agreed with you initially on this David, and still don't mind, but Alan, spd_iv, and Maceij all want to make a move. I am trying my best to go with the most persuasive arguments here, so take it up with those guys. I'll be watching. The game has begun. I hope that move 1 doesn't require all 10 days, and to answer your earlier question - I'll devote the vacation days of my MCx account to this game, as needed.
NOW HOLD ON WHILE I FIX SOMETHING: I accidentally started the game using Mister Cat. This was a mistake. I now need to resign that game, RESTART another under my MCx id, and re-post messages and links over at the Forum. I'll do all this now. Bit of a pain in the neck, but MY MISTAKE. Hang tight. (meow)
On 2019-12-31 at 14:33, MisterCat said:
WAIT, I can't do that, because David is STILL pursuing the 'pass' option, ie. letting The Bot go first. I thought we were done with that. Today IS New Year's Eve, and I wish to get this party started. I can't even restart the game until this is decided, since you select who goes first when the game is created. I'll give you guys a couple of more hours to bat this idea around. As it stands, if nothing new is added, I intend to go first as analyzed above.
On 2019-12-31 at 14:37, twixter said:
Are you certain that Maciej and Michal both understand this is an option? Or do they assume we are moving first?
Maybe you could wait until that point is clarified?
On 2019-12-31 at 14:45, MisterCat said:
Hard to believe that they do not know that, but I assume that they are reading this. You WILL note that Alan prefers to move, as posted above (WAY above). I've cancelled the mistake game, re-posted at the Forum, and now, all I can do is wait a few hours, and see who says what.
David, the idea is not to convince ME to 'pass'; the idea is to convince The Team. I'll have no choice if I MUST make a final decision, but I'm doing all I can to avoid being in that position. Right now, as it stands, it appears that the above discussion points towards The Team making the first move. Let us see if anything changes in the next few hours. mc
On 2019-12-31 at 14:51, twixter said:
Thanks for waiting.
On 2019-12-31 at 16:22, twixter said:
In Poland it’s now 8 p.m. on New Year’s Eve, so you could go ahead and start the game with humans to move if you like. Thanks again for your patience.
On 2019-12-31 at 20:46, MisterCat said:
The deed is done. The game has begun. It is The Team to move first. We have 240 hours, plus vacation days. Hopefully it does not take that long. If you are new here, see the above threads for some analysis regarding move 1.
I will try to hyperlink to the game here: http://www.littlegolem.net/jsp/game/game.jsp?gid=2140447
If that did not work, go to game #2140447, or just see the post at the Forum. (meow)
OK, game has just started. I still think we should play |1.j6 (we are almost sure that TB will not swap it), I also agree to start |1.d6 looks very strong (if TB will not swap it :) ) I would not recommend to start with |1.b6 because there is a chance thatTB will not swap it.
can't we play a second game at the same time where TB makes first move?
On 2020-01-01 at 10:13, spd_iv said:
As Maciej has the highest ranking I think we can simply calculate that his vote counts as 2 and rest of the team counts as 1. What do you think?
|1.d6 was played against TB in games 2039204 (me) and 2092626 (CosimoC), and TwixtBot's responses were 2.d4 and 2.d3. That is the kind of tactical opening that the bot seems to like, but we humans are unfamiliar with. For that reason, I think we'll do better with |1.j6.
Well, the Alan's last comment convinced me to play |1.d6.
My reasoning: 1) Let's play |1.d6 suppose bot answer is 2.d4. 2) Remove in our head the d6 peg. 3) The state of the game is like this |1.d42.swap (note that swap comes from the bot). 4a) We know that TB do not swap the |1.d4, 4b) plus we have a d6 peg for free.
Going back to |1.j6 case: I have analyzed the |1.j62.l133.p134.p18 state by playing several games (me vs me) and all the time black wins. So *) I am not able to see the potential of |1.j6 and how to work with it correctly. OR *) the game would be a tough one...
spd, |1.j62.l133.p134.p185.t186.s157.t138.t209.h1510.l18 ... ? in both cases black player will be the one to force us (by playing sth like 10.l18) to move first at the bottom so he can respond accordingly to the bottom state.
On 2020-01-01 at 22:56, MisterCat said:
You guys KNOW that following the depth of reasoning I'm reading here is quite beyond me. I'm just looking enough to see if I can catch the thread so as to zero in on a move. That being said, I'm glad to see that Maceij is now AGREEING with my comment many posts back - that 1. d6 seems safer than the scary 1. j6. Hooray!
I'm still not doing anything. Please keep up the work ('good'?) so as to agree on a move 1, before too long. I know that this game might easily be lost or won by move 10, so the early moves are real important, but if we use up too much time, we shall have to resort to making it back with a few quick moves (and 36 hour bonuses).
Hopefully move 1 tomorrow. Oh, spd_iv, I will respond to some of your comments from above back in The Forum thread.
On 2020-01-02 at 08:17, spd_iv said:
Thanks Maciej,
I am connvinced that: |1.j62.l133.p134.p185.t186.s157.t138.t209.h1510.l18 doesn't look like an easy game for white :) maybe 5.t18 is not perfect but as Maciej suggested |1d.4 looks much better for me after his explanation. In worst case TB will swap it :)
Alan, will you forgive us for 1. d6 ? I mean, I don't wish to disappoint you terribly. The team has not been persuaded by your 1. j6 reasoning above, and we have spd_iv and Maceij (and me) pushing for 1. d6 . This is after noting your most recent comments. Try not to gloat too much if we lose, as a result!
Within a few hours from this post, I shall play a move, so if there are any last thoughts, now is the time. mc
On 2020-01-02 at 10:10, MisterCat said:
Have not forgotten you, David; you posted while I was writing MY post. I will check back here in a few hours from now, and hopefully this issue is resolved. mc.
What would REALLY be hilarious, is if tb actually PLAYS OUT one of these lines. Like Alan, I don’t know how to deal with this |1.d62.d4 edge attack stuff that the bot favors. That’s the real reason I prefer 1.j6.
Suppose 1.d4 should be swapped, even though the bot does not, to counter Maciej’s argument. In that case it seems likely that d6 should also be swapped. But since TB has not swapped it, again this would mean we would gain an advantage.
Is there a chance TB will swap, despite its past behavior and Mr. Jordan’s disclosure that its rating peaked in June? I don’t really know. Which scenario seems worse, bot swaps j6 or bot swaps d6? Also what if the bot answers d6 with something else entirely?
I don’t really mind if we play d6. This j6 analysis has value to me in and of itself. But I would still appreciate another day to discuss it. Thank you.
On 2020-01-02 at 18:22, MisterCat said:
I'll grant another day. You know that we have the time NOW. We won't have the time LATER, and using this day may come back to haunt us. I hope the team can make it up - just a few moves made in under 24 hours, so the 36 hour bonuses will add up.
I really was expecting to play 1. d6 now. Instead, I will do nothing until tomorrow. This gives ME time to look through all the above commentary, and what follows. As I said, I am trying to get a feel for what people want, and act accordingly. But this means I have to be AWARE of what is being said, whether I can follow it or not. I'll continue to do my best.
black has many possibilities here. I still think we can play J6 but playing as Maciej and David discussed there is big space on the left of the board which can be dengerous for us. That is why I vote for D6.
If TB will swap D6 i think we can play J9 or somwhere in the middle
On 2020-01-03 at 00:32, nie_wiesz said:
For me the scenario where bot swaps d6 is the worse one (because i feel that d6 is slightly stronger). I don't really know how to respond it. On the other hand we expect bot to make mistakes in the early phase of the game. The earliest phase is the very first move. So I do believe that TB won't change his habits and still will be playing |1.d62.d4.
and another good black move would be |1.j62.l133.p134.o10 ! Which is a tough one.
On 2020-01-03 at 00:59, twixter said:
I would like to continue discussion of J6, but in the end it would likely be counter-productive. I probably have more errors in my analysis. So please go ahead with d6, the sooner the better, unless anyone else has an objection. I will be busy tomorrow and may not have time to analyze until around 8 pm Polish time.
On 2020-01-03 at 05:39, Florian Jamain said:
We just should play like 50 games with all starting points, see if he swaps or not, resign and then chose the variation we prefer for the real game.
Don't know if I'm the only one but... I don't think TB is playing strongly to TwixT. We just don't have enough time usually to study the positions, he tries to put the game in a tactical stuff and then he gonna win cause you gonna fail, but even if he is the best in tactics I don't consider him as a good player in strategy, he is playing many bad moves to me in the early game which shows that he does not really understand what is happening.
Maybe it is just the contrary and he is also the best there this is why I don't understand what he is doing :D I don't believe it at the moment.
On 2020-01-03 at 08:41, nie_wiesz said:
Florian can be right in one thing - there are probably some openings which are "better for us" if the bot do not swap it - for example |1.k7. But on the other hand if we could not choose the move from two possible. It would be much harder for us to decide when there will be more of these moves - it could take another few days before we start the game.
If he is not that strong we can easily flip a coin and choose one of the two discussed starting moves. No need to check 50 openings. For me it can be either J6 or D6.
So my suggestion is that we should flip a coin and let fate choose for us ;-)
On 2020-01-03 at 08:51, Florian Jamain said:
I mean you need to take the advantage in the opening cause it is the moment he is not best, so we cannot play anything and say it is okay we gonna crush him :D
After the 3 or 4 moves, if you are lot ahead it's over I imagine, this is the moment we need to do the difference.
But yeah, juste start the game with one of these moves :P
On 2020-01-03 at 09:44, MisterCat said:
1. d6 has been played. We are back to 215 hours. I agree that Bot does not APPEAR to play the opening well at all, and therefor gaining a winning advantage early in the game is the best chance for the Team. I will add that, to me, this move is a more normal, human start, so you guys should know what to do with it. Depending on the reply, of course. (meow)
|1.d62.g4 I agree with most of what Maciej said. But in my vast experience of beating TwixtBot once, I responded to tactical opening moves with tactical moves, and won ;-)
I like |1.d62.g43.m12 a lot, but I don’t think TB would make so many consecutive moves in the bottom half. Just as d10 is a strong resource for white, if black can play k4 or L4 or j5 as part of a battle in the top right, that might also threaten black c4. For example 4.m165.r166.m107.l108.k4 That’s how TB does these edge attacks. They start out looking ridiculous, then they continue to grow until they become overwhelming.
I don’t like 3.h6 because something like black j5 or k4 could still be a threat, and a central move grabs more influence, to borrow Go terminology.
On 2020-01-04 at 01:33, MisterCat said:
Maceij, I do not have the expertise needed to follow your examples. Certainly the other guys do. In the spirit of discussion, I will explain my INSTINCTIVE move to 3. j6 :
This is to me what Alan would describe as a tactical approach; we are immediately hitting the Bot back with a reaction to its move. Bot seemed to like being 3 spots right and 2 spots up, so I thought the same: 3 spots right, 2 spaces down.
White wants to reach the top, so if black blocks on the left with 4. e3, we reach the top with 5. k4 . On the other hand, a block on the right allows us to reach on the left.
Of course, this ignores the whole rest of the board, but I don't know how to think that far ahead. I'll trust that your examples effectively shoot down the move.
In his instructive comments in the Twixt Wiki (still available), Alan warns AGAINST local battles, and suggests engaging in global battles. I would say your brother's suggestion of 3. q15 would be a more global approach.
So Alan, if you follow YOUR OWN sage wisdom, your suggestion of the tactical 3. h6 would be a local battle, and not good. (despite your win, from last February)
Alas, I don't comprehend how anybody could plan far enough ahead, with any degree of certainty, to evaluate a move like 3. q15 being any more correct than a 3. q14 , 3. q16 , 3. r15 3. p 15, or anything else, for that matter. Perhaps somebody will enlighten me.
Now a comment about practicality: there are 22 x 22 - 2 choices available for our next move. That works out to 482 possible moves. Considering what happened in the above analysis of move #1, I fear that if EVERYONE keeps suggesting different moves with accompanying projected lines, this game will be lost on TIME FORFEIT before we can reach a decision! I think that we need to have more of a meeting of minds - perhaps if folks can agree on the TYPES of moves they are searching for, then those 482 choices might be narrowed down significantly.
As far as specific tactical lines, I'm COUNTING on the expertise of the team here to catch outright blunders, oversights, and serious strategical mistakes.
That was a pep talk. GO TEAM!
On 2020-01-04 at 10:43, nie_wiesz said:
|1.d62.g4 Ok, 3.j6 may be a good try but is not the best. *) J6 alone is a weak move. Why? It's considered a 50-50 when it comes to swap. But we are not in swap phase. So it is better to play a strong move such as 3.m12. **) Furthermore J6 do not have a big impact on G4 so I don't think it's a tactical move i.e it would be great if J6 would block the G4 but it doesn't. Why? Black player can easily divide D6 from J6 as he needs (for example: 3.j64.g85.h56.h6 - this is the example of what could be done in the future, I don't really think seriously of 5.h5), which is not good for White.
The 3.h6 is a better try because it has bigger impact on G4. If that's not enough then maybe someone more eloquent than me can explain it?
We have an advantage right now, so probably there are a lot of moves which are winning for us. It can be 3.h63.m123.q15 (and maybe even 3.j6) etc.
Do we still fight for the case? 5.g86.i5 -> don't go that way! 5.f76.i5 -> don't go that way! 5.j7 or 5.i8 -> and there is a black F6 -> J5 threat (maybe this is what David had in mind). It can be really dangerous later on.
Personally I do not like this local battle.
On 2020-01-04 at 11:16, nie_wiesz said:
|1.d62.g43.m12 David, moves like black K4 or white D10 alone are much more scary if they are connected to the edge. So TB should be much more scared of white D10 than we are of black K4. G4-K4 combo do not connect left, so we can block him 4.m165.r166.m107.l108.k49.f5 But if we are so afraid of the G4-K4 we can go D10 first 7.d10. And he can't block us from top edge |1.d62.g43.m124.m165.r166.m107.d108.c49.f510.l411.h4 -> now he gain L4 for free but we gained D10. The problem for black is that L10 is alone but D10 is already connected.
Going back to the reasoning why I do not like the top left local battle. Leaving the top left corner untouched gives us opportunity to respond to it as we like (D10 or F5 or H5 depending on the board state).
On 2020-01-04 at 11:51, MisterCat said:
Maceij, I am not trying to defend the move 3. j6 with my above post; I explained the simple analysis I used, as best to illustrate ALL I DON'T KNOW! I'm a 1700 player, and perhaps with a year or more of dedicated lessons from one of the 2300's about (not going to happen), I might be able to grasp the more advanced ideas I'm reading.
Note - I just played through ALL your above examples of why 3. j6 is NOT good, and they were quite instructive. THANK YOU! I was too sleepy to do that last night!
Continuing to explain my thought process (why?), after 4. g8, my instinct would be 5. n6. Now, we can connect to the top from any peg, and black must try to block from the other side, such as 6. n10. But then white trips black up with 7. j10 ! I looks to me now that white wins the top, so black must take the game to the other side, like 8. i16 . Then the fight starts anew.
Hey - this move-ahead stuff is starting to be FUN! Maybe I'll get better at it! OK, so you see - that's the way I think, totally flawed may it be.
So to summarize, besides my 3. j6, trashed above, we have 3. m12 from David ('center' play, which Maceij approves), 3. h6 from Alan ('top right' play; Maceij does not like), 3. q15 from spd_iv ('bottom right' play, approved by Maceij), and no other suggestions on the table for move 3, for now, unless I missed something above. Great - that's much less than 242 moves! I wouldn't mind a few more suggestions thrown around at this point - maybe some new folks will chime in, but then hopefully we will come together on a preferred move.
189 hours remain, at the time of this post.
On 2020-01-04 at 11:54, nie_wiesz said:
|1.d62.g4 @MisterCat the 3.q15 comes from experience. It has really strong connection to the bottom & it's closer ((*) but not too close) to the left edge than for example 3.o15. So if we play something like this 3.o154.o13 and 3.q154.q13 it feels like the black countermove has bigger impact in first scenario. And 4.q13 do not seems to be strong for black ((*) is too close to the right)
Another thing is that I've often played it successfully (vs humans). At least I feel that way. It can be easily checked by any of you.
On 2020-01-04 at 12:02, nie_wiesz said:
|1.d62.g4 another great thing about 3.q15 is for example this: 3.q154.p195.t18 -> where 4.p19 kind of moves don't work and Q15 is 100% connected to bottom. But for example 3.r15 (is too close to the left) and this scenario 4.q195.u18 do not work here. White need to play 5.u19 which is not that big threat.
Perhaps 3.q14, 3.q16 could also be fine but 3.q15 seems like a good average of those two :-)
On 2020-01-04 at 12:10, nie_wiesz said:
in both comments when I said "left" I meant "right" ;-) i am fine with other direction but left and right sometimes tricks me ;-)
On 2020-01-04 at 12:18, Alan Hensel said:
In my searches a few days ago, I missed this TwixtBot vs. TwixtBot game that BonyJordan posted back in July. Those games have slipped to the 2nd page already. Anyway, that one seems relevant because it starts with |1.d6 (though not 2.g4), the first 7 moves are local battle, and White prevails.
3.h64.h75.i8 looks like a White advantage to me, but I don't think 4.h7 looks like something TB would play. A more Twixtbot-like line might be 4.j75.h116.e57.i48.h8.
M.C., regarding the words I wrote on the Wiki many years ago, in the pre-TwixtBot days: I think it's still good advice for beginners, but it is possible that effectively executing opening tactical moves is an advanced technique.
How is it possible that |1.d62.g43.q15 is closer to the left edge than o15? Did you mean the right edge? And why does this matter?
On 2020-01-04 at 12:40, nie_wiesz said:
|1.d62.g4 @Alan as I said earlier I think we are in big advantage right now. So whatever we choose (local battle or global approach) we would still be at winning position. That's why 3.h64.h75.i8 looks good. But it can be tricky and hard for not making a mistake.
4.h7 is more crazy and scary for me than 4.j75.h116.e57.f7 -> seems to be good for white and not that crazy.
If we know that bot would not respond 4.h7 I can vote for 3.h6
On 2020-01-04 at 13:27, twixter said:
I vote that we make a move soon. There is another clock counting down besides the game clock: how much longer the brothers Celuch will be around to advise us. Maybe we could even put in two moves today? Either m12 or q15 look fine. I am willing to go along with h6 if that is the consensus.
On 2020-01-04 at 13:44, MisterCat said:
Watching, David; I would rather not have to make ANY decision here, if you guys can do it for me. I ALREADY feel a bit guilty about being the one to pick d6 over j6 on move 1! I will check back as much as I can, and when there appears to be a decision on a single move, I'll play it.
Those 36 hour increments are pretty handy, but if we get just one of them today, we should be ok for analysis of move 5. At this posting, 187 hours to go.
On 2020-01-04 at 14:13, nie_wiesz said:
|1.d62.g43.h6 Why I am scared of 4.h7 ? 5.i86.j10 and I will show some examples which shows what bothers me:
7.h118.f69.g910.h14 thanks to J5 thread black has an opportunity to get a free top-right corner move which makes the game complicated from here.
for me the local battle is very risky that we miss sth. I prefer to play 3.m12 or 3.q15. Both moves are good in my opinion so lets just pick one of them.
On 2020-01-04 at 14:51, MisterCat said:
I guess we SHALL pick one of these. If the TEAM won't narrow in to ONE choice in a little while today, I guess I will play 3. m12. My reasoning? Fairness. m12 is twixter's choice,and he did NOT get his first choice played on move 1. I guess it's his turn. Reasoning is hardly profound.
Save me from making this decision, Michael. Convince David that q15 is to be preferred, and I'll do that. I'll check back. (meow)
On 2020-01-04 at 14:59, twixter said:
M12 was not my choice. Maciej asked me to examine it. M12 and Q15 look equally good to me.
On 2020-01-04 at 15:06, DrJochum said:
I agree: M12 and Q15 seem to be equally good to me. I would prefere M12, because it seems more flexible to me (just a feeling).
Here is a line where black blocks j15 with j17 and wins. I did not spot that being tested above. Listen, I know that the line is probably total nonsense at the level of play here, but this is becoming fun. I'm sort of getting the hang of constructing these advanced lines. So, if you would, look at my line, wince in pain at the horror of it all, show me the error of my ways, and I will learn something from it.
Meanwhile, I'll try to construct a line where white WINS with j15, and I'll ALSO see if I can come up with another suggestion for move 5. mc.
Did you see the latest TB slaughter of Florian in only 11 moves? That is the level of play we are facing here, so I feel that NOW is the time for most effort and caution. We can use the hours, use the vacation days, and IF I ABSOLUTELY MUST, I will buy ten more vacation days, should it come to that. (meow)
@MisterCat: |1.d62.g43.m124.m175.j156.j17 I think that in your example, the mistake is 7.g16. Why would we give away left bottom corner that simply and gaining almost nothing at all? 7.f177.e17would be stronger in this part of the board. But of course here 6.j17 we could also play around right side like David said.
|1.d62.g43.m124.m17 I feel like we're fixating on 5.j15, leaving us no good comparison to anything else. It seems clear to me that 5.j15 leads into a mess, and TwixtBot is good at emerging victorious from messes. Be careful.
Alan you may be right but thanks to Maciej and David we can predict which moves TB can play and how to react on them. Now we are analysing only 2 moves so if we agree than one of these moves are good enough we should just play it.
On 2020-01-05 at 18:00, spd_iv said:
I am wondering what is Maciej suggestion on our last comments 5.r16 also looks good for me. I think we should focus only on these 2 moves and pick the better one.
On 2020-01-05 at 19:32, twixter said:
This position provides a good case in point for Alan’s suggestion.
I think black is doing very well here. Even if you find a j15 line that stumps me, there is no guarantee it will stump TB. IMO j15 will be more difficult and risky than 5.r16. Maybe we haven’t been talking about it because there is nothing much to talk about? That is exactly the kind of move I want to make.
At any rate, we have lots of time for this move. MC sounds willing to use vacation days on this move.
So let’s talk about r16 shall we? Show me what is wrong with it.
On 2020-01-05 at 19:44, spd_iv said:
I didnt say that sth is wrong with it. I accept that move and I agree that r16 is more defensive. I just would like to wait for Maciej suggestions.
I didn’t mean to suggest anyone said anything was wrong with r16. It’s just a manner of speaking. I’d rather find out now what’s wrong with it, instead of finding out from TB.
On 2020-01-05 at 23:30, nie_wiesz said:
|1.d62.g43.m124.m17 5.j15 may be good but I agree with Alan, it's possible that it can go out of control - especially the top side battle. So I give up on that for the Team sake.
5.r16 - really nice try David. It's much easier to control the game state from this point (at least for us humans). There are too many variations on bottom right corner and each one gives something different to black. So probably TB will attack on different side of the board. Like 6.l46.k46.o116.m9. Because it's hard to predict the very one I think we should focus on local variations to be prepared of the threats arising from this.
5.h14 seems fine too. @Spd, how would you respond to 6.h17 or 6.h19 I don't really know what else can black do except 6.k4
On 2020-01-05 at 23:37, MisterCat said:
190 hours remain as of this post. Yes, spd_iv, I encourage you guys to take the time you need to hopefully ensure that we avoid getting wiped out. Besides the hours, the account has 20 vacation days, and then if I kick in money, I can buy 10 more.
But 190 hours plus 30 days is still not infinite, so let us continue to be mindful of time. I love those 36 hour increments - these well come in REAL handy.
Maceij - thanks for the lesson above, which I shall study. Alas, I went wrong with THE FIRST MOVE in my hypothetical line; it may be a move that players like MYSELF would play, but you explained why it is flawed; it is NOT a move that the strongest players would play. But that, I think, illustrates the limitation of long, hypothetical lines - like you guys are using regularly; just ONE deviation from the line, and you've got a whole different game!
I'll do the best I can to keep up with the discussion here - I DO have a day job, you know - and if The Team decides on a move, I'll play it as soon as possible. mc
Oh, that was David who mentioned vacation days above; my bad. I'll use them as needed. For what it's worth, I prefer r16 of the two choices being discussed, and I'm agreeing with the reasoning above - it will be less complicated for The Team to deal with a battle on the right side of the board. ( o o )
Sneaky, Alan; I'd call that a 'mousetrap' - allowing black to reach the right side, knowing that we can connect to the bottom from m12 to m16 and down. Of course, we could have played your i16 right at the start, without that big construct on the right - but maybe it gives us an advantage in connecting to the top.
What I find a bit sad is that Maceij would SEEM to demonstrate that, with ONE LITTLE peg placed right of center (namely black's m17), is is NOT POSSIBLE (?) to prevent black from reaching the right? Meaning - if black were at l17 instead, we COULD stop him, but at m17 it can't be done?? Philosophically, that doesn't sound too swift for the overall 'truth' to the game of Twixt! What I mean is that games will be over - won or lost - theoretically after only 5 moves!
OK then - if Twixtbot felt to block US by placing a peg directly in front of us leaving 4 empty holes (the block from m12 to m17), might we not do the same to it? Block black's path to the left by placing a peg left of m17 with 4 empty holes, namely 5. h17 ? The 'truth' is that h17 can not be blocked from reaching the bottom, and we can aim to connect to the top either by heading for our m12 peg, or straight up to our d6 peg.
There - just some musings, since my efforts to generate those well-played hypothetical lines you guys use seem to fail. meow.
On 2020-01-06 at 07:55, twixter said:
Alan, nw was just talking about local tactics. Your line shows why TB won’t do that on the whole board. Suppose black plays |1.d62.g43.m124.m175.r166.k4 if white just protects with 7.f5 then black might jump into the center with 8.i12 and now the threat of black q14 has become a lot stronger.
I think we should try and play 5.r16. 5.h14 was just an exapmle and I did not anylise it much, but if team decide to play 5.h14 i am fine with that also :)
Yper please do suggest anything you want. Please paste or type into each new comment the entire line you are talking about from the beginning, including the vertical bar at the start. In this case |1.d62.g43.m124.m175.q17 After you do this, you needn’t do it again in the same comment.
I do not see a special difference here for having H14 and there is a room for a mistake (there is a chance that TB will make better use of h17 than we of H14). So I vote for 5.r16
Easy question please - why do you guys keep playing f5 in the upper left in response to k4, instead of playing e4? I mean, with f5, black reaches the left side via, |1.d62.g43.m124.m175. r166. f57. e3 and white is blocked. Is the idea to connect with the center pegs? But if so, white is STILL blocked! On |1.d62.g43.m124.m175. r166. e4 , white can at least play for the upper left corner.
My head is spinning, trying to wade through the advanced lines being posted. Unfortunately, it appears to me that my earlier remark (about black having the advantage on the right side) is being shown to be true. I still think that's a shame, and wish there were a way for white to deal with that m17 peg.
But that being the case, I guess you guys are broadening the scope of your analysis, and playing lines that go ALL OVER THE BOARD! Great, I guess. I'll keep following, and if it appears that a consensus is being reached, I'll move asap. meow.
I posted a useful tool back at The Forum at Little Golem, which I asked Alan to make for us. You may find it useful, so check out the post. mc
On 2020-01-07 at 13:48, twixter said:
All right. Against |1.d62.g43.m124.m175.r16 I don’t see a good line for the bot. Thanks for your patience. Let’s make the move.
On 2020-01-07 at 13:57, nie_wiesz said:
|1.d62.g43.m124.m17 Ok, we rejected a territorial move like 5.j15 because of its complexity. It's difficult to determine the winner of 5.h146.h177.r16 vs 5.r16 battle. So (against bot) I prefer simplicity like 5.r16. No one has suggested any left blocking move like 5.h165.h175.j17 etc. So I assume that we do agree to play a right-side blocking move like 5.r16 or 5.q17, right?
5.q17 is not that simple because it leaves a hole between M12 and Q17 (itself) which gives a lot of possibilities for black - straightforward attack with 6.p14 or leaving P14, P15 as a threat. And its resistance to M21 is almost the same as 5..r16 has ---> 5.q176.m217.o18 vs 5.r166.m217.p17. I'm not 100% sure that those cases are good or bad for white. I just feel that r16 is a better move.
5.r16 does more or less the same as 5.q17 and do not have a hole between M12 peg and itself. That reason alone is sufficient for me to play 5.r16 even though 5.r16 has its own weakness points like for example 6.j20.
I could have some other suggestions for 5th move but I feel like it has no sense to share my thought right now because we would never make a decision :-)
Hey Maceij, I DID suggest the 'left blocking move' 6. h17 way up there, in this thread! Nobody said anything, and of course, I'm not strong enough to accompany it with a useful hypothetical line. I'll forgive that.
I see you guys getting real close on a move now. Instead of making it, I'll go to my therapist appointment. I'll check back in about 2 hours, and if the status remains the same, then I'll make the 5.r16 move as instructed. mc
On 2020-01-07 at 14:54, nie_wiesz said:
|1.d62.g43.m124.m17 , ohh there is too many posts here. It's easy to get lost. I admit that I didn't see your post. Nevertheless i see it now and I like 5.h17. Especially considering one of your posts where you shared your thoughts about mirror type of moves :) 5.h17 is a good try - would be hard for me to reject it because I also like it. As I like 5.r17. But I have some inner self that tells me not to touch the left side. Butr 5.r17 changes the strategy we talk about by 90 degrees... We made a lot of effort to R16 so I'm sticking to it.
On 2020-01-07 at 15:26, spd_iv said:
|1.d62.g43.m124.m17 if we play 5.h17 then probably 6.h107.q12 and 8.m21. I don't like this position so if the Team has a good response for 6.h10 (or any similar move) then we can wait one or 2 days to make more analysis, but I think we should go for 5.r16 as we know much more about it and we lose less time :)
On 2020-01-07 at 16:27, twixter said:
Sorry MC, I didn’t see your suggestion either. H17 might also be winning. But besides the time it could take to analyze, there is also an emotional factor. If we switch to studying H17 it could take the wind out of our sails. Well mine anyway. There are probably several other ways to win but if we are confident with R16 I think we should go with it. Just my opinion of course. We can’t build an entire encyclopedia of openings in one game.
OK, guys. Thanks for the kind nods from Maceij and David. Welcome to the party, Florian!
I will play r16 now, by popular demand. If we lose this game, then next time we'll try something else (if there IS a next time). I was out for 4 hours, but I will move right now .......
There are 183 hours remaining. Expect a reply from Twixtbot within a few minutes! With luck, the pages and pages of analysis above will be proved worthwhile, if it makes one of the predicted moves. Then, much of our work will be done already for the next move!
Then my suggestions are 7.t207.v207.u19 and 7.o19 There is a counterplay to7.t20 (a way to use R20 against T20), isn't it? I do not remember right now so leave it as a suggestion.
Now if you look, it may be true that white reaches the bottom ie. 15. p1716. p15 but what good has it done us? It appears to me that blacks pegs at p15 and o13 give him (it) pretty good chances of blocking our lowly peg at r16 from ever MAKING IT to the top!
But maybe I'm wrong. We need the Team to analyze and post some more concrete hypothetical lines at this stage. (mc)
well, what I meant by "sews up the bottom" is not simply that white could reach the bottom, but it seemed to me that it left black no viable path along the bottom half of the board -- guess I need to go digest some of those hypothetical threads (how DO you guys do that???)
all understood, Maceij; having simply collected the lines above involving r20 , I responded to Bob440 by quoting your post; I had not looked far enough to see twixter's subsequent analysis.
Bob, as you see, my 'critique' was not quite right, since Maceij himself corrected his line. Sure, perhaps Black has no viable path along the bottom, but it's still a big board; white has to find a way to reach the top!
You and I are discussing strategy using exposition of 'general principles'; sometimes these guys do that, but more often their analysis technique is to 'play computer' and compose SPECIFIC lines using a sequence of good, plausible moves. The Commentator diagram at the left makes that easy to do - just play out a sequence on the board, see how it looks, and if it looks good, POST IT! Then, wait for the crowd to shoot you down!
When I try to compose a sequence of future moves, I usually make a blunder within my first few moves - completely negating the value of the rest of the line. It helps if you are a real strong player, since it's more likely that your sequence will become the truth. But it also helps to be real, real careful and try, try again. THESE GUYS ARE PUTTING A LOT OF EFFORT INTO THIS!
Look at what's happening here - the vast majority of lines posted tend to be trashed anyhow; there were VERY FEW posts that considered Twixtbot's latest r20 move. So if you are inclined, post a sequence of moves and feed it to 'The Star Chamber' to be chewed up and spit out. And, if you hit on something that wins the game for us, wouldn't it be great!?
If these moves are possible, then it is probbly better to keep it as a threat and play more in the center of the board, then or we gonna play one of these moves or we can go on the left and our center is better.
If o19 is working, and I don't know if it is, then it is maybe a good option.
To illustrate what I mean, let's see some Maciej analysis, I can see that l15 is interesting in some variations, so is it possible to play it directly or it will allow TB to defend everything with some back up moves? (Or another move like this.)
guys, I will be available on Friday evening/Saturday morning. I hope you will not make any mistakes :)
I like the idea with getting the top with sth like |1.d62.g43.m124.m175.r166.r207.i12 or 7.h14 but 7.t20 looks also not bad (by giving the initiative to TB, Tb will probably play somewhere in top-left corner of the board)
On 2020-01-08 at 14:02, Florian Jamain said:
Sure David! I did not really analyze, was just to point that playing in the bottom right seems small to me, the Bot gonna have the opportunity to play somewhere else. Keeping it as a threat and reinforcing the middle seems better to me.
Surely i12, simple move but quite strong I guess, is one of the moves I wanted to talk about.
We should analyze more and we find a strong move in the middle I am basically convinced that it will be better than a move in bottom right.
155 hours left now; that is 6 days and 11 hours. mc
On 2020-01-09 at 00:37, twixter said:
I therefore conclude that white should play in the lower left after all. For example after |1.d62.g43.m124.m175.r166.r207.t208.p69.k7 White has secured the top, and has a tempo threat on the bottom right, as well as threatening i16. So, it would seem black doesn’t have time for a downstream peg in the top right. By the way, this T20 move is called the Medcalf defense. One way for black to play locally is 8.s189.t1510.p15, but 11.p17 looks like a good local response. On the full board, how about 8.m10
7.i12 and 7.l15 are mutually exclusive so when we get one of them we won't have the other one (or it won't have a desired effect). 7.i12 have pretty good connection to top and divides the black player into two halves. But if there exist straightforward attack from black, then TB will go that way: |1.d62.g43.m124.m175.r166.r207.i128.i179.o1910.p1411.o1312.q16 - black wins since we can't use L15 as a helper - so we need to play 11.p15 I do remember that there was some analysis where we used L15 as additional help to get bottom. So if we would like to play 7.i12 we have to be sure that it is not an instant win for black: 7.i128.i179.o1910.p1411.p1512.n1513.q1314.q1215.r1116.r1017.s918.r519.q1020.l4 -- black straightforward attack (seems white?) 7.i128.i179.o1910.o1411.p1512.p1613.p1714.q1315.l15 - again we can't use L15, so we have to let the opponent fight for the right side. 7.i128.i179.g17 - maybe whites hope 11.r1212.q1313.q1414.r1115.s1016.s917.t818.s4 -- black straightforward attack (does I12 is strong enough to actually gives us the top side? - that situation is almost forced by black so we should analyze it if we want i12) Another options to check are 8.n208.j208.i20. The 8.p6 probably won't work here 9.e1210.c411.n6.
7.l15 - has extremely good connection to R16 but it doesn't provide us a bottom connection nor do not help much on the top side. So moves like 8.j208.p6 are possible.
To be honest I feel like both 7.i12 and 7.l15 can be winning but I propose the right bottom corner. If you don't like 7.v20 or 7.t20 then I vote for 7.o19... When I checked the 7.o198.p6 my first idea was 9.i12 ... Maybe we should investigate harder ;-)
Maciej, I agree with you that move 7 should be bottom right. In my recent posts I show how white i12 or L15 can both be refuted, and I would be glad to demonstrate further. I also agree that if white does play bottom right on move 7, then black doesn’t have time to play p6.
As we said earlier |1.d62.g43.m124.m175.r166.r207.i12 is resistant to moves like 8.p6 thanks to 9.e12 and I can't imagine how black could break through this big wall on the 12th row. Maybe some kind of 10.h20 or 10.f21... But not the 8.p6 is what I am afraid of but the 8.j208.i20!
Of course it's only a strategy (I am only a human and don't know if white could do something better in the process), just trying to find any weakness points for the I12.
So it seems that it's quite hard for me to reject (7.i128.j209.p19 is an option for white) or approve the 7.i12 - probably lack of skill ;-)
The 7.o19 is another story. But we couldn't find any black response for that either.
Maybe we should step back and consider how we should spend our time here.
I think Alan would agree, the variation |1.d62.g43.m124.m175.r166.r207.o198.p109.i12 is a mess. I keep looking at 8.p6 and finding nothing but doom and gloom for white. So maybe we should shift to stuff like 7.o19 or 7.t20. If that also gets messy we might have to revisit i12, but a clean win is ideal, yes?
As I said in my very first message, personnally I don't believe in t20, v20 or u19 right now, only o19 was for me something and only if it is possible, cause it is not an "all in" on the right side, it gives some strong threat in the middle/left, for example with your i18. To be honest in a very high level game, I reject almost directly the three moves t20, v20 or u19 by intuition, it's like if my brain is saying "it cannot be the best", I hope it's not bullshit from me :D
With this analysis of Maciej, I think o19 is playable and my vote goes also for it.
Basically I am for nothing, I just wanted to say that playing in bottom right is very probably not the "best".
Then I would consider o19 AND many other moves in the center that we did not even checked but we tried 2 important ones and it seems that it was not really working.
But finally, if o19 is playable, it just seems to be the best, it is "between" all the moves we considered, it gives many opportunities. We still can go from the right or from the left and use in the end of variation the o19 peg to connect to the bottom.
On 2020-01-09 at 19:54, twixter said:
We have time to look at various responses to |1.d62.g43.m124.m175.r166.r207.o19 , for example how about 8.m20 This seems to be a useful tempo move which blocks o19 from connecting to anything on the left. After 9.p21 (not necessarily the best) black could play 10.p6 with roughly the same plan as before.
It's not like I think this is a bad move 7.t20 - (perhaps there are not non-messy moves for us) but I feel the same as Florian - "that playing in bottom right is very probably not the "best". And finally I see the the Clear difference between 7.t20 and 7.o19. Thank you!
At this point we should play o19, we're losing time and maybe we gonna need it soon.
As expected bottom right seems not good, we could not find a strong move in the middle neither, so... o19, the move between everything stays and seems to be the good one.
Is it something? I mean at least it counters your j20 for black not like i12. :D
On 2020-01-10 at 15:00, twixter said:
Florian do you have Jtwixt? If you wish, I could send you a file which looks at |1.d62.g43.m124.m175.r166.r207.h148.h17 Or you could scroll back and look for previous discussion of this line. Or, you could show us a specific variation. Like you said, time is passing. We need to work on all branches that look promising.
Maciej, against 7.o19 my preferred response is 8.m20.
The position is unclear but we have many possible moves there. I don't know 9.k19 for instance? Or something else on the left which gives a threat on k19.
On 2020-01-10 at 15:39, nie_wiesz said:
David, if you don't mind I would also like your jTwixt file. I do not remember all the lines ;-)
Meanwhile, 7.t208.o169.q910.o1111.g16 carries the the threats of either reaching d6 or q9. White might be able to play u9 at some point, which threatens white T15.
|1.d62.g43.m124.m175.r166.r20 I am starting to getting lost. Too many posts. I have problem of the identification of what is the aim of our discussion. Maybe we should gather all the lines that are for and against the particular move. David I do not use jTwixt for now so I have problems with remembering all the stuff :)
Thanks for that nice summary, Maceij - that way I didn't have to do it! Looks like we're giving this another day, but with 106 hours left NOW (4 days 10 hours), we may have over analyzed this one move. I HOPE that, once we play a move our analysis proves fruitful - somebody accurately predicted TB's reply, so we'll have a good head start for move 9. We could use a couple of those 36 hour increments!
I'll check back tomorrow. By that time, perhaps Michal will have returned to sort out the mess. (meow)
I'll try to explain my point of view David, the fact is I am not "anylizing" with lines like you, in my opinion it does not make any sense, let me explain:
The best move, but only if it "exists" is arount i12, cause it keep the threat bottom right, add a threat bottom left and reinforce the center. Now, is this move existing?
We tried 7.i12 but it failed, why? Cause of the 8.j20 from Maciej, this move defend move threat on left and right, so here I STOP to analyze, it is useless to continue cause the reason I played my previous move is checkmate, maybe it's still ok but it means that I have better.
Same, we tried 7.t20 directly, already I don't like, it is a "direct" approach, try to win fast when center is not strong enough to do it. Just 8.p10 and I STOP my analysis here, why? Cause I prefer the previous position, it is useless to continue, I prefer white here |1.d62.g43.m124.m175.r166.r20 than here |1.d62.g43.m124.m175.r166.r207.t208.p10 so if I would have an interesting variation there, I just play it before t20.
We know that we have 7.o19, a move that provide many possibilities cause we can go to it from left and right sides. We did not find the good counter for black at the moment, it seems interesting. 8.m20 is clearly something, but after something like 9.k19 same, I STOP my analysis cause I don't think this was a good response for black, white seems better than before the 2 last moves.
Now 7.h14 which seems nice visually, we have this problem: 8.h199.o1910.m21 and I STOP my analysis, why? Cause I don't like white last move with this black response, it seems that the position was better just before, so I would ask before saying 7.h14 is to forget, what about t20?
Now we have h14, so t20 is more relevant, maybe it is something Maciej or you have also a bad variation for this?
On 2020-01-11 at 11:15, MisterCat said:
It took some time and effort for me to look through the above, with lines I really did not understand or follow, to summarize. I did what I could, and correct me if need be.
Of these, the most positive consideration appears to be with 7. p19 , 7. o19 , 7. t20 , 7. i15 .
The blocks on the right are 7. t20 , 7. t17 , 7. v20 and 7. u19. Of these, the most analysis appears to be on 7. t20 . If nothing else, then there is a greater chance we have included TB's next move in the analyzed lines, giving us a head start for move 9. I just hope that 7. t20 is effective in blocking or giving us a connection up and down; otherwise to me it seems like a waste of time.
The tactical interference moves are 7. p19 , 7. p17 and 7. o19 . 7. p17 is not favored, 7. o19 is OK, but more serious attention is devoted to 7. p19 , so there'd be a better chance of predicting TB's reply. However, 7. o19 has fans, like Bob and Alan, and the rest of the people are starting to come around. I really have no preference between these two, and personally I'd rather see one of these interference moves than 7. t20 .
The 'profound', whole board moves are 7. h14 , 7. i12 and 7. i15. People are afraid of TB gaining an edge on the lower left, and this is why 7. h14 is preferred. Those lines are pretty wacky, and I do not know how to visualize a board 20 moves ahead, with stuff all over the place. If we are here to have the most FUN (at higher risk of losing), then I like 7. h14 . If we are trying our utmost to win at any cost, then I prefer one of the blocks, and if you want me to pick one, I'll go with 7. o19 . Why? This was an 'instinctive' choice, before all the analysis, and I'd like to put some faith in humanity.
I hope this helps narrow down the discussion. The remaining time is now 95 hours, or 3 days, 23 hours. I advise that excess time used NOW is REALLY GOING TO HURT US in these next few moves! (roar)
But it's interesting cause if a move like this works, then white gonna probably win the position.
On 2020-01-11 at 12:45, MisterCat said:
So if you saw my post David, the discussion of 'rating' is a moot point, and irrelevant to this discussion. Remember to view this whole game as an 'experiment', more than a 'contest'. If we win, something will be learned, if we lose, something will be learned, and wouldn't a draw be wild?! I plan to discuss my thoughts on this more, sometime in a Forum post.
Looks to me above, that you are moving AWAY from h14 and starting to campaign for T20. I guess you are not fond of the blocks at o19 and p19 any more? The Celuch brothers kind of liked these moves, as I recall. Let's hear from more people, now.
I should point out that, while Maceij has a STELLAR record in games against humans, he remains the only player among the top 10 who has NOT had the 'pleasure' of being trounced by Twixtbot. There is something to be said for experience. I'm delighted he's participating, since in his case, he is not out for vengeance!
Florian likes h14 FIRST, and followed by T20 IF the bot responds to h14 with h19. Otherwise, back to the drawing board. I THINK David likes T20 straight away. Perhaps another of my silly questions, David, but if 7. T20 and the bot simply pushes right with 8. T19 , then I don't understand how T20 has helped us. In that case, we are STILL going to be playing to block with O19 or p19 , or play on the left side with h14 (or my h17 - heh). Certainly we can't allow TB to win easily across the bottom.
And if T20 first and h14 second IS THE SAME as h14 first and T20 second, then I believe we can 'flip a coin'. If we are including a blocking move, then we flip a 4 sided coin. Really - we can agree to use a result from a racetrack of sports score, if it comes down to that. mc
I have a feeling that current move is an important one. Against any human it would take me about 10 - 15 min to decide which one of those moves to play 7.i12 or 7.o19 or 7.t20 depending on my mood of the day.
Whatever you choose I'm fine with it. It could be either 7.t20 or 7.o19 or anything else ;-)
I will have some break now so good luck guys!
On 2020-01-11 at 14:08, MisterCat said:
Florian, thanks for the explanation. This WHOLE experiment has really been a way for me to get some FREE LESSONS! Ha -ha!
We are still looking then, at coin flipping? T20, O19, P19, H14 .
Eventually, I go out to do laundry, so a few hours will go by if no move before then. Relax, Florian - while not the route we want to take, those vacation days WILL click in if needed (20 of them), and all we need are a few quick moves - at 36 hours bonus per move, we'll work our way up to 240 hours in no time; provided the game is not over by then (LOL).
92 hours remain. That's 3 days and 20 hours.
mc
On 2020-01-11 at 14:20, Florian Jamain said:
Then we wait Michal and Alan, they give a move and we play.
At this point, if I had to play now something, I would pick o19 probably. It's sad that we could not find a move around i12, maybe h14, maybe something else that could convinced us.
The funny part is that I 100% agree with Maciej when he is saying "Against any human it would take me about 10 - 15 min to decide which one of those moves to play 7.i12 or 7.o19 or 7.t20 depending on my mood of the day."
On 2020-01-11 at 16:55, nie_wiesz said:
Just for info: My brother won't have time during the weekend. So he won't be available until Monday.
On 2020-01-11 at 17:11, Florian Jamain said:
Michal told me on facebook he will check tomorrow!!!!!!!!! :D
On 2020-01-11 at 17:52, twixter said:
Maciej’s last line against T20 seems to bust my plan. I must accept that his positional judgement is more compelling than my emotional commitment. I accept o19 for our next move, and I apologize for slowing the process down.
On 2020-01-11 at 18:56, Alan Hensel said:
I feel most comfortable with o19. I'm increasingly uncomfortable with the clock.
On 2020-01-11 at 22:20, Florian Jamain said:
Well, then just go, we play o19 I guess.
On 2020-01-11 at 22:27, MisterCat said:
Okey doke - O19 it is; about to play it ... Done; I expect a reply with 15 minutes.
I hope, I hope that Twixtbot plays a move we've done some serious analysis on already; that will speed up move 9 for us.
Thanks to Bob, for suggesting O19 FOUR DAYS AGO! That's how long it took everyone else to come around. No, I'm kidding - I know that other people looked at this move, and lots and lots of work has been done.
Still, if we win, thank Bob. And if we lose, blame Bob. Ha ha ha ha ha ha ha ha ha ha ha ...............!
Did anybody predict that? Did we do some analysis already?? One needs to go back and look.
Now, we have 120 hours left; that's exactly 5 days. This should be doable, but I think, maybe, a lesson has been learned regarding time management.
Back to work, Team! (mc)
On 2020-01-11 at 22:46, Florian Jamain said:
Well...
I don't like this cause now the game will be massively tactical after this move, it's like the strategic part is over, so I hope we are ahead cause it will be harder.
We can check 9.q8 first, which is probably failing but still need to analyse it.
And again, we have all the moves around i12, we could not find a good one previous move maybe now...
I did the research. Bad news. Nobody. Nada. Nothing. Back to the drawing board. Either L8 is too brilliantly profound for anybody here to have fathomed, or it is flawed, in which case we crush it! There may be no law of excluded middle, so possibly neither is true.
Time becomes more of an issue now, since we are starting from scratch. I'll go get my scratching post. (meow)
Florian throws out 9. Q8 . I'll throw out 9. H10 . My turn to actually contribute a move. There is nothing profound about my move, and I'm certain that everyone sees what I have in mind, so I won't elaborate right now. mc
On 2020-01-12 at 00:20, twixter said:
Black L8 cropped up in analysis of |1.d62.g43.m124.m175.r166.r207.t20 Black also had a peg on G8, and in those circumstances, white was locally able to play D11. So with 7.o198.l8 on the board, we might consider 9.D11. This seems more forcing than white H10. On the other hand, 9.d10 might coordinate better with M12.
It’s like Maciej said. Black has more reason to fear white D10 than white should fear black K4. Black would have to try to block below d10 somewhere. 9.q8 might be a good move, but black could respond in many places on the right side. (Did I get that right? Yeah) With fewer black responses to consider, maybe we can find a plan with more confidence of success.
On 2020-01-12 at 00:23, nie_wiesz said:
|1.d62.g43.m124.m175.r166.r207.o198.l8 Amazing, he really surprised me! TB has started to create a mess. Is he really in trouble? Or it's just his normal I feel it's just the beginning of the true messiness.
Long time ago I said something like this: "if black plays on the top half, then we respond D10".
If |1.d62.g43.m124.m175.r166.r207.o198.l89.d1010.e8 then 11.f9 seems to improve threats to the bottom AND to reach m12 without losing connection to the top. Maybe black will be able to play E8 at some future point.
There are examples of bot vs bot games where one side seems to suddenly turn and smash into a brick wall. Maybe we got lucky here.
10.l411.h6 : - 12.i1113.f15 - pair L4 - L8 is not connected to left - 12.i1113.t7 : we can still play on the right against L4 (T7 and R16 like each other thanks to a connector T11/R12 and T7 and M12 like each other thank to P9/Q10)
M.C. I will check the 9.h10 later - if it's not connected to the top then it's not a good try. If it is connected to the top then it could be a good try;-)
On 2020-01-12 at 10:04, MisterCat said:
Thanks;
"- even though white wins it's not an easy win." Somebody is expecting an 'easy' win against the silicon monster??
OK, I started looking at lines with my 9. h10 . I'm not going to post one riddled with terrible moves, but I'll tell you what I saw - h10 connected to d6, d6 connected to the top, h10 connected to m12, everything going great only m12 had problems connecting to r16. Of course, we don't HAVE to go through r16 - there is a nice big space on the left side to work with.
Looking forward to your analysis. mc
On 2020-01-12 at 12:44, spd_iv said:
example of easy win:
https://www.littlegolem.net/jsp/game/game.jsp?gid=2133161&nmove=7 of course black made mistake in move 8, but after 7 moves I believe black had 100% win.
I think TB is not a real monster, he still makes mistakes like in the game with passenger and I belivie he is in trouble right now. I doesn't matter if we play now 9.d10 or 9.h14 in both cases we already win the game easily :)
If someone has different idea, then probably we will waste the time to find "the best" move. I think we should make a move in 24 hours from now. Just everyone please vote for your move.
On 2020-01-12 at 21:11, nie_wiesz said:
|1.d62.g43.m124.m175.r166.r207.o198.l8 Michal, I agree that TB seems not to be perfect, at least his very first moves. But the further the game goes, the closer TB is to perfection.
9.d10 is a strong move but it can make things pretty complicated (for humans). Example in my previous post. 9.h14 is also pretty complicated - I really don't know what to think about it. See example: 10.l411.f512.h1013.t7 ... ? 9.q8 - as most of moves in the right top corner - is sort of "all in" move. I mean if black somehow connect to the right side and in the process will get the L4 then the battle in the left would be very hard. See example: 9.q810.q1211.i1212.l413.f514.h1015.f1116.e917.e818.g10 9.h10 - if it's connected to the topside then it can be a strong move. We should check for a trap here. See example - the trap: 9.h1010.f611.d1012.e813.i714.l415.g316.h217.c818.d2 (David this is similar to your previous line for 9.d10).
Maybe we should follow the plan: 1) Look closer to 9.q8 and check for a quick win. If there is not such thing we should check next move. If there is a quick win, we just play it. 2) 9.h10 if it's connected to the top we should play it, if it's not then go to the next move. 3) 9.h14 what if black plays top middle 10.l4? is 11.t7 good enough to stop black from connecting the right (example9.h1410.l411.f512.h1013.f1514.f615.t716.q10). If it's strong enough we should play it, if not the go to the next move. 4) if we are here then we should play 9.d10
On 2020-01-12 at 21:27, nie_wiesz said:
In previous post I have shown some my concerns about each of the candidates (in the examples) except for D10. D10 is complicated but it's strong (or we will find something that shuts down the D10). So if there is a democracy in this forum then I vote for D10.
If I were the TB, in response to 9.d10, I'd play 10.e12
but I can't take it any further than that
On 2020-01-12 at 22:38, bob440 said:
by the way, I will accept neither credit nor blame for 7.o19
someone else suggested it (Alan?); I merely chimed it to say (likely erroneously) that it looked like it sewed up the bottom
On 2020-01-12 at 22:54, MisterCat said:
Maceij, thanks for the continuing kind words for my h10 suggestion. The move has yet to be carefully analyzed. If we end up playing it (and winning), I shall be delighted, but also nervous as all hell that it will lead to a loss. I may take a stab at posting a (flawed) line or two, just for show.
There are several other candidate moves being thrown around now, including two from your brother, and a different approach from Florian. Don't think of this as a 'democracy' or 'government'. Think of it more as a 'think tank'. This game is not a 'contest'; it is an 'experiment'. I am counting on the discussions to narrow in on basically acceptable moves - and if not provably the absolute best, then at least not flawed, and acceptable to The Team. No voting, no ratings.
Don't bother with the hypothetical question for me: what shall I do every every single Team member posts a different preferred move, and with rational justification to accompany it? While I hope that does not happen, I WILL deal, and without looking at ratings or votes. I have some other ideas, should the need arise, that I'll keep to myself for now.
spd_iv, I expect to NOT spend 5 days on this move like we did the last one, but it's only been one day so far. We are still all right with time. I will do my job seriously, as this whole project was 'my baby'; of course, I'm delighted at the enthusiastic response. I will NOT allow the game to be lost on time forfeit.
Team: I believe a 'lesson learned' from the time spent on the last move is that it is ALMOST USELESS to attempt to predict the response from Twixtbot; at least for now, on a wide open board. Thus, many of you analyzed and posted long hypothetical lines that turned out to be trash. In order to get our time up, I believe MORE effort should be devoted to predicting the response from Twixtbot, so we have a reply already analyzed, and LESS time should be spent on analyzing lines 20 or moves deep, which are bound to NOT HAPPEN.
That's just a suggestion; naturally, analyze however you like! At this moment, we have 96 hours - that's 4 days, plus 20 more vacation days, if we need them. (Master of Ceremonies)
Bob, thanks for your post, and I said I was joking; I also saw Alan's earlier post. And here, you did exactly what I asked for above - you read my mind! I think that we've got to try to be more accurate predicting the bot's reply, so that's what you've done!
So, for example, The Team can analyze now whether e12 busts d10; if it's a bust, eliminate d10; if not, consider other candidate moves for The Bot in response to d10, so we are READY with concrete analysis for move 11. (meow)
Maciej looked at 10.j20 and I looked at 10.g21. It seems to me, black is forced to try to block white’s threat to win on the left. I welcome any suggestion where else black might play. Please poke holes in our analysis.
Please first show me what is wrong with D10, that it needs improvement, and then I will look at Q8. Maciej and I have posted several D10 branches. We cannot keep taking so much time to move any more. I’m sorry but that is the reality we face.
On 2020-01-13 at 04:56, Florian Jamain said:
Like Michal, I don't think it needs it cause I think white is winning this, and I believe d10 is one of the possible winning moves, morevoer d10 is really safe, we are not taking risks with it.
But we need to keep in mind that if we forgot one very complex variation, TB will probably play it, so if we have "safe" little improvents we need to consider it.
The idea is just to use q8 to fortify the middle and then play the d10 move, I don't know if it's possible and moreover if it is possible without risk. Cause this is for me the great thing of d10, it is very safe and "seems" enough to win, but vs TB I'm still anxious that we are trying to win too easily and he has a possible very complex line that crush us.
This is why I'm just trying to see if it is psosible to "prepare" a little d10 staying safe.
(I see lines where we play until t15 before d10, but all olf this is maybe risky.)
On 2020-01-13 at 06:11, spd_iv said:
I will write some of my analysis this evening (after work) about D10, Q8 etc..
On 2020-01-13 at 06:18, twixter said:
You’re worried there might be a flaw with D10, but you don’t want to search for it. You’d rather look at a “likely bad” move on the chance that it might be better? That’s not what what I would call efficient use of our dwindling time.
I'm okay with the fact that l4 is enough to defend this and we lose our d10. I don't think it is better to play something like this and then f5 than playing d10 directly.
Actually, I'm okay for d10, we still wait a little, Michal will give some ideas tonight?
I'm paying attention to the back and forth, and the anxiety with time. D10 or Q8 . Now P9, seemingly with a win (cute - I played through that). Of course I haven't the skill to REALLY make the best evaluation. In fact, I just was about to make a comment, but removed it. This is beyond me.
I have to go to work, but I can come back. I'll try to get that move in TODAY, and before too long. A few more comments will be nice. mc
On 2020-01-13 at 11:36, twixter said:
I wish someone would look at the D10 line. No one has yet posted a single challenge to it, other than to say maybe TB will find something against it. I find this behavior illogical. |1.d62.g43.m124.m175.r166.r207.o198.l89.d10 For move 10, we have looked at black moves e8, f9, e12, j10, G21, and i20. Our verdict In each case is white wins. Show me a specific refutation of one of these lines or propose a different black move 10. I regard any other analysis as a waste of time. A case in point is my refutation of 9.q8. Did it serve to convince anyone to return their attention to 9.d10? Apparently not.
On 2020-01-13 at 13:05, MisterCat said:
David; relax; you are loved! If it means anything, I'M looking at your lines. For example, I absolutely LOVED your reply to Bob440, with:
Look at that absolutely NEAT box at D6, D10, H6, H10! Clever.
Meanwhile, I haven't exactly seen MY H10 given too much thought. What I see about Florian's ideas are that - instead of directly making an attack down the left side, if we PREFACE that with activity on the upper right, it might act as a safety net. I mean, whether Q8 (on the outs now, apparently) or P9 were played, if we succeed in reaching the top on the upper right, then GREAT. If TB blocks our direct attempt to win that way, the block won't extend across the whole board - we still have the edge with D6 on the upper left. And after a scuffle, perhaps what we have now in the center will HELP our chances of reaching that D6 peg.
I know; stupid play, and a waste of everyone's time. However, the idea is to use our strength on the upper left to allow a connection through the middle to the lower right, through peg R16. But what I fear is that the connection between M12 and R16 is not guaranteed. Go back a bit:
This is not the only example, but if the attempt to reach the top right were to fail, then white has a guaranteed connection here between M12 and R16, so can work from the top LEFT.
I think that Florian's idea then is not necessarily to guarantee a win on the top right, but by playing on the right side a little bit, we'll achieve some strength in the middle, to support a later attempt to win - either with YOUR D10 or even MY H10.
OK, I have to go back to work. FYI, a little about me - I am a local TAXI driver, but I work in my hometown. That means I can take an occasional break from driving to return to my home computer - like now, to post things like this.
I TRY to follow what is happening on my cell phone in my car, but sometimes I'm busy; also, I'm not logged into Commentator on my cell phone; even if I WERE, it's a pain dealing with the teeny screen on my phone. Furthermore, I'm not logged into my fake account MCx, so can't play actual game moves in my car.
So I'm running around, trying to do everything. I'll manage. Keep up the good work. I am ready and willing, if it is needed, to fork over the money (a donation to Richard, actually) and buy 10 more vacation days. We DO have time.
I know that some folks are getting 'antsy' - MOVE ALREADY. Relax. Have faith. I'm on top of all this, and will stay 'till the end. (meow)
On 2020-01-13 at 14:03, Florian Jamain said:
I stopped to analyse d10 cause I think it's a good move, and I believe everybody is okay to play it. So I'm trying to provide at least one other interesting move with the hours which left.
I have nothing against d10 actually, so I'm trying to see if we have something even better, it's like we found a probable checkmate in 12, just trying to see if there is an obvious one in 6 moves.
I could not find any strong response against d10, but my feeling is that it is still complex cause center is not fortified at the moment. If we could play a forced variation where we arrive to put 1 peg in the center, then playing d10 would be a sure win.
I studied q8, but we arrive to a position that Michal does not like, me I guess I could like it but less than d10 right now so it is stupid :D Then I tried a very easy variation with p9, actually I had no time to analyse it but the variation I studied was an easy win for us.
I clearly don't think I will find something better than d10 :D
But since we are waiting to play it, we can study the position, at the moment we gonna play it, the Bot gonna respond and we gonna start to study this further.
We wait this evening if Alan, Maciej, Michal or someone else wanna make some comments and we play tonight the move.
h10 - if it's connected to the topside then it can be a strong move. We should check for a trap here. See example - the trap: 9.h1010.f611.d1012.e813.i714.l415.g316.h217.c818.d2 (David this is similar to your previous line for 9.d10).
On 2020-01-13 at 16:05, MCx said:
OK, I created a new identity for Commentator called MCx, so I can post from my car. Let me get my reading glasses on ... that's better. I still can't move at LG, since I'd have to get my password to log in; could do that if I must.
David, I DID note and play through Maceij's trap line after my h14 - yes, it IS a serious and clever trap. But I didn't take it as a bust - just something to be careful to avoid.
The basis for that trap is that black block's White's attempt to reach the top; would not that sort of trap (or similar) apply to any, or at least most attempts at play on the upper left side? ( I have not checked specific lines.)
Michal wants to move; Alan's been silent; David, Maceij, Florian all push for D10. I'd have to scroll up to check for Bob, or any other opinions.
So I won't wait all day: 4pm NY time I move. That's 2 hours from now if any desperate words to stop me.
The reality David, is that I did not find something against d10 but my feeling is that since the center is not fortified, maybe somethinf exists and I'm not seeing right now, but there is also very good chances that just nothing exists for black!
But I'll try to study d10 more to try to find something for black right now!
So 24.m10 is necessary. Or something in the bottom half like 24.j20. I still prefer black but must give up my arrogant claim that I refuted Q8. This is messy too.
So, that being the case, I hope you’re still satisfied with 10.D9
What, MUST I decide??! I was all set to play D10, but there is a bit of clamour for P9. Somebody say something - I'm home right now, but can't stay very long. mc
On 2020-01-13 at 18:49, MisterCat said:
You know what? I actually preferred Florian's P9 for the reason I gave - I AGREED with him that some action in the center and upper right might serve to make D10 more effective later.
Had I not looked at the above posts - practically all current 'research' is being done on D10. That being the case, I'll play D10 right now. At the very least, hopefully all this hard work has not gone for naught.
...
Done. I've got to get back to work; can't hang around my home waiting for the reply. I can post from my car now, so WILL keep you posted, and meow.
Note: This site is starting to load quite some time for me. Maybe we can start a new forum http://twixt-commentator.duckdns.org/game/2140447 ? We could always go back here to see our previous lines.
Ok, then we have the first candidate 11.p17 proposed by Michal (but it does not really matter who proposed which move, does it?) then it would go probably that way 12.m2013.p21.
My thought about it are: It's kinda "all in". We give up the bottom threats just to be 100% connected. This can be risky. Some responses to check are: 14.o1114.p1114.p1415.o1316.o1217.p1518.k11 etc.
|1.d62.g43.m124.m175.r166.r207.o198.l89.d1010.h21 For: 11.i18 - it has a good connection to D10 -> G13 or F15 - it has a good connection to M12 -> L15(!) or J15 - it collaborate well with O19 -> M20 - even alone creates a threat to the bottom -> H20 - for now I can't find anything for black ;/ 12.j20 countered by 13.d20 - what else do we want from a single move?
I'm just letting you guys roll with the analysis for a bit, now. However, I still think that it is worthwhile to consider a connection from top left, to D6, to D10, across to M12, over to R16, down to O19, and then straight down to the bottom. I know - easier SAID than DONE. My H10 move might just come in handy, now!
At this moment, we have 107 hours; that's 4 days, 11 hours. See? A couple of nice quick moves every now and then, and things will be all right. Tomorrow I have the day off from work, so I can be more attentive.
Maceij - if you are having trouble scrolling to the bottom, I posted instructions from Alan at The LG Forum (Twixt vs. The World part II) - you can put a bookmark on your Chrome browser bar that automatically takes you to the bottom in seconds! Try that, if you have not already.
If you guys WANT me to create another fake game for all the new analysis, I can do that; it might get confusing though, since whenever I do that, SOME of the analysis for a particular move will be still over HERE, and SOME will be over there. But let me know, and I'll do it.
I'll say this game is now looking to be a real complicated battle - maybe one for the ages. I'm delighted. I hope everyone else is. (meow)
On 2020-01-14 at 03:57, Florian Jamain said:
Well, the obvious move is i18, and I think we can almost play it instantly.
Against m20 we had k19, against H21 it is i18.
We play on the left to reinforce this part and we threat to win directly this way but we keep a good connection to o19 so that all our pieces seem connected, remember that it was the argument of o19... We played this and not t20 precisely for this reason.
This is the move which makes sense for our game and justify our previous moves.
Of course a move like p17 is interesting but it's like... I think if we play this we have to be 100% it's a quick win or we gonna lose. And I don't think it's a quick win for us.
For these reasons I'm already okay for playing i18, this move is totally obvious to me, when d10 for instance was great but maybe "could" need a preparation.
On 2020-01-14 at 04:44, spd_iv said:
agree with Maciej and Florian, we can play i18 instantly.
Remaining time 132 hours. That's 5 days, 12 hours. mc.
On 2020-01-14 at 14:36, twixter said:
MC I have a favor to ask. Please would you create a new analysis page? I agree with Maciej’ that this page is taking a long time to load, in addition to all that scrolling. Thanks very much.
Here’s a new analysis page. Less time loading and scrolling! http://twixt-commentator.duckdns.org/game/2142725
On 2020-01-14 at 16:35, MisterCat said:
H13 kind of put the kabosh on my idea of connecting from D10 to M12. Oh well. Alas - we will never know if that would have worked out. meow.
OK, I did what was requested: I created a NEW, fake game stub for analysis. The links to the game and analysis page are posted at The Forum ('The World vs. Twixtbot part 3'). Let me try to do it here:
New commentary page: http://twixt-commentator.duckdns.org/game/2142726
as always, the actual game can be found here: http://www.littlegolem.net/jsp/game/game.jsp?gid=2140447
Now, I'm going to try to copy and paste all the recent comments about move 13 from here, over to the new commentary page. However, I'm only grabbing the most recent ones; if there are other lines of value above, you'll have to refer back here. (meow)
On 2020-01-14 at 17:11, twixter said:
Sorry use http://twixt-commentator.duckdns.org/game/2142726 instead
Twixtbot vs. The World.
The game is unrated, and the victory or loss will be recorded under my fake ID "MCx".
Awaiting commencement.
Now that you are HERE, you see how this goes - post comments below, and once the game starts, I will use the comments to decide on moves. I have not even decided on whether to go FIRST or SECOND, and it gives you that option when you make a challenge.
Comment?
I hope Alan H. can help!
If you let TB go first, it'll probably play on row 5, 6, 7, or 8, and that is unfamiliar territory right away for a human. If you go first and try to lead with a move that far from the edge, TB tends to make a tactical battle out of it, which is also unfamiliar territory for us humans. If you go first and pick something like d3 that seems even-handed for humans, TB seems to think that's way too weak and knows how to crush it.
Of these bad options, I personally think our best strategy is to go first, and choose a first move that TB would play, and keep playing it at its own game - throw its own kind of tactical opening moves back in its face. If you try to make it play the way a human plays, things seem to deteriorate quickly. We have a pretty good idea what first moves TB swaps, so If you want to effectively go second, you can just choose one that we know TB swaps.
Here are my TwixtBot first move stats for reference.
TB likes to lead with h8, but swaps it. It also likes j6, which it doesn't swap. It likes h6, which looks like it's probably in the no-swap zone, but we can't be sure until somebody tries it. Or, if you're really feeling adventurous, we haven't seen any TB game start on i7 at all.
I'd like to recommend a pre-first-move movie: "Miracle" ;-) Go Team!
I am purposefully trying to stay out of this, as much as I can. I hope some more comments chime in, and we can agree - at least on the first move!
I wish this game to begin by New Year's Eve, at the latest. I will start it earlier if there appears to be more of a consensus. (meow)
This is NOT the official game move, but at least the board won't look empty!
(meow)
Alan sounds much more confident, but I, for one, haven't the foggiest notion as to how to play TB "at its own game."
[ sigh ] none of that leads to a suggestion for a 1st move.
So I say pass. (mc)
wouldn't that mean that it evaluates all of its opening moves as roughly 50-50 to result in a win (whether or not its opponent swaps)?
so, are we overthinking the 1st move stragedy? (yes, I know that's not a word)
what about C8, this move looks quite strong, not as good as C12 but still offers many possibilities (good connection to the top by C4). I think the TEAM should make a first move which TB did not play very often.
I am not good at the beginning and I will try to visit this site from time to time to make some comments.
In my opinion moves like H8, J6, H6, I7 looks like to be easily swapped. If TB doesn't swap J6 maybe we should try this one. Disadventage of this move is that TB played it many times and knows how to play against it.
TwixtBot seems to believe that certain first moves are worth attacking directly. For example, it appears that TwixtBot thinks a good response to |1.f8 is 2.f4. Would you feel comfortable playing 2.f4? Would you know how to follow through with that? What if the best response to |1.c8 is actually something weird like that, and anything we're familiar with is something TwixtBot easily turns into an advantage for itself?
I'm not hopeful that basing decisions on what we think TwixtBot is familiar with is useful. TwixtBot has played thousands of times against itself starting on every legal spot on the board. It should know how to play against any first move, as comfortably as putting on an old shoe. Our own familiarity is a factor.
Yes, we know TB doesn't swap J6. If we start there, we'll get 3rd peg. I feel that's easier to analyze than 2nd peg. If spd_iv is ok with |1.j6, I'm comfortable with that.
However, as I said in the LG forums, I think we should let the highest ranking Twixt player in the room make final decisions. At the moment, that is spd_iv.
When faced with |1.j6 2.L13, TwixtBot responded with 3.p13, as seen here.
Games 2054062 and 2082323 also involve 1.j6 and TwixtBot, but are not as informative.
As the master of ceremonies here, I hope NEVER to be put in a position where I have to make the final, faulty choice of a move. I am really hoping the discussions will swirl around agreed moves. But if I MUST, I will read the discussion and try to go with a 'majority rule' - not strictly going with the highest rated player. Three 2200's will overrule one 2400, etc., Remember, practically everybody has a losing record against Twixtbot now, and I hope that we can prevail by working as a TEAM.
Bony Jordan HAS given approval for this experiment, and he is welcome to contribute his own, HUMAN analysis.
Tomorrow is New Year's Eve, everybody. Happy New Year, and the game WILL commence. (meow)
|1.c8 2.swap 3.e6 - this is first move my intuition says but 4.e4 seems to be too good,
|1.c8 2.swap 3.i9 - or
|1.c8 2.swap 3.o15 (example) |1.c8 2.swap 3.o15 4.n11 5.i11 6.j7 7.l8 8.m9 9.j9 10.l7 11.j5 12.k5 13.i7 14.j3 15.h4
from the other hand I will always swap |1.j6 - I do not remember any of my game starts with j6 but my intuitions tells me this move is very strong:
|1.j6 2.l13 3.h13 4.h10 5.l11 6.m11 (game nr 2075839) what would TB play after 7.f10?
I think both j6 and c8 looks fine for the first move and I agree to use one ot them.
I have no idea why TB swaps |1.b6 - seems much weaker than |1.d6 (which TB do not swap). I think that all the advantages |1.b6 has are included in |1.d6...
So I would recommend to play:
1.b6 2.swap 3.j9 (here 3.j9 can be changed to an adjacent field)
OR
1.d6 and we'll see what is the TB response
|1.b6 2.swap 3.j9 can be countered by 4.j9 so I would recommend to leave left top corner battle for later and play 3rd move in right bottom corner:
|1.b6 2.swap 3.q15 ...
Case: |1.d6 -- I would recommend this one (I'm curious what would be the TB response)
Case: |1.j6
|1.j6 2.l13 3.p13 suppose TB plays 4.p18 and then:
5.t18 6.s15 OR
5.t13 6.o20 OR
5.o16 6.m17 7.p17 8.q16 9.o19 10.p8 ...
Each time feels like black wins.
Case: |1.h6 -- similar to |1.j6
--------------------------------------------
So I vote for |1.d6 (eventually |1.b6)
Noting Alan's above comments on j6, it appears that if we do that, the Bot will move right in the middle; that seems scary to me. c8 IS a good move, but Maceij is just guessing, since it's never been played; b6 doesn't look like a good move to me. I'd rather play d6 personally. Unfortunately, spd_iv is not too keen on d6, but maybe has been persuaded.
I'm just trying to make sense and summarize the comments I'm reading above, and maybe not doing too good a job at it. I'm going to get the game underway NOW. That gives us 240 hours to move, plus vacation days, and I sure hope we don't blow it all on move #1!
We could of course resign and start again if we choose to.
NOW HOLD ON WHILE I FIX SOMETHING: I accidentally started the game using Mister Cat. This was a mistake. I now need to resign that game, RESTART another under my MCx id, and re-post messages and links over at the Forum. I'll do all this now. Bit of a pain in the neck, but MY MISTAKE. Hang tight.
(meow)
Today IS New Year's Eve, and I wish to get this party started. I can't even restart the game until this is decided, since you select who goes first when the game is created. I'll give you guys a couple of more hours to bat this idea around.
As it stands, if nothing new is added, I intend to go first as analyzed above.
Maybe you could wait until that point is clarified?
David, the idea is not to convince ME to 'pass'; the idea is to convince The Team. I'll have no choice if I MUST make a final decision, but I'm doing all I can to avoid being in that position. Right now, as it stands, it appears that the above discussion points towards The Team making the first move. Let us see if anything changes in the next few hours.
mc
I will try to hyperlink to the game here: http://www.littlegolem.net/jsp/game/game.jsp?gid=2140447
If that did not work, go to game #2140447, or just see the post at the Forum.
(meow)
I would not recommend to start with |1.b6 because there is a chance thatTB will not swap it.
can't we play a second game at the same time where TB makes first move?
|1.d6 - 2,5 (Maciej + 0,5 from me)
|1.j6 - 2 (Me + David)
Alan what is your suggestion?
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.h15 10.i10 11.n16
what is next:
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.i16 10.l9 11.m16 12.q11 or
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.i16 10.l9 11.g11 12.g9 13.i10 14.j8 ?
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.r19 10.o20 11.q21 12.u16 13.i16 14.l9 15.g11 16.g9 17.r14 then 18.q7 19.n8 how’s that?
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.r19 10.o20 11.q21 12.u16 13.i16 14.i18?
I think giving O20 to black is not worth. Instead of I16 I recommend to play G15 (like in my previous post): |1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.h15
My reasoning:
1) Let's play |1.d6 suppose bot answer is 2.d4.
2) Remove in our head the d6 peg.
3) The state of the game is like this |1.d4 2.swap (note that swap comes from the bot).
4a) We know that TB do not swap the |1.d4,
4b) plus we have a d6 peg for free.
Going back to |1.j6 case:
I have analyzed the |1.j6 2.l13 3.p13 4.p18 state by playing several games (me vs me) and all the time black wins. So
*) I am not able to see the potential of |1.j6 and how to work with it correctly.
OR
*) the game would be a tough one...
The j6 examples:
I agree to spd that |1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.r19 10.o20 11.q21 12.u16 13.i16 14.i18 belongs to black --> here white made a mistake of giving 10.o20 for free. Since white can avoid making this mistake by playing immediately 9.i16 we can focus on something else:
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.i16 10.l18 and what next?
my example of how it could be
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.i16 10.l18 11.g17 12.j14 13.g12 14.f12 15.n14 16.k16 17.r14 18.q16 19.h14 20.h11 21.i11 22.j10 23.k10 24.l9.. again black wins
I'm still not doing anything. Please keep up the work ('good'?) so as to agree on a move 1, before too long. I know that this game might easily be lost or won by move 10, so the early moves are real important, but if we use up too much time, we shall have to resort to making it back with a few quick moves (and 36 hour bonuses).
Hopefully move 1 tomorrow. Oh, spd_iv, I will respond to some of your comments from above back in The Forum thread.
I am connvinced that: |1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.h15 10.l18 doesn't look like an easy game for white :) maybe 5.t18 is not perfect but as Maciej suggested |1d.4 looks much better for me after his explanation. In worst case TB will swap it :)
Within a few hours from this post, I shall play a move, so if there are any last thoughts, now is the time.
mc
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.i16 10.l18 11.m17 12.m16 13.l19 14.k20.
Anyway, I am good with |1.d6, especially after thinking about Maciej's comments.
Maciej, you don’t have to repeat the entire line from the start within the same comment. For example 14.n19 15.r19
Let's begin with: |1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.t20 9.i16 10.l18
11.m17
--> 12.m16 13.l19 14.k20 15.o16 16.m21
11.k17 (interesting one)
--> 12.m16 13.g12 14.h14 15.h13 16.g16 17.g17 18.i12
-----> 15.h15 16.g16 17.i13 18.j19
-----> 15.j13 looks pretty nice for white so need to undo the M16 peg
--> 12.o10
--->13.g12 14.h14 15.j13 16.j10 17.l12 18.m9 19.t8 20.t10 21.s10 22.n12 23.m16 24.q14 25.q15 26.p12
--------->19.t9 20.r9 21.s7 22.s11 23.u11 24.r13 25.u16 26.r19
but do we know how to answer for 10.l9?
How about
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.o10 11.t8 12.q14 13.n12 14.m11 15.r12 16.o15
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.o10 11.t8 12.q14 13.n12 14.m11 15.r12 16.o15 17.m14 in your position it's actually good for white so i'm not convinced that black would go that way
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9 it looks like a lot of room on the left for white to get something going. Maybe 19.g15 20.g13 21.k11
So. Definitely d6 right? Heh
Here’s one more look at my g15 idea. |1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9 19.g15 20.j14 21.e14 22.g11 23.e10 24.f8 25.e9 26.g6 27.m9
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9 19.g15 20.j14 21.e14 22.g11 23.e10 24.f8 25.e9 26.g6 27.k8 28.h9 29.n12
Mr. Cat I hope you are willing to wait another day for me to convince the brothers Celuch. Thanks.
Suppose 1.d4 should be swapped, even though the bot does not, to counter Maciej’s argument. In that case it seems likely that d6 should also be swapped. But since TB has not swapped it, again this would mean we would gain an advantage.
Is there a chance TB will swap, despite its past behavior and Mr. Jordan’s disclosure that its rating peaked in June? I don’t really know. Which scenario seems worse, bot swaps j6 or bot swaps d6? Also what if the bot answers d6 with something else entirely?
I don’t really mind if we play d6. This j6 analysis has value to me in and of itself. But I would still appreciate another day to discuss it. Thank you.
I really was expecting to play 1. d6 now. Instead, I will do nothing until tomorrow. This gives ME time to look through all the above commentary, and what follows. As I said, I am trying to get a feel for what people want, and act accordingly. But this means I have to be AWARE of what is being said, whether I can follow it or not. I'll continue to do my best.
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9 19.g15 20.j14 21.e14 22.g10 23.k10 24.m11 25.i11 26.h15 27.i16 28.g13 29.e12 30.f12 31.f14 32.f11 or
22.g9 23.m10 24.o11 25.h11 26.h15 27.i16 28.i10 or
20.h13 21.e14 22.i15 23.g11 24.g18
black has many possibilities here. I still think we can play J6 but playing as Maciej and David discussed there is big space on the left of the board which can be dengerous for us. That is why I vote for D6.
How black can use D4 or D3? by playing sth like this: |1.d6 2.d4 3.l12 4.i4 or 4.h4. of course not as soon, so I think we could play the sth like: |1.d6 2.d4 3.l9 4.l11 5.q15
If TB will swap D6 i think we can play J9 or somwhere in the middle
Going back to |1.j6. A tiny error was there:
|1.j6 2.l13 3.p13 4.p18 5.t18 6.s15 7.t13 8.u16 9.n17 10.i19 11.l18 12.o15 13.o16 14.p8 15.r8 16.n13 17.q15 18.r9 19.n9 --> white wins
so black has to play 18.q10 19.g15 and then
i don't like the 20.j14 because of
--> 21.e14 22.h15 23.i16 24.g10 25.i10 or
--> 21.e14 22.h15 23.i16 24.g9 25.i10 26.g13 27.o11 28.n9 29.m10 or
--> 21.e14 22.h15 23.i16 24.g9 25.i10 26.g13 27.o11 28.m9 29.m10 30.l11 31.p7 32.o6 33.s11 34.t7 35.t9 36.q5 37.u5
but the spd's 20.h13 seems to be better
21.j12 22.k11 23.k10 24.l9 25.i14 26.j8 27.e8 28.f8
21.i10 22.m11 23.h12 24.i15
21.f12 22.i15 23.h11 24.h17 25.n12 26.m10 27.p7 28.s11 29.m9 30.l8 31.n11 32.n7 33.l10 34.j9
21.f12 22.i15 23.h11 24.h17 25.o11 26.n9 27.m10 28.l10 29.k11 30.j9
and another good black move would be |1.j6 2.l13 3.p13 4.o10 ! Which is a tough one.
Don't know if I'm the only one but... I don't think TB is playing strongly to TwixT. We just don't have enough time usually to study the positions, he tries to put the game in a tactical stuff and then he gonna win cause you gonna fail, but even if he is the best in tactics I don't consider him as a good player in strategy, he is playing many bad moves to me in the early game which shows that he does not really understand what is happening.
Maybe it is just the contrary and he is also the best there this is why I don't understand what he is doing :D
I don't believe it at the moment.
If he is not that strong we can easily flip a coin and choose one of the two discussed starting moves. No need to check 50 openings. For me it can be either J6 or D6.
So my suggestion is that we should flip a coin and let fate choose for us ;-)
After the 3 or 4 moves, if you are lot ahead it's over I imagine, this is the moment we need to do the difference.
But yeah, juste start the game with one of these moves :P
(meow)
www.littlegolem.net/jsp/game/game.jsp?gid=2140447
1.d6 2.g4
www.littlegolem.net/jsp/game/game.jsp?gid=2140447
(meow)
Imagine this scenario:
--> |1.d7 2.swap.
And compare it to the current state of the game:
--> |1.d6 2.g4
-----------------------------------------------------------------
Ok, let's divide the board into 5 areas:
--> Center (my vote: + "good")
--> Top Left (my vote: - "bad)
--> Top Right (my vote: 0 "neutral")
--> Bottom Left (my vote: 0 "neutral")
--> Bottom Right (my vote: + "good")
What about we first decide which area we would play and then decide the very move?
I vote for Center (3.m12) or Bottom Right (as spd pointed out 3.q15).
-----------------------------------------------------------------
Center examples: 3.m12 (David is m12 worth a try as a center move?, I'm asking you as David a very experienced guy in this field)
4.m16 5.r16 6.r20 7.o19 8.m20 9.j15 10.h20 11.f15 12.f19 13.p21
4.m16 5.r16 6.r20 7.o19 8.m20 9.j15 10.n18 11.j19 12.m8 13.d10 (look at the d10 - seems like has a connection to top edge)
4.m8 5.p9 6.p7 7.t9 8.t4 9.d10
Each one seems good for white.
-----------------------------------------------------------------
Bottom right examples: 3.q15
4.o16 5.n12 6.q17 7.s16 8.r19 9.j16
--> 6.p21 7.k17
4.p13 5.m13 6.m9 7.k9
4.q20 5.u19 6.t17 7.s16 8.s19 9.u15 10.v18 11.o19
--> 8.r18 9.u17
4.o13 5.l13 6.i19 7.r8 8.q17 9.o16 10.o18 11.g16 12.g18 13.k18
--> 6.h19 7.r8 8.q17 9.l17 10.l10 11.k11 12.k8 13.m10
Each one seems good for white.
-----------------------------------------------------------------
Some words about 3.j6. There are quite a few reasons why I dislike 3.j6.
1) First of all I do not like to play anywhere in Top-Left area for now.
2a) J6 do not cover/surrounds/blocks the G4 peg, so G4 still has some value.
2b) Furthermore G4 is probably slightly better than D6 when it comes to duel between those two (just from Twixt Bot First move stats doc).
2c) So scenario 1.j6 is probably better (for us) than scenario 1.d6 2.g4 3.j6.
2d) We rejected the 1.j6.
3a) Quick examples:
3b) 1.d6 2.g4 3.j6 4.j8 5.f10 6.e8 7.g8 8.h9 9.d11 10.g7 11.c8 12.d2 13.f5 14.l4 15.h4 16.b11 17.b10 18.g16
3c) 5.n9 6.n11
3d) 5.f11 6.e8 7.e9 8.h14 9.f15 10.h18 11.f19 12.g16 13.e17 14.f6 15.i8 16.g12 17.e13 18.f10 19.c8 20.d2
-----------------------------------------------------------------
Ok, I've put some work here. Now I'm waiting for your response, corrections, other suggestions, examples etc.
The first response that comes to my mind is 3.h6.
That’s how TB does these edge attacks. They start out looking ridiculous, then they continue to grow until they become overwhelming.
I don’t like 3.h6 because something like black j5 or k4 could still be a threat, and a central move grabs more influence, to borrow Go terminology.
This is to me what Alan would describe as a tactical approach; we are immediately hitting the Bot back with a reaction to its move. Bot seemed to like being 3 spots right and 2 spots up, so I thought the same: 3 spots right, 2 spaces down.
White wants to reach the top, so if black blocks on the left with 4. e3, we reach the top with 5. k4 . On the other hand, a block on the right allows us to reach on the left.
Of course, this ignores the whole rest of the board, but I don't know how to think that far ahead. I'll trust that your examples effectively shoot down the move.
In his instructive comments in the Twixt Wiki (still available), Alan warns AGAINST local battles, and suggests engaging in global battles. I would say your brother's suggestion of 3. q15 would be a more global approach.
So Alan, if you follow YOUR OWN sage wisdom, your suggestion of the tactical 3. h6 would be a local battle, and not good. (despite your win, from last February)
Alas, I don't comprehend how anybody could plan far enough ahead, with any degree of certainty, to evaluate a move like 3. q15 being any more correct than a 3. q14 , 3. q16 , 3. r15 3. p 15, or anything else, for that matter. Perhaps somebody will enlighten me.
1.d6 2.g4
Now a comment about practicality: there are 22 x 22 - 2 choices available for our next move. That works out to 482 possible moves. Considering what happened in the above analysis of move #1, I fear that if EVERYONE keeps suggesting different moves with accompanying projected lines, this game will be lost on TIME FORFEIT before we can reach a decision! I think that we need to have more of a meeting of minds - perhaps if folks can agree on the TYPES of moves they are searching for, then those 482 choices might be narrowed down significantly.
As far as specific tactical lines, I'm COUNTING on the expertise of the team here to catch outright blunders, oversights, and serious strategical mistakes.
That was a pep talk. GO TEAM!
*) J6 alone is a weak move. Why? It's considered a 50-50 when it comes to swap. But we are not in swap phase. So it is better to play a strong move such as 3.m12.
**) Furthermore J6 do not have a big impact on G4 so I don't think it's a tactical move i.e it would be great if J6 would block the G4 but it doesn't. Why? Black player can easily divide D6 from J6 as he needs (for example: 3.j6 4.g8 5.h5 6.h6 - this is the example of what could be done in the future, I don't really think seriously of 5.h5), which is not good for White.
The 3.h6 is a better try because it has bigger impact on G4.
If that's not enough then maybe someone more eloquent than me can explain it?
We have an advantage right now, so probably there are a lot of moves which are winning for us. It can be 3.h6 3.m12 3.q15 (and maybe even 3.j6) etc.
|1.d6 2.g4 3.h6 4.h7 and things are going messy.
Do we still fight for the case?
5.g8 6.i5 -> don't go that way!
5.f7 6.i5 -> don't go that way!
5.j7 or 5.i8 -> and there is a black F6 -> J5 threat (maybe this is what David had in mind). It can be really dangerous later on.
Personally I do not like this local battle.
G4-K4 combo do not connect left, so we can block him 4.m16 5.r16 6.m10 7.l10 8.k4 9.f5
But if we are so afraid of the G4-K4 we can go D10 first 7.d10. And he can't block us from top edge
|1.d6 2.g4 3.m12 4.m16 5.r16 6.m10 7.d10 8.c4 9.f5 10.l4 11.h4 -> now he gain L4 for free but we gained D10. The problem for black is that L10 is alone but D10 is already connected.
Going back to the reasoning why I do not like the top left local battle. Leaving the top left corner untouched gives us opportunity to respond to it as we like (D10 or F5 or H5 depending on the board state).
Note - I just played through ALL your above examples of why 3. j6 is NOT good, and they were quite instructive. THANK YOU! I was too sleepy to do that last night!
Continuing to explain my thought process (why?),
after 4. g8, my instinct would be 5. n6. Now, we can connect to the top from any peg, and black must try to block from the other side, such as 6. n10. But then white trips black up with 7. j10 ! I looks to me now that white wins the top, so black must take the game to the other side, like 8. i16 . Then the fight starts anew.
Hey - this move-ahead stuff is starting to be FUN! Maybe I'll get better at it! OK, so you see - that's the way I think, totally flawed may it be.
So to summarize, besides my 3. j6, trashed above,
we have 3. m12 from David ('center' play, which Maceij approves),
3. h6 from Alan ('top right' play; Maceij does not like),
3. q15 from spd_iv ('bottom right' play, approved by Maceij),
and no other suggestions on the table for move 3, for now, unless I missed something above. Great - that's much less than 242 moves! I wouldn't mind a few more suggestions thrown around at this point - maybe some new folks will chime in, but then hopefully we will come together on a preferred move.
189 hours remain, at the time of this post.
3.o15 4.o13 and 3.q15 4.q13 it feels like the black countermove has bigger impact in first scenario. And 4.q13 do not seems to be strong for black ((*) is too close to the right)
Another thing is that I've often played it successfully (vs humans). At least I feel that way. It can be easily checked by any of you.
3.q15 4.p19 5.t18 -> where 4.p19 kind of moves don't work and Q15 is 100% connected to bottom.
But for example 3.r15 (is too close to the left) and this scenario 4.q19 5.u18 do not work here. White need to play 5.u19 which is not that big threat.
Perhaps 3.q14, 3.q16 could also be fine but 3.q15 seems like a good average of those two :-)
3.h6 4.h7 5.i8 looks like a White advantage to me, but I don't think 4.h7 looks like something TB would play. A more Twixtbot-like line might be 4.j7 5.h11 6.e5 7.i4 8.h8.
M.C., regarding the words I wrote on the Wiki many years ago, in the pre-TwixtBot days: I think it's still good advice for beginners, but it is possible that effectively executing opening tactical moves is an advanced technique.
|1.d6 2.g4 3.m12 4.m16 5.r16 6.m10 7.d10 I don’t see anything for black. So I agree with you about m12.
4.h7 is more crazy and scary for me than 4.j7 5.h11 6.e5 7.f7 -> seems to be good for white and not that crazy.
If we know that bot would not respond 4.h7 I can vote for 3.h6
Those 36 hour increments are pretty handy, but if we get just one of them today, we should be ok for analysis of move 5. At this posting, 187 hours to go.
7.h11 8.f6 9.g9 10.h14 thanks to J5 thread black has an opportunity to get a free top-right corner move which makes the game complicated from here.
9.d11 10.i9 11.c4 12.d9 13.e8 14.h14 15.d15 16.f13 17.e13 18.e11 19.c9 20.e17 -> something went wrong for white.
17.f12 18.e15 19.j12 20.k14 -> something went wrong for white.
15.j12 16.l16 -> from here I checked some left battles and couldn't stop black from the left (neither 17.d16 nor 17.j15)
7.d11 8.e13 9.d16 10.f17 -> crazy, isn't it?
Save me from making this decision, Michael. Convince David that q15 is to be preferred, and I'll do that. I'll check back. (meow)
I would prefere M12, because it seems more flexible to me (just a feeling).
1.d6 2.g4
Welcome to DrJochum.
Well then, I'm still stuck. I KNOW that 3. q15 was spd_iv's choice, so I'd go with THAT, but the NEW guy likes 3. m12.
Oh me, oh my, the sadness of the world; how shall I cope?? Sniff! What to do?? Help. meow.
Move List
1.d6 2.g4 3.m12
1.d6 2.g4 3.m12 4.m17
(meow)
6.k16 7.q12
-->8.m21 9.i17 and check for bottom connection:
---->10.h21 11.f20 12.g20 13.j19 14.k22 15.g22
------>12.e22 13.c21 14.d21 15.j19 16.k22 17.g22 18.g23 19.d23 20.e19 21.e21
---->10.i21 11.j19 12.k20 13.f21 14.h19 15.f18 16.h15 17.h14
------>16.e18 17.d17 18.g19 19.e19 20.f17 21.g18
-->Ok, I assume that 9.i17 is connected to bottom...Correct me if I'm wrong
-->10.j14 11.i13 12.i12 13.h11 14.h10 15.g9 16.e5 17.f7 18.c6 19.i5
---->14.g8 15.j10
-->And it seems that J15 is connected to top... So probably 6.k16 is a mistake or there exist something strange like 7.q12 8.l21
6.i17 7.r17 (here i17 can be blocked by G17 or E17)
--> if black plays on the top half, then we respond D10
--> 8.q15 9.l16
--> 8.m21 9.l16 10.n19 11.p18
----> 10.r21 11.r22
6.i19 7.k18 8.n19 9.i17 10.k20 11.d19
-->8.k16 9.q12 10.n19 11.i17 12.k20 13.d19
6.m10 7.d10
6.e5 7.q12
any ideas?
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i17 7.r17 8.m21 9.l16 10.r21 11.r22 12.p16
Threatening black s19
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.k16 9.q12 10.n19 11.i17 12.k20 13.d19 14.g9 threatening black f20 or j5.
For that matter how about 6.j5
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.i17 10.k20 11.d19 12.g10 13.h14 14.m9
I still think 6.j5 might present difficulties.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.p19 8.m20 9.h17 10.j14 11.r18 12.p21 13.n20
Meanwhile, I'll try to construct a line where white WINS with j15, and I'll ALSO see if I can come up with another suggestion for move 5. mc.
1.d6 2.g4 3.m12 4.m17 5.j15 6.j17 7.g16 8.g20 9.d19 10.e19 11.o17 12.n15 13.p15 14.o13 15.q13 16.p11 17.r11 18.q7 19.p10 20.o8 21.n9 22.m9 23.l10 24.k10 25.c4 26.f10 and black wins.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.j17 one reply is 7.q16 and if 8.p14 9.o17 threatening white n15, but if 10.n15 11.k17 12.l18 13.m18
|1.d6 2.g4 3.m12 4.m17 5.j15 6.j17 7.q16 8.p14 9.o17 10.o18 would be better for black but then 11.s15 12.q19 13.u19 14.t17 15.t16
maybe then 7.r17 8.m21 9.l16 10.r21 11.k18 ? If there was a battle on bottom then it might look like:
12.h21 13.f20 14.e22 15.c21 16.g21 17.j21 18.k20 19.i19 20.j22 21.d19 22.h19 23.g16 or 22.f19 23.e17 so I think here 11.k18 is connected to bottom.
10.l15 would be another option 11.f15 12.f20 13.k18 14.k22 - black wins, so we can't go this way 11.f15, so we have to play 11.k18, which is risky. For example:13.i12 14.l4 15.f5 16.g10 17.o7 18.l10 19.k11 20.n5 21.p5 22.n9 23.l9 24.m7 25.k7 26.f6 27.i6 28.i3 29.e8 30.g8...
Another approach can be 5.j15 6.i17 7.l16, and then 8.l15 9.f15 (10.g8 11.q12 12.m21 13.h16 14.k13 15.j11 or ) or 8.l19 9.r17 (10.o16 11.q15 12.p13 13.r12 14.q10 15.o13) or 10.p9 11.n17 or 10.o16 11.q15 12.p9 13.h16 14.g16 15.o11
5.j15 6.i19:
David, what about 7.k18 8.k16 9.q12 10.n19 11.i17 12.k20 13.d19 14.g9 15.i13 16.i10 17.h11 ?
5.r16:
What about 6.r20 7.p19 8.o21 9.o22 10.n19 11.q21 12.p16 ?
@MisterCat: |1.d6 2.g4 3.m12 4.m17 5.j15 6.j17 I think that in your example, the mistake is 7.g16. Why would we give away left bottom corner that simply and gaining almost nothing at all? 7.f17 7.e17would be stronger in this part of the board. But of course here 6.j17 we could also play around right side like David said.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.i17 10.k20 11.d19 12.g10 threatens black f20, 13.h14 14.m9 15.n10 16.n7 17.l9 18.k4 19.f5 20.h6
Or 9.l20 10.k16 11.q12 12.l14 protects against white r18 13.i17 14.j13 15.e13 16.f10 17.f11 18.e8 19.g9 20.i5 threatens to reach j13 or scoot across the top
Perhaps the crucial line is 11.i17 12.h10 13.q12 14.n7 15.q7 16.j14 17.r18 18.q14 19.q16 20.r12 21.p14 22.p6 23.s6 24.q8 25.r9 26.p11
|1.d6 2.g4 3.m12 4.m17 5.j15, 6.i19 7.d19 8.g9 9.q12
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.p19 8.o21 9.h17 10.k14 11.r18
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.l20 10.k16 11.i17 12.h10 13.j11?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.p14?
I like both moves 5.j15 and 5.r16 I tried to find any other moves but I am not sure about them(5.i12? - hard to say what would be the black response)
I would pick 5.j15 - Maciej made great job with his analysis (I cannot find any mistakes)
Michal,
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.l20 10.k16 11.i17 12.h10 13.j11 14.l7 your move.
If 5.r16 6.p14 7.o13 8.q16 9.p17 10.s15 11.l15
5.h14 will take me a while.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.l20 10.k16 11.i17 12.h10 13.j11 14.l7 15.j7
but as Maciej wrote before: 7.d19 seems to be much better for white.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.l20 10.k16 11.i17 12.h10 13.j11 14.l6 instead.
7.d19 8.f20 9.c21 10.e17 to me looks worse for white not better.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.d19 8.f20 9.q12 10.q17 11.i17 12.j21 13.c21 14.e17 15.h19 16.g20 17.f18 or sth like this:
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.d19 8.f20 9.q12 10.n19 11.r18
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.k18 8.n19 9.l20 10.k16 11.i17 12.h10 13.j11 14.l6 15.o7
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.d19 8.f20 9.q12 10.o7 11.d10 12.c4 13.f5 14.f6 15.e3 16.e8 17.c8 18.f11
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.d19 8.f20 9.q12 10.o7 11.d10 12.c4 13.u12 If white has bottom then the game is over.
Alan you may be right but thanks to Maciej and David we can predict which moves TB can play and how to react on them. Now we are analysing only 2 moves so if we agree than one of these moves are good enough we should just play it.
|1.d6 2.g4 3.m12 4.m17 5.j15 6.i19 7.d19 8.f20 9.q12 10.o7 11.d10 12.c4 13.f15 14.j5 15.q7 16.r12
I think black is doing very well here. Even if you find a j15 line that stumps me, there is no guarantee it will stump TB. IMO j15 will be more difficult and risky than 5.r16. Maybe we haven’t been talking about it because there is nothing much to talk about? That is exactly the kind of move I want to make.
At any rate, we have lots of time for this move. MC sounds willing to use vacation days on this move.
So let’s talk about r16 shall we? Show me what is wrong with it.
black can play sth like: |1.d6 2.g4 3.m12 4.m17 5.r16 6.l4 or
6.p10 7.i12 8.l4 9.c4 or
6.o13
5.j15 may be good but I agree with Alan, it's possible that it can go out of control - especially the top side battle. So I give up on that for the Team sake.
5.r16 - really nice try David. It's much easier to control the game state from this point (at least for us humans). There are too many variations on bottom right corner and each one gives something different to black. So probably TB will attack on different side of the board. Like 6.l4 6.k4 6.o11 6.m9. Because it's hard to predict the very one I think we should focus on local variations to be prepared of the threats arising from this.
Let's focus on some black local possible responses (I focus only on bottom right corner here to see the outcome of possible variations)
6.q14 (Q14 not necessarily as 6th move)
--> 7.l15 or 6.r20 7.o19 8.q14 9.l15 10.l19 11.i18 also good but what if black plays 8.j20 then what do we choose 9.p17 9.m18 9.h19? We should remember the J20 threat.
--> 7.s13 8.r11 9.o13 10.p16 11.p17 12.o18 13.q12 14.q19 15.u19 16.s20 17.v21 18.u16 19.t15 20.t18 21.v16 22.w17 - seems good for white, black can connect to the right edge at this point.
6.o13 7.n15 or 6.r20 7.o19 8.o13 9.n15 10.n19 -> ouch, 6.r20 is a real threat (7.o19 8.n19 or 7.o19 8.j20) hmm, maybe 7.p19 then as David mentioned? But I do not like it also...
--> Ok I stop there... I am too tired.
5.h14 seems fine too. @Spd, how would you respond to 6.h17 or 6.h19 I don't really know what else can black do except 6.k4
But 190 hours plus 30 days is still not infinite, so let us continue to be mindful of time. I love those 36 hour increments - these well come in REAL handy.
Maceij - thanks for the lesson above, which I shall study. Alas, I went wrong with THE FIRST MOVE in my hypothetical line; it may be a move that players like MYSELF would play, but you explained why it is flawed; it is NOT a move that the strongest players would play. But that, I think, illustrates the limitation of long, hypothetical lines - like you guys are using regularly; just ONE deviation from the line, and you've got a whole different game!
I'll do the best I can to keep up with the discussion here - I DO have a day job, you know - and if The Team decides on a move, I'll play it as soon as possible. mc
7.s13 8.r11 9.o13 10.p16 11.p17 12.o18 13.q12 14.q19 15.u19 16.s20 17.v21 18.u16 19.t15 20.t18 21.v16 22.w17 (my apologies of another bug).
But that doesn't mean that TB will play that. It just shows that there is a potential for black to connect right edge.
Move List
1.d6 2.g4 3.m12 4.m17
What I find a bit sad is that Maceij would SEEM to demonstrate that, with ONE LITTLE peg placed right of center (namely black's m17), is is NOT POSSIBLE (?) to prevent black from reaching the right? Meaning - if black were at l17 instead, we COULD stop him, but at m17 it can't be done?? Philosophically, that doesn't sound too swift for the overall 'truth' to the game of Twixt! What I mean is that games will be over - won or lost - theoretically after only 5 moves!
OK then - if Twixtbot felt to block US by placing a peg directly in front of us leaving 4 empty holes (the block from m12 to m17), might we not do the same to it? Block black's path to the left by placing a peg left of m17 with 4 empty holes, namely 5. h17 ? The 'truth' is that h17 can not be blocked from reaching the bottom, and we can aim to connect to the top either by heading for our m12 peg, or straight up to our d6 peg.
There - just some musings, since my efforts to generate those well-played hypothetical lines you guys use seem to fail. meow.
David what about 5.r16 6.k4 7.i12?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 with the idea of being a negative ladder catcher for black q14? 9.f5 10.m21 for example.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17
White can also force another possibility: |1.d6 2.g4 3.m12 4.m17 5.r16 6.q14 7.s13 8.r11 9.o13 10.p16 11.q12 12.r20 13.p17 14.n19 15.p20 16.p19 17.n16 18.o21 but I think the first scenario is better.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.o13 this one is also a tough one |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.o13 9.n15 10.n19 ... ? And for example: 11.m20 12.l14 13.n14 14.o17
Michal you meant that |1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 7.r16 8.q14 9.s13 10.r11 11.o13 12.p16 13.p17 14.o18 15.q12 16.q19 17.j17 18.j14 19.h18 20.i12 ?
|1.d6 2.g4 3.m12 4.m17 5.h14
Michal, what about 6.h19 ?
7.r16 is no longer good here
8.q14 9.s13 10.r11 11.o13 12.p16 13.p17 14.o18 15.q12 16.q19 17.u19 18.u16
but there is
7.p16
8.p18 9.k18 10.j18 11.m19 or
10.j14 11.n15 12.m21 13.p20 14.q20 15.o18 16.o21 17.j20 or
10.k16 11.p20 12.o16 13.g16 14.i21 15.i17 16.n21 17.l20 18.l22 19.n19
|1.d6 2.g4 3.m12 4.m17 5.h14 6.h19 7.p16 seems not bad, what you guys think?
|1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 probably is better for black then?
Do we still consider 5.h14?
5.h14 is simillar to 5.j15 and can provide to the mess.
I think we should try and play 5.r16. 5.h14 was just an exapmle and I did not anylise it much, but if team decide to play 5.h14 i am fine with that also :)
|1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 7.r16 8.q14 9.s13 10.r11 11.o13 I don’t think the bot will stay in the bottom half for long. 12.o7 and if 13.t15 14.j5 15.f5 16.f10, or 13.f5 14.h10 15.m8 16.p16 17.p17 18.o18 19.q12 20.q19 21.j17 22.l15
Also 5.r16 6.r20 7.o19 8.o13 9.l15 looks better for white than n15.
Michal maybe 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17 12.h10 13.f11 14.f14 15.g13 16.q21
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17 12.h10 13.f11 14.f14 15.g13 16.q21 17.s20 18.t22 19.s21 20.r21 21.q20 22.p19 23.u22
j15
h17
r16
Can I suggest 5.q17 or Is it naive?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17 12.h10 13.f11 14.f14 15.g13 16.q21 17.s20 18.t22 19.s21 20.r21 21.q20 22.p19 23.u22 24.r18 25.s18 26.u21 27.v20 28.t19 29.u19 30.w21 no it doesn’t?
Black might respond with 6.p20 7.r19 8.q22 9.s21 10.p6 11.n14 12.k4 13.f5 14.f10 or if 11.f5 12.p14 13.p15 14.o16 15.q13 16.q12 17.r11 18.p10 and either way, I believe black wins.
1.d6 2.g4 3.m12 4.m17 5.q17 6.p20 7.r19 8.q22 9.s21 10.p6 11.f5 12.p14 13.p15 14.o16 15.q13 16.q12 17.r11 18.r10 19.s9 20.t6 21.q8 22.o8 looks like black wins.
6.p14 7.q12 8.m21 - hmm we starting to loose game control, or
6.p14 7.p15 8.n15 9.q13 10.q12 11.r11 12.r10 13.s9 14.m21 ?
--> 15.r7 16.r21
---- > 17.t20 18.u22 or 17.r20 18.q19
--> 15.q21 16.r5
---- > 17.q8 18.p5 19.o7 20.m6
---- > 17.p6 18.s8 19.t7 20.t4 21.r8 22.m5 23.o4 24.q7
---- > 17.q7 18.o6 19.s6 20.t6 21.u7 22.v7 23.n6 24.m5 25.p7 26.p6
Both cases seems not good for white and I don't know if there is any better use of 5.q17 than 6.p14 7.p15.
At least that's why I don't like the 5.q17 6.p14.
The reason behind 5.r16 is that it is connected to the bottom side and has nice connection to M12. For example
6.p14 7.o13
6.q14 7.l15 8.k18 9.p15
@David,
|1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 7.r16 8.q14 9.s13 10.r11 11.o13 12.o7 not 13.f5 but 13.d10 and then for example 14.l4 15.f5 16.p16 17.q12 18.r20 19.k17. That can also be played like this 5.r16 6.q14 7.s13 8.r11 9.o13 10.o7 11.d10 12.l4 13.f5 14.p16 15.q12 16.r20 ... In the second example white have more space in bottom left corner.
I do not see a special difference here for having H14 and there is a room for a mistake (there is a chance that TB will make better use of h17 than we of H14).
So I vote for 5.r16
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17 12.h10 13.f11 14.f14 15.g13 16.l15 17.o13 18.q19
Easy question please - why do you guys keep playing f5 in the upper left in response to k4, instead of playing e4? I mean, with f5, black reaches the left side via, |1.d6 2.g4 3.m12 4.m17 5. r16 6. f5 7. e3 and white is blocked. Is the idea to connect with the center pegs? But if so, white is STILL blocked! On |1.d6 2.g4 3.m12 4.m17 5. r16 6. e4 , white can at least play for the upper left corner.
My head is spinning, trying to wade through the advanced lines being posted. Unfortunately, it appears to me that my earlier remark (about black having the advantage on the right side) is being shown to be true. I still think that's a shame, and wish there were a way for white to deal with that m17 peg.
But that being the case, I guess you guys are broadening the scope of your analysis, and playing lines that go ALL OVER THE BOARD! Great, I guess. I'll keep following, and if it appears that a consensus is being reached, I'll move asap. meow.
about
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5
10. e3 and white is blocked.
about
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.e4
white can still play for the upper left corner. That's what I meant. mc.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.n20
mc
19.f5 20.i10 21.k7 22.k9 23.l9 24.p16 25.q12 26.r20 27.k17 28.l15 29.j15 30.k13 or
19.j5 20.i3 21.f5 22.i10 23.k11 24.j8 25.k7 26.f6
|1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 7.r16 8.q14 9.s13 10.r11 11.o13 12.o7 13.d10 14.e12 15.f11 16.g11 17.j9 18.k11
|1.d6 2.g4 3.m12 4.m17 5.r16 6.k4 7.i12 8.q6 9.f5 10.m21 11.p17 12.h10 13.f11 14.f14 15.g13 16.l15 17.o13 18.q19 19.s19 I think this should work
I posted a useful tool back at The Forum at Little Golem, which I asked Alan to make for us. You may find it useful, so check out the post.
mc
|1.d6 2.g4 3.m12 4.m17 5.r16 I don’t see a good line for the bot. Thanks for your patience. Let’s make the move.
Ok, we rejected a territorial move like 5.j15 because of its complexity. It's difficult to determine the winner of 5.h14 6.h17 7.r16 vs 5.r16 battle. So (against bot) I prefer simplicity like 5.r16. No one has suggested any left blocking move like 5.h16 5.h17 5.j17 etc.
So I assume that we do agree to play a right-side blocking move like 5.r16 or 5.q17, right?
One more thing about 5.q17.
1) black straightforward attack can result in this (last time I made a mistake to not show this case)
|1.d6 2.g4 3.m12 4.m17 5.q17 6.p14 7.p15 8.n15 9.q13 10.q12 11.r11 12.r10 13.s9 14.m21 15.q21 16.r5 17.q10 18.l4
that case alone is pretty complicated. If that case is winnable for black I feel that TB will go that way. But even if a straightforward attack won't work, black can play a move on the top right edge that threats the connection to the right side like 6.p6 (and then for example 7.i12 8.l4 9.f5 10.h10 11.m8 12.f6). So black can gain an extra move on top-right corner with a P14 threat. Furthermore M21 peg is also playable for black 6.p6 7.i12 8.m21 9.o18 10.n15 11.q12 12.q14 13.p15 14.o13 15.r14 16.p11 or 6.p6 7.i12 8.m21 9.s16 10.r21.
5.q17 is not that simple because it leaves a hole between M12 and Q17 (itself) which gives a lot of possibilities for black - straightforward attack with 6.p14 or leaving P14, P15 as a threat. And its resistance to M21 is almost the same as 5..r16 has ---> 5.q17 6.m21 7.o18 vs 5.r16 6.m21 7.p17. I'm not 100% sure that those cases are good or bad for white. I just feel that r16 is a better move.
5.r16 does more or less the same as 5.q17 and do not have a hole between M12 peg and itself. That reason alone is sufficient for me to play 5.r16 even though 5.r16 has its own weakness points like for example 6.j20.
I could have some other suggestions for 5th move but I feel like it has no sense to share my thought right now because we would never make a decision :-)
So I vote for 5.r16 and to make a decision!
I see you guys getting real close on a move now. Instead of making it, I'll go to my therapist appointment. I'll check back in about 2 hours, and if the status remains the same, then I'll make the 5.r16 move as instructed.
mc
if we play 5.h17 then probably 6.h10 7.q12 and 8.m21. I don't like this position so if the Team has a good response for 6.h10 (or any similar move) then we can wait one or 2 days to make more analysis, but I think we should go for 5.r16 as we know much more about it and we lose less time :)
Can imagine "something like" :
|1.d6 2.g4 3.m12 4.m17 5.r16 6.p10 7.i12 8.m21 9.t15
Or some variations of this and it seems okay for white.
At this point I think the win is still in our hands, let's see what the Bot gonna find.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.p10 7.i12 8.m21 9.t15 10.p14 11.g17 12.g20 13.j20 14.i21 15.l19 16.k22 17.p17 18.n19 19.o13 20.s21 21.s20
I will play r16 now, by popular demand. If we lose this game, then next time we'll try something else (if there IS a next time). I was out for 4 hours, but I will move right now .......
There are 183 hours remaining. Expect a reply from Twixtbot within a few minutes! With luck, the pages and pages of analysis above will be proved worthwhile, if it makes one of the predicted moves. Then, much of our work will be done already for the next move!
Move List
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
Let's look above and see how much analysis is devoted to that move. Back to work, Team!
(meow)
twixter said:
Maybe it would be better for white to play on the right, then black may be less likely to stir up trouble along the top.
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.p19 8.m20 9.h17 10.j14 11.r18 12.p21 13.n20
Also 5.r16 6.r20 7.o19 8.o13 9.l15 looks better for white than n15.
Maceij:
5.r16:
What about 6.r20 7.p19 8.o21 9.o22 10.n19 11.q21 12.p16 ?
or 6.r20 7.o19 8.q14 9.l15 10.l19 11.i18 also good
but what if black plays 8.j20 then what do we choose 9.p17 9.m18 9.h19?
or 6.r20 7.o19 8.o13 9.n15 10.n19 -> ouch, 6.r20 is a real threat (7.o19 8.n19 or 7.o19 8.j20) hmm, maybe 7.p19 then as David mentioned? But I do not like it
this one is also a tough one |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.o13 9.n15 10.n19 ... ? And for example: 11.m20 12.l14 13.n14 14.o17
spd_iv said:
1.d6 2.g4 3.m12 4.m17 5.h14 6.h17 7.r16 8.r20 9.o19 10.q14 11.s13 12.r11 13.o13 14.p16 15.p17
Alan Hensel said:
1.d6 2.g4 3.m12 4.m17 - If 5.r16 6.r20, why wouldn't our local response be 7.o19?
Maybe I'm not a great Twixt player, but I'm pretty good with cut and paste (LOL)! Hopefully this helps give a head start. (mc)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
7. o19 and 7.p19
I prefer p19, for example 8.p14 9.o13 10.q16 11.p17 12.s15 13.l15
Or 8.o21 9.h17 10.i15 11.r18
seems to me like 7.o19 pretty well sews up the bottom (for the moment) -- now we go play up top
7.p17 8.n19 or 7.q18 8.p20 don't seem to be an option.
I do remember that I didn't like both 7.o19 and 7.p19. But maybe 7.o19 has some potential... David the problem with 7.p19 for me is that we have to involve other part of the board for it to work. For me this is risky. Examples: 7.p19 8.o21 9.h17 10.h19 11.n20 12.l19 (13.r18 14.p14 15.o13 16.q16 17.q14 18.t19 or 13.l21 14.p16 15.p17 16.o18 17.t15 18.q19 19.t20 20.u19).
Then my suggestions are 7.t20 7.v20 7.u19 and 7.o19
There is a counterplay to7.t20 (a way to use R20 against T20), isn't it? I do not remember right now so leave it as a suggestion.
7.t20
--> 8.s18 9.t15 10.u17 11.o19
--> 8.v20 9.p17 10.n19 11.r21
so probably black will play somewhere else on the board: 8.p15 or 8.k4 or 8.o13. Does anyone remember how can black counter the 7.t20 ?
7.v20
--> 8.t20 9.t21 (10.s18 11.t15 or 10.u18 11.q18)
--> 8.w20 9.t21 (10.s18 11.t15 or 10.u19 11.q18)
--> 8.u18 9.r18
If both 7.t20 and 7.v20 are crushed by you guys then I suggest a weaker & safer alternative
7.u19
--> 8.t19 9.t17 10.v20 11.o19
7.o19
--> 8.p15 9.l15 (10.k18 11.p17 or 10.m21 11.i18)
--> 8.q14 9.l15 (10.l19 11.n16 12.p19 13.p17 14.n20 15.h18 16.i16 17.g15 18.h13 19.j16) or 9.s13
this one is also a tough one |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.o13 9.n15 10.n19 ... ? And for example: 11.m20 12.l14 13.n14 14.o17
Now if you look, it may be true that white reaches the bottom ie. 15. p17 16. p15 but what good has it done us? It appears to me that blacks pegs at p15 and o13 give him (it) pretty good chances of blocking our lowly peg at r16 from ever MAKING IT to the top!
But maybe I'm wrong. We need the Team to analyze and post some more concrete hypothetical lines at this stage. (mc)
We still need to focus and re-check our analysis because we are humans and we do all make mistakes, don't we?
David, my example about 7.p19 isn't the best. My bad. White would play 8.o21 9.h17 10.h19 11.o22 ... but it also would be defeated by something like this: |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.p19 8.o21 9.h17 10.k4 11.f5 12.h10 13.m8 14.h19 15.o22 16.q20 17.j18 18.j14 19.n20 20.o16 - here I did not focus on the very moves but more on strategy.
7.p17 is still not an option
Bob, as you see, my 'critique' was not quite right, since Maceij himself corrected his line. Sure, perhaps Black has no viable path along the bottom, but it's still a big board; white has to find a way to reach the top!
You and I are discussing strategy using exposition of 'general principles'; sometimes these guys do that, but more often their analysis technique is to 'play computer' and compose SPECIFIC lines using a sequence of good, plausible moves. The Commentator diagram at the left makes that easy to do - just play out a sequence on the board, see how it looks, and if it looks good, POST IT! Then, wait for the crowd to shoot you down!
When I try to compose a sequence of future moves, I usually make a blunder within my first few moves - completely negating the value of the rest of the line. It helps if you are a real strong player, since it's more likely that your sequence will become the truth. But it also helps to be real, real careful and try, try again. THESE GUYS ARE PUTTING A LOT OF EFFORT INTO THIS!
Look at what's happening here - the vast majority of lines posted tend to be trashed anyhow; there were VERY FEW posts that considered Twixtbot's latest r20 move. So if you are inclined, post a sequence of moves and feed it to 'The Star Chamber' to be chewed up and spit out. And, if you hit on something that wins the game for us, wouldn't it be great!?
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
177 hours remain. That's 7 days, 9 hours.
(mc)
If these moves are possible, then it is probbly better to keep it as a threat and play more in the center of the board, then or we gonna play one of these moves or we can go on the left and our center is better.
If o19 is working, and I don't know if it is, then it is maybe a good option.
To illustrate what I mean, let's see some Maciej analysis, I can see that l15 is interesting in some variations, so is it possible to play it directly or it will allow TB to defend everything with some back up moves? (Or another move like this.)
Is it possible? Or something else in this way?
Maybe 7.i12?
I like the idea with getting the top with sth like |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 or 7.h14 but 7.t20 looks also not bad (by giving the initiative to TB, Tb will probably play somewhere in top-left corner of the board)
Surely i12, simple move but quite strong I guess, is one of the moves I wanted to talk about.
We should analyze more and we find a strong move in the middle I am basically convinced that it will be better than a move in bottom right.
@David, there is a tiny blunder in your line: |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.l15 8.p6 9.k7 10.l20 11.k17 12.j19 13.t20 14.s18 15.t15 16.u17 17.o19 18.o23 19.o21 20.n21 21.q22 22.q18 23.n19 24.m18 and not 25.o17 but 25.p17 and things are pretty good for white. But I think black would go 14.p14 and there is nothing I can find for white on the right side (15.o13 16.o16 17.o19 18.n19 or even worse case 15.p15 16.o16 17.q13 18.s18).
So yeah 7.l15 weakness is bottom continuation like 8.l20 or 8.j20.
I will be back in several hours and will try to make a list of Pros and Cons of each of these moves: 7.t20 7.v20 7.u19 7.o19 7.l15 7.i12
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.l15 8.p6 9.k7 10.l20 11.i18 12.j11 of course this depends on white’s response to p6 13.n10 14.k9 15.m8 16.j7 17.i6 18.h20 19.h21 20.j19 21.g19 22.i17 23.h16 24.h15 25.g14 26.g13 27.f12 28.f9 29.g10 30.h6 31.c4 32.d8 but white has lots of choices. Maybe L15 works?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 9.d10 10.l20 11.f18 12.e12 13.f11 14.g11 15.h10 16.g16
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.l15 8.p6 9.m7 10.l20 11.i18 12.k11 13.n10 14.l13 15.n14 16.h20 17.t20 18.s18 19.t15 20.u17 21.o19 22.m18 23.p21 24.o16 25.p17 26.p14 27.r13
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.l15 8.p6 9.m7 10.n7
I can’t find a good response for white, in either the 7.i12 or 7.l15 line, to 8.p6
155 hours left now; that is 6 days and 11 hours.
mc
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.p6 9.k7
White has secured the top, and has a tempo threat on the bottom right, as well as threatening i16.
So, it would seem black doesn’t have time for a downstream peg in the top right.
By the way, this T20 move is called the Medcalf defense. One way for black to play locally is 8.s18 9.t15 10.p15, but 11.p17 looks like a good local response. On the full board, how about 8.m10
7.i12 and 7.l15 are mutually exclusive so when we get one of them we won't have the other one (or it won't have a desired effect).
7.i12 have pretty good connection to top and divides the black player into two halves. But if there exist straightforward attack from black, then TB will go that way:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.i17 9.o19 10.p14 11.o13 12.q16 - black wins since we can't use L15 as a helper - so we need to play 11.p15
I do remember that there was some analysis where we used L15 as additional help to get bottom. So if we would like to play 7.i12 we have to be sure that it is not an instant win for black:
7.i12 8.i17 9.o19 10.p14 11.p15 12.n15 13.q13 14.q12 15.r11 16.r10 17.s9 18.r5 19.q10 20.l4 -- black straightforward attack (seems white?)
7.i12 8.i17 9.o19 10.o14 11.p15 12.p16 13.p17 14.q13 15.l15 - again we can't use L15, so we have to let the opponent fight for the right side.
7.i12 8.i17 9.g17 - maybe whites hope
11.r12 12.q13 13.q14 14.r11 15.s10 16.s9 17.t8 18.s4 -- black straightforward attack (does I12 is strong enough to actually gives us the top side? - that situation is almost forced by black so we should analyze it if we want i12)
Another options to check are 8.n20 8.j20 8.i20. The 8.p6 probably won't work here 9.e12 10.c4 11.n6.
7.l15 - has extremely good connection to R16 but it doesn't provide us a bottom connection nor do not help much on the top side. So moves like 8.j20 8.p6 are possible.
To be honest I feel like both 7.i12 and 7.l15 can be winning but I propose the right bottom corner. If you don't like 7.v20 or 7.t20 then I vote for 7.o19... When I checked the 7.o19 8.p6 my first idea was 9.i12 ... Maybe we should investigate harder ;-)
In this position |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.i17 9.g17 white seems maybe okay.
Hard to predict.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 9.d10 10.l20 11.f18 12.e12 13.f11 14.g11 15.h10 16.g16
Or 9.g17 10.k4 11.f5 12.g12
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 9.e12
First of all |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.j20 9.o19 10.p19 11.p17 12.n20 or 11.m18 12.l19 -> an examples that shows that R16 peg is not kind of blocked here. (of course there is an open 9.p19 possibility now, right David? ;-) but I will leave that scenario for now)
@Florian, Yes 7.i12 8.i17 9.g17 can be a hope for white (as I said it earlier). But let's build a strategy for black to actually us it favorably.
1) Black can go that way first 7.i12 8.j20 9.h19 10.i17 11.g17 - now the R16 is not connected to the bottom. So if there exist an escape from the ladder in the left 7.i12 8.j20 9.h19 10.i17 11.g17 12.c4 13.m8 14.h15 15.f15 16.g13 17.e13 18.f11 19.d11 20.e9 21.c8 22.h6, 7.i12 8.j20 9.h19 10.i17 11.g17 12.h15 13.f15 14.g13 15.e13 16.c4 17.g12 18.l6 19.o7 20.n10 21.n9 22.l11 23.k11 24.j10 25.h10 26.p11 then black could go that way directly. I don't think there exist. But forcing this reasoning It would be better for black to first gain some pegs on top side and then do the attack.
2) 7.i12 8.j20 9.h19 10.k4 11.f5 12.h10 13.f11 14.f6 15.d10 16.f14 17.g13 18.i17 -> black wins! So moves 8.j20 8.i20 shouldn't be underestimated in this case (I am showing them for quite some time and there is a time to show their strength).
Of course it's only a strategy (I am only a human and don't know if white could do something better in the process), just trying to find any weakness points for the I12.
So it seems that it's quite hard for me to reject (7.i12 8.j20 9.p19 is an option for white) or approve the 7.i12 - probably lack of skill ;-)
The 7.o19 is another story. But we couldn't find any black response for that either.
Let's try the very basic:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.j20 9.h19 10.i17 11.g17 12.c4 13.g13
Maybe we still can connect on the right side?
And we can have some hopes like:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.j20 9.h19 10.i17 11.g17 12.c4 13.p17 14.n19 15.n18 16.p20 17.t17 18.t19 19.v18 20.v20 21.g13
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10
or
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.p10
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 9.e12 10.h19 11.f19 12.h15 13.e17 14.g21 15.e21 16.g13 17.d14 18.k4 19.f5 20.e8
Florian, my second line was cleaner |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.j20 9.h19 10.i17 11.g17 12.h15 13.f15 14.g13 15.e13 16.c4
For now I can't find any good response for |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.j20 and for |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10
Both
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.i12 10.j20
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.j9 10.j20 seems like black to me. Maybe tomorrow I'll find something. So maybe somewhere in the top left which is the counter to J20?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 can also be a tough one if 9.e12 will be refuted by David.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 8.p6 9.e12 10.h19 11.j18 12.l19 13.k20 14.j14 and if 15.d19 16.g13 threatens e8 again.
Does 7.T20 run into trouble?
But |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.i12 10.j20 11.i18 seems much better for white.
@Florian, I would play |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.i12.
I think Alan would agree, the variation
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.i12
is a mess. I keep looking at 8.p6 and finding nothing but doom and gloom for white. So maybe we should shift to stuff like 7.o19 or 7.t20. If that also gets messy we might have to revisit i12, but a clean win is ideal, yes?
I favor Steven’s idea 7.t20 how about 8.m10?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.i12 is a mess, 8.p6 is one reason why, so let’s look at 7.o19 or 7.t20, like 8.m10 maybe.
I mean this "But |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.p10 9.i12 10.j20 11.i18 seems much better for white."
This is also my thought.
As I said in my very first message, personnally I don't believe in t20, v20 or u19 right now, only o19 was for me something and only if it is possible, cause it is not an "all in" on the right side, it gives some strong threat in the middle/left, for example with your i18. To be honest in a very high level game, I reject almost directly the three moves t20, v20 or u19 by intuition, it's like if my brain is saying "it cannot be the best", I hope it's not bullshit from me :D
With this analysis of Maciej, I think o19 is playable and my vote goes also for it.
Is anyone who still believe in 7.i12? -- should we wait for my brother spd_iv who actually was the first to suggest it?
So the battle is between 7.o19 and 7.t20, right?
Then I would consider o19 AND many other moves in the center that we did not even checked but we tried 2 important ones and it seems that it was not really working.
But finally, if o19 is playable, it just seems to be the best, it is "between" all the moves we considered, it gives many opportunities.
We still can go from the right or from the left and use in the end of variation the o19 peg to connect to the bottom.
This seems to be a useful tempo move which blocks o19 from connecting to anything on the left. After 9.p21 (not necessarily the best) black could play 10.p6 with roughly the same plan as before.
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
132 hours remain; that's 5 days 12 hours. mc.
I was briefly considering 7.t17, but it has similar issues: 8.o16 9.q14 10.t21 11.o19 12.n21 13.p21 14.q20 15.p17 16.p22 though that might not work out quite as bad: 17.k19 18.l20 19.i18 20.k18 21.i14.
But if 7.o19 8.m20, how about 9.t17 then?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 9.t17 10.s18 11.u19 12.u23 13.r22 14.t21 15.q20 16.p14
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 maybe 9.q9 and then
10.s18 11.g16 or
10.i17 11.t15 12.r13 13.u13 14.s11 15.u10
Instead of o16 how about 8.m10
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 9.q9 10.i17 11.t15 12.r13 13.u13 14.s11 15.u10 16.t8 17.t7 18.s6 19.u9 20.t9 21.v8 22.t4 23.s8 24.o5 25.l6 26.n8 27.r6 28.q5 29.p5 30.p3 31.m8 32.m6 33.n10 34.k5 35.j5 36.i4 37.h4 38.h2 39.f5
10.o11 11.o13 12.q12 13.p15 14.s18 15.t15 16.u17 - if Q12 is connected to black then there is no hope for white
10.o11 11.t15 12.q12 - the same situation like above
10.o11 11.p11 12.q12 13.n10 14.h19
OR
11.s8 12.p14
12.o4 13.u7 14.p14 15.t15 16.s18 17.u18 18.s11 19.t13 20.v15
12.o4 13.u7 14.p14 15.t15 16.s18 17.u17 18.s13 or
11.u9 12.o4 13.u5 14.t11
11.u9 12.s4 13.q5 14.p5 15.o6 16.n6 17.o4 18.s11
Personally I would pick 10.o11 for black and can't see anything for white.
It's not like I think this is a bad move 7.t20 - (perhaps there are not non-messy moves for us) but I feel the same as Florian - "that playing in bottom right is very probably not the "best". And finally I see the the Clear difference between 7.t20 and 7.o19. Thank you!
I meant - if Q12 is connected to the RIGHT then there is no hope for white.
11.s8 12.p14
12.o4 13.u7 14.p14 15.t15 16.s18 17.u18 18.s11 19.t13 20.v15
12.o4 13.u7 14.p14 15.t15 16.s18 17.u17 18.s13 or
11.u9 12.o4 13.u5 14.t11
11.u9 12.s4 13.q5 14.p5 15.o6 16.n6 17.o4 18.s11
As expected bottom right seems not good, we could not find a strong move in the middle neither, so... o19, the move between everything stays and seems to be the good one.
Again, why we dropped a move like 7.h14 ?
I still believe somewhere around the middle, there is a better move than o19...
It's like t20 is by far too direct, o19 is nice but still too direct.
I was talking about h14 right now.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.h14
Is it something? I mean at least it counters your j20 for black not like i12. :D
Or you could scroll back and look for previous discussion of this line.
Or, you could show us a specific variation. Like you said, time is passing. We need to work on all branches that look promising.
Maciej, against 7.o19 my preferred response is 8.m20.
If 7.t20 8.o16 9.q9 10.o11 I prefer 11.g16 not H16.
Well there |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.h14 8.h17 we play o19 or t20 I would accept it easily. We can even start with f15 first.
I don't think |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 is a problem for us.
The position is unclear but we have many possible moves there. I don't know 9.k19 for instance? Or something else on the left which gives a threat on k19.
7.h14 8.h17 9.o19 10.p6 11.k7 12.o4 13.j5 14.l20 15.p21 16.p14
If you play this |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 9.k19 10.l18 I just gonna play on the left, certainly not 11.p21 ... just visually I can say that white lead is probably huge.
In the other variation, since we put h14 maybe t20 is the move, not o19, not clear.
15.21 seems just stupid in this variation? Maybe you wanted to play m20 for black instead of l20?
Even there it is complicated and p21 is still not good, probably some threats around k19...
It is too hard to analyse like this cause there are MANY possible moves.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 9.k19 10.l18 11.j17 this is how I interpret “play on the left” 12.k13
I agree my h17 line was flawed. How about 7.h14 8.h19 and now 9.t20 doesn’t look good for white because of 10.o16. So if 9.o19 maybe 10.m21 then if 11.p21 12.p6 13.k7 14.o4 15.j5 16.p14
Meanwhile, 7.t20 8.o16 9.q9 10.o11 11.g16 carries the the threats of either reaching d6 or q9. White might be able to play u9 at some point, which threatens white T15.
Or 14.h14 15.f13 what do I miss?
17.p14 18.l8 or 17.l10 18.h14 19.p14 20.n13
Meanwhile after 7.o19 8.m20 9.k19 10.l18 11.j17 12.k13 I don’t see black’s doom.
My current perspective (considering your notes as well)
Against 7.t20 -- 7.t20 8.o16 9.q9 10.o11 11.g16 12.i17 13.e17 14.g8 15.t15 16.q12 (17.p14 18.l10)(17.u9 18.s11 19.u13 20.t9 21.s8 22.v10 23.j13 24.l8)
For: 7.o19 -- 8.m20 9.k19 10.l18 11.p21 - why it is not good?
For: 7.o19 -- 8.m21 9.n21
Open: 7.o19 -- 8.j20 9.m18 10.m10 11.d10 12.k18 13.p17 14.k11
For 7.h14 -- 8.h19 9.o19 10.m21 11.p21 12.p6 13.k17 14.n19 15.h18 16.i17 17.i16 18.g17 19.o13 (20.k4 21.f5 22.h10 23.n6) (20.l10 21.h10 22.g8 23.n6 24.m8 25.i8
Against 7.h14 -- 8.h19 9.o19 10.m21 11.p21 12.o13 13.n15 14.l15
Against 7.h14 -- 8.h19 9.o19 10.m21 11.k17 12.n19 13.h18 14.i17 15.i16 16.g17 17.p21 18.o13 (19.n15 20.o17) (19.q13 20.p11 21.r11 22.q9 23.s9 24.s4 25.r7 26.q5)
I'll check back tomorrow. By that time, perhaps Michal will have returned to sort out the mess. (meow)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 9.q9 10.o11 11.g16 12.i17 13.e17 14.g8 15.t15 16.q12 17.u9 18.s11 19.u13 20.t9 21.s8 22.v10 23.j13 24.l8 25.j7 26.i7 27.d11 28.d9 29.e8 30.e11 31.g9 32.g12 33.i10 34.i13 35.n14 36.p14 37.k16 38.j15 39.l18
Or 26.j9 27.h11 28.h10 29.d11 30.d8 31.e9 32.e6 33.f7
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 9.k19 10.l18 11.p21 12.p6
Or 11.j17 12.k13
Against H14: 7.h14 8.h19 9.o19 10.m21 11.p21 12.p6 13.k17 14.k4 15.f5 16.h10
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 9.q9 10.o11 11.g16 12.i17 13.e17 14.g8 15.t15 16.q12 17.u9 18.s11 19.u13 20.t9 21.s8 22.v10 23.j13 24.l8 25.j7 26.j9 27.d11 28.d9 and now 29.h11
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 9.q9 10.o11 11.g16 12.i17 13.e17 14.g8 15.t15 16.q12 17.u9 18.s11 19.u13 20.t9 21.s8 22.v10 23.j13 24.l8 25.d11 with two main branches:
26.d8 27.j7 with two twigs:
28.i7 29.e9 30.e6 31.h6 or
28.i9 29.e9 30.e6 31.f7
The other main branch is
6.d9 27.e8 28.e11 29.g9 30.g12 31.i10 32.i13 33.n14 34.p14 35.k16 36.j15 37.l18 38.k13 39.l14 40.l11 41.l10
In this position |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
The best move, but only if it "exists" is arount i12, cause it keep the threat bottom right, add a threat bottom left and reinforce the center.
Now, is this move existing?
We tried 7.i12 but it failed, why? Cause of the 8.j20 from Maciej, this move defend move threat on left and right, so here I STOP to analyze, it is useless to continue cause the reason I played my previous move is checkmate, maybe it's still ok but it means that I have better.
Same, we tried 7.t20 directly, already I don't like, it is a "direct" approach, try to win fast when center is not strong enough to do it. Just 8.p10 and I STOP my analysis here, why? Cause I prefer the previous position, it is useless to continue, I prefer white here |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 than here |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.p10 so if I would have an interesting variation there, I just play it before t20.
We know that we have 7.o19, a move that provide many possibilities cause we can go to it from left and right sides.
We did not find the good counter for black at the moment, it seems interesting.
8.m20 is clearly something, but after something like 9.k19 same, I STOP my analysis cause I don't think this was a good response for black, white seems better than before the 2 last moves.
Now 7.h14 which seems nice visually, we have this problem: 8.h19 9.o19 10.m21 and I STOP my analysis, why? Cause I don't like white last move with this black response, it seems that the position was better just before, so I would ask before saying 7.h14 is to forget, what about t20?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.h14 8.h19 9.t20
Now we have h14, so t20 is more relevant, maybe it is something Maciej or you have also a bad variation for this?
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20
I'm seeing 7. h14 , 7. p19 , 7. t20 , 7. o19 , 7. i12, 7. t17 , 7. i15 , 7. p17 , 7. v20 , 7. L15 , 7. u19 .
Of these, the most positive consideration appears to be with 7. p19 , 7. o19 , 7. t20 , 7. i15 .
The blocks on the right are 7. t20 , 7. t17 , 7. v20 and 7. u19. Of these, the most analysis appears to be on 7. t20 . If nothing else, then there is a greater chance we have included TB's next move in the analyzed lines, giving us a head start for move 9. I just hope that 7. t20 is effective in blocking or giving us a connection up and down; otherwise to me it seems like a waste of time.
The tactical interference moves are 7. p19 , 7. p17 and 7. o19 . 7. p17 is not favored, 7. o19 is OK, but more serious attention is devoted to 7. p19 , so there'd be a better chance of predicting TB's reply. However, 7. o19 has fans, like Bob and Alan, and the rest of the people are starting to come around. I really have no preference between these two, and personally I'd rather see one of these interference moves than 7. t20 .
The 'profound', whole board moves are 7. h14 , 7. i12 and 7. i15. People are afraid of TB gaining an edge on the lower left, and this is why 7. h14 is preferred. Those lines are pretty wacky, and I do not know how to visualize a board 20 moves ahead, with stuff all over the place. If we are here to have the most FUN (at higher risk of losing), then I like 7. h14 . If we are trying our utmost to win at any cost, then I prefer one of the blocks, and if you want me to pick one, I'll go with 7. o19 . Why? This was an 'instinctive' choice, before all the analysis, and I'd like to put some faith in humanity.
I hope this helps narrow down the discussion. The remaining time is now 95 hours, or 3 days, 23 hours. I advise that excess time used NOW is REALLY GOING TO HURT US in these next few moves! (roar)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.h14 8.h19 9.t20 10.o16 is a quick win for black, no?
After 7.t20 8.p10 9.k13 white threatens to connect to the top several ways, and to the bottom in widely separated ways.
Are you willing to accept Maciej as the highest rated player?
But it's interesting cause if a move like this works, then white gonna probably win the position.
Looks to me above, that you are moving AWAY from h14 and starting to campaign for T20. I guess you are not fond of the blocks at o19 and p19 any more? The Celuch brothers kind of liked these moves, as I recall. Let's hear from more people, now.
I should point out that, while Maceij has a STELLAR record in games against humans, he remains the only player among the top 10 who has NOT had the 'pleasure' of being trounced by Twixtbot. There is something to be said for experience. I'm delighted he's participating, since in his case, he is not out for vengeance!
Florian likes h14 FIRST, and followed by T20 IF the bot responds to h14 with h19. Otherwise, back to the drawing board. I THINK David likes T20 straight away. Perhaps another of my silly questions, David, but if 7. T20 and the bot simply pushes right with 8. T19 , then I don't understand how T20 has helped us. In that case, we are STILL going to be playing to block with O19 or p19 , or play on the left side with h14 (or my h17 - heh). Certainly we can't allow TB to win easily across the bottom.
And if T20 first and h14 second IS THE SAME as h14 first and T20 second, then I believe we can 'flip a coin'. If we are including a blocking move, then we flip a 4 sided coin. Really - we can agree to use a result from a racetrack of sports score, if it comes down to that.
mc
I feel 7.t20 really risky, I believe the Bot gonna play something and take the lead from now but maybe I am wrong.
7.h14 is really strong if... it's not losing in 2 or 3 moves and I'm not sure about that :D
7.o19 is maybe the safer move.
Actually, Alan and the Celuch brothers need to say something and we need to PLAY!!!!!
We gonna lose by time :D
white would just respond 9.t15 and threatening both u18 and o19, this position is probably no more than a clear white win.
M.C, 7.p19 was rejected, wasn't it?
7.i15 has not been analyzed so I don't know what to think about it. My intuition says that it looks not so strong but I may be wrong.
Both 7.t20 and 7.o19 creates a completely different course of events:
a) |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.t20 8.o16 9.q9 10.o11
b) |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.m20 9.k19 10.l18 11.p21 12.p6
We take the initiative in different parts of the board. Both have their own pros and cons. I feel that 7.o19 is safer. But I can't really prove it.
David, another example:
7.t20 8.o16 9.q9 10.o11 11.g16 12.g8 13.t15 14.q12 15.u9 16.t11 17.s10 18.s13 19.j13 20.f14 - black threats I17 or L8
I have a feeling that current move is an important one. Against any human it would take me about 10 - 15 min to decide which one of those moves to play 7.i12 or 7.o19 or 7.t20 depending on my mood of the day.
Whatever you choose I'm fine with it. It could be either 7.t20 or 7.o19 or anything else ;-)
I will have some break now so good luck guys!
We are still looking then, at coin flipping? T20, O19, P19, H14 .
Eventually, I go out to do laundry, so a few hours will go by if no move before then. Relax, Florian - while not the route we want to take, those vacation days WILL click in if needed (20 of them), and all we need are a few quick moves - at 36 hours bonus per move, we'll work our way up to 240 hours in no time; provided the game is not over by then (LOL).
92 hours remain. That's 3 days and 20 hours.
mc
At this point, if I had to play now something, I would pick o19 probably.
It's sad that we could not find a move around i12, maybe h14, maybe something else that could convinced us.
The funny part is that I 100% agree with Maciej when he is saying "Against any human it would take me about 10 - 15 min to decide which one of those moves to play 7.i12 or 7.o19 or 7.t20 depending on my mood of the day."
Done; I expect a reply with 15 minutes.
I hope, I hope that Twixtbot plays a move we've done some serious analysis on already; that will speed up move 9 for us.
Thanks to Bob, for suggesting O19 FOUR DAYS AGO! That's how long it took everyone else to come around. No, I'm kidding - I know that other people looked at this move, and lots and lots of work has been done.
Still, if we win, thank Bob. And if we lose, blame Bob.
Ha ha ha ha ha ha ha ha ha ha ha ...............!
Move List
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
Did anybody predict that? Did we do some analysis already?? One needs to go back and look.
Now, we have 120 hours left; that's exactly 5 days. This should be doable, but I think, maybe, a lesson has been learned regarding time management.
Back to work, Team! (mc)
I don't like this cause now the game will be massively tactical after this move, it's like the strategic part is over, so I hope we are ahead cause it will be harder.
We can check 9.q8 first, which is probably failing but still need to analyse it.
And again, we have all the moves around i12, we could not find a good one previous move maybe now...
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
I did the research. Bad news. Nobody. Nada. Nothing. Back to the drawing board. Either L8 is too brilliantly profound for anybody here to have fathomed, or it is flawed, in which case we crush it! There may be no law of excluded middle, so possibly neither is true.
Time becomes more of an issue now, since we are starting from scratch. I'll go get my scratching post.
(meow)
1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
Florian throws out 9. Q8 . I'll throw out 9. H10 . My turn to actually contribute a move. There is nothing profound about my move, and I'm certain that everyone sees what I have in mind, so I won't elaborate right now.
mc
It’s like Maciej said. Black has more reason to fear white D10 than white should fear black K4. Black would have to try to block below d10 somewhere. 9.q8 might be a good move, but black could respond in many places on the right side. (Did I get that right? Yeah) With fewer black responses to consider, maybe we can find a plan with more confidence of success.
Amazing, he really surprised me! TB has started to create a mess. Is he really in trouble? Or it's just his normal I feel it's just the beginning of the true messiness.
Long time ago I said something like this: "if black plays on the top half, then we respond D10".
So that's my first suggestion: 7.d10
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e8 then 11.f9 seems to improve threats to the bottom AND to reach m12 without losing connection to the top. Maybe black will be able to play E8 at some future point.
There are examples of bot vs bot games where one side seems to suddenly turn and smash into a brick wall. Maybe we got lucky here.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e8 11.f9 12.f6 13.i7 14.g9 15.j9
For 9.d10 :
10.e8 11.c8 12.f6 13.g3 14.f2 15.c4
- 16.f11 17.e13 18.e14 19.c14 20.g13 21.d17 or 20.e18 21.g14 : we win the corner and are able to defend straightforward attack
10.e12 11.d15 12.c13 13.j9 14.j7 15.f11 : we have very good connection to M12 peg
10.l4 11.h6 :
- 12.i11 13.f15 - pair L4 - L8 is not connected to left
- 12.i11 13.t7 : we can still play on the right against L4 (T7 and R16 like each other thanks to a connector T11/R12 and T7 and M12 like each other thank to P9/Q10)
here is my very first hypothetical line:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.j20 11.i18 12.e8 13.c8 14.f6 15.g3 16.f2 17.c4 18.g19 19.g17 20.f17 21.j21 22.k18 23.h20 24.h16 25.q8 26.i14 27.p17 28.q12 29.o13 30.n9 31.j9 32.k6 33.i7 34.i5 35.h10
- even though white wins it's not an easy win.
"- even though white wins it's not an easy win." Somebody is expecting an 'easy' win against the silicon monster??
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
OK, I started looking at lines with my 9. h10 . I'm not going to post one riddled with terrible moves, but I'll tell you what I saw - h10 connected to d6, d6 connected to the top, h10 connected to m12, everything going great only m12 had problems connecting to r16. Of course, we don't HAVE to go through r16 - there is a nice big space on the left side to work with.
Looking forward to your analysis. mc
https://www.littlegolem.net/jsp/game/game.jsp?gid=2133161&nmove=7 of course black made mistake in move 8, but after 7 moves I believe black had 100% win.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 now I like only 9.d10 or 9.h14 (my brain told me these 2 moves are the best here).
About H14, If TB will move in top-left then we play in top-right. If TB will play in bottom-left then we in top-right/middle:
example:|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.h14 10.m21 11.j19 12.g10 13.n8 14.n9 15.o10 16.o11 17.p12 18.p13 19.q14 20.n19 21.o13
I think TB is not a real monster, he still makes mistakes like in the game with passenger and I belivie he is in trouble right now. I doesn't matter if we play now 9.d10 or 9.h14 in both cases we already win the game easily :)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.g21 11.e16 and now
12.f19 13.f18 14.d20 15.h19 16.j20 17.m20 18.k18 19.o13 20.l4 21.q8 22.q4 23.s5 24.r5 25.q6 26.t4 27.l2 or
12.e20 13.h19 14.j20 15.m20 16.o13 17.k21 18.i22 19.j19 20.g17 21.i17 22.f15 23.h10 two branches
24.j15 25.l14 26.l16 27.n15 28.o16 29.q14 30.p14 31.r12 or
4.i12 25.k11 26.l14 27.h14 28.g13 29.j13 30.h9 31.j9 32.l4 33.f5 34.g11 35.j5
White probably has something simpler.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.g21 11.e16 12.e20 13.h19 14.j20 15.m20 16.o13 17.k21 18.i22 19.j19 20.g17 21.i17 22.f15 23.h10 24.i12 25.k11 26.l14 27.h14 28.g13 29.j13 30.h9 31.j9 32.l4 33.f5 34.g11 35.j5
I vote for 9.h14
but 9.d10 is also playable so I will accept this.
If someone has different idea, then probably we will waste the time to find "the best" move. I think we should make a move in 24 hours from now. Just everyone please vote for your move.
Michal, I agree that TB seems not to be perfect, at least his very first moves. But the further the game goes, the closer TB is to perfection.
9.d10 is a strong move but it can make things pretty complicated (for humans). Example in my previous post.
9.h14 is also pretty complicated - I really don't know what to think about it. See example: 10.l4 11.f5 12.h10 13.t7 ... ?
9.q8 - as most of moves in the right top corner - is sort of "all in" move. I mean if black somehow connect to the right side and in the process will get the L4 then the battle in the left would be very hard. See example: 9.q8 10.q12 11.i12 12.l4 13.f5 14.h10 15.f11 16.e9 17.e8 18.g10
9.h10 - if it's connected to the topside then it can be a strong move. We should check for a trap here. See example - the trap: 9.h10 10.f6 11.d10 12.e8 13.i7 14.l4 15.g3 16.h2 17.c8 18.d2 (David this is similar to your previous line for 9.d10).
Maybe we should follow the plan:
1) Look closer to 9.q8 and check for a quick win. If there is not such thing we should check next move. If there is a quick win, we just play it.
2) 9.h10 if it's connected to the top we should play it, if it's not then go to the next move.
3) 9.h14 what if black plays top middle 10.l4? is 11.t7 good enough to stop black from connecting the right (example9.h14 10.l4 11.f5 12.h10 13.f15 14.f6 15.t7 16.q10). If it's strong enough we should play it, if not the go to the next move.
4) if we are here then we should play 9.d10
As for D10 this variation is not ok:
9.d10 10.e8 11.f9 12.f6 13.i7 14.l4 15.g3 16.h2 17.c8 18.d2
We have to play this:
9.d10 10.e8 11.c8 12.f6 13.c4 I accept this position. Do we accept this? (it's somehow similar to Go game, isn't it?)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.h14 10.l20 not m21 11.j19 12.h6 13.f5 14.g10 15.n8 16.n9 17.o10 18.i17 19.i16 20.j15 21.h18 22.k13 23.i12 24.l11 and now
25.o11 26.p15 27.p17 28.q13 29.n14 30.m18 31.o15 32.o23 33.o21 34.p21 or
25.o13 26.q14 27.q12 to block black o11 28.m18 29.p21 30.p17 31.q18 32.q15 or
27.p15 28.r18 29.n12 30.m18 31.p21 32.p17 or
29.q18 30.o11 31.p12 32.p13 33.r13 34.r14
If I were the TB, in response to 9.d10, I'd play 10.e12
but I can't take it any further than that
someone else suggested it (Alan?); I merely chimed it to say (likely erroneously) that it looked like it sewed up the bottom
There are several other candidate moves being thrown around now, including two from your brother, and a different approach from Florian. Don't think of this as a 'democracy' or 'government'. Think of it more as a 'think tank'. This game is not a 'contest'; it is an 'experiment'. I am counting on the discussions to narrow in on basically acceptable moves - and if not provably the absolute best, then at least not flawed, and acceptable to The Team. No voting, no ratings.
Don't bother with the hypothetical question for me: what shall I do every every single Team member posts a different preferred move, and with rational justification to accompany it? While I hope that does not happen, I WILL deal, and without looking at ratings or votes. I have some other ideas, should the need arise, that I'll keep to myself for now.
spd_iv, I expect to NOT spend 5 days on this move like we did the last one, but it's only been one day so far. We are still all right with time. I will do my job seriously, as this whole project was 'my baby'; of course, I'm delighted at the enthusiastic response. I will NOT allow the game to be lost on time forfeit.
Team: I believe a 'lesson learned' from the time spent on the last move is that it is ALMOST USELESS to attempt to predict the response from Twixtbot; at least for now, on a wide open board. Thus, many of you analyzed and posted long hypothetical lines that turned out to be trash. In order to get our time up, I believe MORE effort should be devoted to predicting the response from Twixtbot, so we have a reply already analyzed, and LESS time should be spent on analyzing lines 20 or moves deep, which are bound to NOT HAPPEN.
That's just a suggestion; naturally, analyze however you like! At this moment, we have 96 hours - that's 4 days, plus 20 more vacation days, if we need them. (Master of Ceremonies)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
So, for example, The Team can analyze now whether e12 busts d10; if it's a bust, eliminate d10; if not, consider other candidate moves for The Bot in response to d10, so we are READY with concrete analysis for move 11.
(meow)
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e12 how about 11.h10 then if 12.e8 13.h6
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e12 11.j9 the j9 is now connected to the topside.
12.j7 13.f5 14.i9 15.h10
12.j7 13.f5 14.h8 15.h10
12.j7 13.f5 14.f6 15.f9
Having j9 -> m12 connected to the topside is pretty good, isn't it?
After |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e12 11.h10, I think I'd play 12.j12
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 immediately 10.e8 seems sub optimal because after 11.c8 12.f6 13.c4 black can no longer threaten to take the top left with k4 or L4.
Maciej looked at 10.j20 and I looked at 10.g21. It seems to me, black is forced to try to block white’s threat to win on the left. I welcome any suggestion where else black might play. Please poke holes in our analysis.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.i20 11.h18 12.g21 13.k19 14.l19 15.m20 16.j18 17.o13 18.l4 19.q8 20.q4 21.s5 22.r5 23.q6 24.t4 25.l2 is the same as a 10.g21 branch I looked at earlier.
Even if TB plays something we didn’t examine, we may still be able to use one or more of these patterns.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.q12 11.i12 12.l4 13.d10
Is it something?
But we need to keep in mind that if we forgot one very complex variation, TB will probably play it, so if we have "safe" little improvents we need to consider it.
But, is this bad for us? |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.o11 11.o13
The idea is just to use q8 to fortify the middle and then play the d10 move, I don't know if it's possible and moreover if it is possible without risk.
Cause this is for me the great thing of d10, it is very safe and "seems" enough to win, but vs TB I'm still anxious that we are trying to win too easily and he has a possible very complex line that crush us.
This is why I'm just trying to see if it is psosible to "prepare" a little d10 staying safe.
(I see lines where we play until t15 before d10, but all olf this is maybe risky.)
Actually, I'm okay for d10, we still wait a little, Michal will give some ideas tonight?
I'm fine with 9.q8
That's bad for white?
Florian, about your last comment, I don't like |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.o11 11.o13 12.l4 13.q12 14.q6 15.d10 16.c4 17.p6 18.o5 19.n5 20.m4 21.r6 22.s5 23.t5 24.u4 25.v4 26.w3 27.g3. I cannot say why right now, just my intuition says that it is better for black.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.p9 10.p7 11.r8 12.r6 13.t7 14.t5 15.u5 16.v4 17.w4 18.x3 19.o7 20.n6 21.m6 22.l5 23.k5 24.j4 25.i4 26.h3 27.d10 with F2 as a threat.
But then what is the answer of black to p9?
I have to go to work, but I can come back. I'll try to get that move in TODAY, and before too long. A few more comments will be nice. mc
For move 10, we have looked at black moves e8, f9, e12, j10, G21, and i20. Our verdict In each case is white wins. Show me a specific refutation of one of these lines or propose a different black move 10. I regard any other analysis as a waste of time. A case in point is my refutation of 9.q8. Did it serve to convince anyone to return their attention to 9.d10? Apparently not.
bob440, after
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.e12 how about 11.h10 then if 12.e8 13.h6
Look at that absolutely NEAT box at D6, D10, H6, H10! Clever.
Meanwhile, I haven't exactly seen MY H10 given too much thought. What I see about Florian's ideas are that - instead of directly making an attack down the left side, if we PREFACE that with activity on the upper right, it might act as a safety net. I mean, whether Q8 (on the outs now, apparently) or P9 were played, if we succeed in reaching the top on the upper right, then GREAT. If TB blocks our direct attempt to win that way, the block won't extend across the whole board - we still have the edge with D6 on the upper left. And after a scuffle, perhaps what we have now in the center will HELP our chances of reaching that D6 peg.
For instance, let's look at my H10 with a line where black plays naively, at first:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8
9.h10 10.f8 11.e7 12.d7 13.g8 14.f6 15.j5 16.i5 17.i7 18.j11 19.k11 20.l12 21.j9 22.k4 23.l4 24.o18 25.p17 26.p21 27.n21 28.n13 29.o13 30.p16 31.p15
I know; stupid play, and a waste of everyone's time. However, the idea is to use our strength on the upper left to allow a connection through the middle to the lower right, through peg R16. But what I fear is that the connection between M12 and R16 is not guaranteed. Go back a bit:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.h10 10.f8 11.e7 12.d7 13.g8 14.f6 15.j5 16.i5 17.i7 18.j11 19.k11 20.l12 21.j9 22.k4 23.l4
but THIS time, TB successfully blocks the connection. Example with: 24.p14 25.q14 26.r13 27.o15 28.n15
So now look at one of Florian's Q8 lines:
Same for David's line |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.o11 11.o13 12.l4 13.q12 14.q6 15.d10 16.c4 17.p6 18.o5 19.n5 20.m4 21.r6 22.s5 23.t5 24.u4 25.v4 26.w3 27.g3
This is not the only example, but if the attempt to reach the top right were to fail, then white has a guaranteed connection here between M12 and R16, so can work from the top LEFT.
I think that Florian's idea then is not necessarily to guarantee a win on the top right, but by playing on the right side a little bit, we'll achieve some strength in the middle, to support a later attempt to win - either with YOUR D10 or even MY H10.
OK, I have to go back to work. FYI, a little about me - I am a local TAXI driver, but I work in my hometown. That means I can take an occasional break from driving to return to my home computer - like now, to post things like this.
I TRY to follow what is happening on my cell phone in my car, but sometimes I'm busy; also, I'm not logged into Commentator on my cell phone; even if I WERE, it's a pain dealing with the teeny screen on my phone. Furthermore, I'm not logged into my fake account MCx, so can't play actual game moves in my car.
So I'm running around, trying to do everything. I'll manage. Keep up the good work. I am ready and willing, if it is needed, to fork over the money (a donation to Richard, actually) and buy 10 more vacation days. We DO have time.
I know that some folks are getting 'antsy' - MOVE ALREADY. Relax. Have faith. I'm on top of all this, and will stay 'till the end.
(meow)
I have nothing against d10 actually, so I'm trying to see if we have something even better, it's like we found a probable checkmate in 12, just trying to see if there is an obvious one in 6 moves.
I could not find any strong response against d10, but my feeling is that it is still complex cause center is not fortified at the moment. If we could play a forced variation where we arrive to put 1 peg in the center, then playing d10 would be a sure win.
I studied q8, but we arrive to a position that Michal does not like, me I guess I could like it but less than d10 right now so it is stupid :D
Then I tried a very easy variation with p9, actually I had no time to analyse it but the variation I studied was an easy win for us.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.p9 10.l20 11.j19 12.f6 13.r10 14.r12 is anything but an instant win.
Against Q8
8.l8 9.q8 10.o11 11.o13 12.l4 13.q12 14.q6 15.d10 16.c4 17.p6 18.o5 19.n5 20.m4 21.r6 22.s5 23.t5 24.u4 25.v4 26.w3 27.g3 28.g8
Or 15.q4 16.r8 17.s9 18.p4 19.o3 20.t9 21.n5 22.p9 23.m7 24.l9 25.d10 26.e12 27.g11 28.f6 29.c4 30.e8
We could spend hours talking about these lines. Do you really think you will find something simpler than D10?
For D10
9.d10 10.e12 11.j9 and then
> 12.j7 13.f5 14.i9 15.h10 or
> 14.h8 15.h10 or
> 14.f6 15.f9
10.j20 11.i18 12.e8 13.c8 14.f6 15.g3 16.f2 17.c4 18.g19 19.g17 20.f17 21.j21 22.k18 23.h20 24.h16 25.q8 26.i14 27.p17 28.q12 29.o13 30.n9 31.j9 32.k6 33.i7 34.i5 35.h10
10.g21 11.e16 and now
12.f19 13.f18 14.d20 15.h19 16.j20 17.m20 18.k18 19.o13 20.l4 21.q8 22.q4 23.s5 24.r5 25.q6 26.t4 27.l2 or
12.e20 13.h19 14.j20 15.m20 16.o13 17.k21 18.i22 19.j19 20.g17 21.i17 22.f15 23.h10 two branches
24.j15 25.l14 26.l16 27.n15 28.o16 29.q14 30.p14 31.r12 or
24.i12 25.k11 26.l14 27.h14 28.g13 29.j13 30.h9 31.j9 32.l4 33.f5 34.g11 35.j5
10.i20 11.h18 12.g21 13.k19 14.l19 15.m20 16.j18 17.o13 18.l4 19.q8 20.q4 21.s5 22.r5 23.q6 24.t4 25.l2 is the same as a 10.g21 branch I looked at earlier.
But since we are waiting to play it, we can study the position, at the moment we gonna play it, the Bot gonna respond and we gonna start to study this further.
We wait this evening if Alan, Maciej, Michal or someone else wanna make some comments and we play tonight the move.
Here is what Maciej had to say about |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.h10
h10 - if it's connected to the topside then it can be a strong move. We should check for a trap here. See example - the trap: 9.h10 10.f6 11.d10 12.e8 13.i7 14.l4 15.g3 16.h2 17.c8 18.d2 (David this is similar to your previous line for 9.d10).
David, I DID note and play through Maceij's trap line after my h14 - yes, it IS a serious and clever trap. But I didn't take it as a bust - just something to be careful to avoid.
The basis for that trap is that black block's White's attempt to reach the top; would not that sort of trap (or similar) apply to any, or at least most attempts at play on the upper left side? ( I have not checked specific lines.)
Michal wants to move; Alan's been silent; David, Maceij, Florian all push for D10. I'd have to scroll up to check for Bob, or any other opinions.
So I won't wait all day: 4pm NY time I move. That's 2 hours from now if any desperate words to stop me.
OK?? (mc)
I did some work; this was Bob's concern, but it looks to me like 13. i12 handles this. (meowx)
Against q8
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.o11 11.o13 12.l4 13.q12 14.q6 15.q4 16.r8 17.s9 18.p4 19.o3 20.t9 21.n5 22.p9 23.m7 my suggested 24.l9 loses to 25.l10, but 24.K10 25.d10 26.e12 27.g11 28.f6 29.c4 30.e8 works for black.
Against p9
9.p9 10.l20 11.j19 12.f6 13.r10 14.r12 15.o13 16.q5 17.m6 18.n9 threatening p13 19.o11 20.o7 21.q7 22.o6 23.u6 24.r7 25.t8 26.s9 27.t9 28.v6
So the odds are we will play D10. Why wait to discuss it? If you think TB has an ace up its sleeve we could start to look for it now.
For instance in the last variation of David I would play something like this:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.p9 10.m20 11.h14
and now
12.h17 13.p21 14.p14 15.p15 16.n15 17.q13 18.q12 19.r11 20.r10 21.s9 22.r5 23.q8 at this point, black has to block the top right and the answer is the famous p10.
And if black defends directly on top it failed also:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.p9 10.p7 11.o7 12.n6 13.q7 14.r6 15.s6 16.t5 17.u5 18.v4 19.w4 20.x3 21.m6 22.l5 23.k5 24.j4 25.i4 26.h3 27.e4
But I'll try to study d10 more to try to find something for black right now!
lets start with 9.p9
if 10.l4 then 11.d10 (seems to be not as strong) so maybe 11.h14, 11.i12 or just 11.s6 would be better:
11.s6 12.p12 13.r12 14.s13 15.t13 16.q12 17.r14 18.r10 - looks complicated for white
11.s6 12.p12 13.j15 14.i17 15.h16 16.g16 17.f15 18.e15 19.l16 20.l15 21.m18 22.k13 looks much better
11.s6 12.q5 13.u5 14.q13 15.p15 16.o12 17.i12 18.m11 19.l10 20.l9 21.k8 22.e5 23.m7 24.n5 25.q7
but I think TB can find sth between P9-M12-R16
H14:
11.h14 12.p6 not sure if this is good for white 13.r8 14.r5 15.t7 16.t4 17.v6 18.v3 19.f5 20.f10
I12:
11.i12 12.p6 13.r8 14.r5 15.t7 16.t4 17.f5 18.f10 simmilar situation, so I believe we should play somewhere in top-left like 11.s6
D10:
black can play for example 10.h15 but then 11.i12 and I cannot find anything good for black so I believe one of the best move for black is 10.l4:
11.f5 12.i20 and what is next? 13.e19 doesn't look good enough
sth like 13.k16 14.k18 15.j18 16.i19 17.h19 18.g20 19.f20 20.e21 21.d21 22.c22 23.b22 24.a23 25.p17 26.o13 27.r11 28.q12 29.q13 30.r10 31.s9 32.r5 33.q8 34.p4 35.n10?
black can play for example 10.h15 but then 11.i12 and I cannot find anything good for black so I believe one of the best move for black is 10.l4:
11.f5 12.i20 and what is next? 13.e19 doesn't look good enough
sth like 13.k16 14.k18 15.j18 16.i19 17.h19 18.g20 19.f20 20.e21 21.d21 22.c22 23.b22 24.a23 25.p17 26.o13 27.r11 28.q12 29.q13 30.r10 31.s9 32.r5 33.q8 34.p4 35.n10?
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.l4 11.r12
Game over no? Black seems crushed to me :D
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.q8 10.o11 11.o13 12.l4 13.q12 14.q6 15.q4 16.r8 17.s9 18.p4 19.o3 20.t9 21.n5 22.p9 23.m7 24.k10 also loses for black 25.l9 26.m9 27.o9 28.n9 29.m8 30.r10 31.n10
So 24.m10 is necessary. Or something in the bottom half like 24.j20. I still prefer black but must give up my arrogant claim that I refuted Q8. This is messy too.
So, that being the case, I hope you’re still satisfied with 10.D9
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.l4 11.r12 12.q6 13.f5 14.h15 now it doesnt look like black is crashed, did I miss something?
Had I not looked at the above posts - practically all current 'research' is being done on D10. That being the case, I'll play D10 right now. At the very least, hopefully all this hard work has not gone for naught.
...
Done. I've got to get back to work; can't hang around my home waiting for the reply. I can post from my car now, so WILL keep you posted, and meow.
12.l19 13.h19 or
12.j20 13.m18 14.k18 15.r12 or
12.j14 13.f16 14.f20 15.k18 16.l20 17.n21 18.l19 19.r12 20.q6 21.k11 or
12.l4 13.r12 14.q6 15.f5
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.h15 12.j20 13.m18 14.k18 15.r12 16.p19 so white has to play 11.h15 12.j20 13.m18 14.k18 15.p17 which still looks good
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.p17 12.l4 13.f5 14.e12 15.o11 16.q6 17.j9 18.f6 19.k7 20.k6 21.i6 22.h2 23.e8 24.g8 25.f11
What about something like |1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.p9 10.l4 11.t9
Can it be a quick win for us?
Are we connected to the top???
I think we are still ahead. I dont think we should play somewhere on the left, I cant find anything good for black after 11.p17
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21
Some first thoughts:
11.p17 - for now I don't know what to think about it, it's more or less all in, but another "what if" 12.m20 13.p21 14.o13 ?
15.o11 16.q12 17.l15 18.n15 19.k17 20.l19 21.j20.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.p17 12.m20 13.p21 14.o13 15.o11 16.q12 17.l15 18.n15 19.k17 20.k18 21.j19 22.j20?
Note: This site is starting to load quite some time for me. Maybe we can start a new forum http://twixt-commentator.duckdns.org/game/2140447 ? We could always go back here to see our previous lines.
Ok, then we have the first candidate 11.p17 proposed by Michal (but it does not really matter who proposed which move, does it?) then it would go probably that way 12.m20 13.p21.
My thought about it are: It's kinda "all in". We give up the bottom threats just to be 100% connected. This can be risky. Some responses to check are: 14.o11 14.p11 14.p14 15.o13 16.o12 17.p15 18.k11 etc.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.p17 12.m20 13.p21 14.o13 15.p13 is better? 16.q14 17.n12 18.p12 19.o14 20.i11 21.n8
ok, what happen if we dont play 11.p17 for example 11.e17:
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.e17 12.h17 13.h14 14.d21 15.p17
maybe playing 11.p17 now is to early
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.p17 12.m20 13.p21 14.o13 15.p13 16.l4 17.f5 18.n11 19.n14 20.l12
That means 14.o13 is not an easy one for us...
we can also play 11.m20 or somewhere on the left
My suggestion is: 11.i18
12.j20 13.d20 14.e18 15.d15 16.c17 17.g17 18.f20 19.m20
For: 11.i18
12.j20 13.d20 14.e18 15.d15 16.c17 17.g17 18.g19 19.m20
12.m21 13.f15
-- 14.i19 15.k19 16.k20 17.p21
-- 14.j20 15.k17 16.n19 17.p21 (D20 or R12 white threats)
12.g19 13.f15 14.m21 15.k19 16.k22 17.p21
12.p19 13.f15 14.j20 15.m20
For: 11.i18
- it has a good connection to D10 -> G13 or F15
- it has a good connection to M12 -> L15(!) or J15
- it collaborate well with O19 -> M20
- even alone creates a threat to the bottom -> H20
- for now I can't find anything for black ;/ 12.j20 countered by 13.d20
- what else do we want from a single move?
Last but not least: 12.j20 was used earlier in our analysis (10.j20 11.i18) and it seems the same as (10.j20 11.i18 12.h21 which is not too good for black) for now I think 13.d20 really destroys black but of course we could also play something like 13.m20 or 13.g17 14.d21 15.m20 as we wish
For example (just to show one of the top battle possibilities): 13.m20 14.c6 15.b5 16.e7 17.c7 18.d9 19.b9 20.e11 21.e12 22.f13 but in spite of all 12.g15 looks good for white
If we have a battle in the top right at some point, probably not immediately, q8 is stronger than I first thought.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.q8 this is stupid right here 12.o11 13.o13 14.l4 15.q12 16.q6 17.q4 18.r8 19.s9 20.p4 21.o3 22.t9 23.n5 24.p9 25.m7 and the only local answer is 26.m10
Maybe we could use this later.
At this moment, we have 107 hours; that's 4 days, 11 hours. See? A couple of nice quick moves every now and then, and things will be all right. Tomorrow I have the day off from work, so I can be more attentive.
Maceij - if you are having trouble scrolling to the bottom, I posted instructions from Alan at The LG Forum (Twixt vs. The World part II) - you can put a bookmark on your Chrome browser bar that automatically takes you to the bottom in seconds! Try that, if you have not already.
If you guys WANT me to create another fake game for all the new analysis, I can do that; it might get confusing though, since whenever I do that, SOME of the analysis for a particular move will be still over HERE, and SOME will be over there. But let me know, and I'll do it.
I'll say this game is now looking to be a real complicated battle - maybe one for the ages. I'm delighted. I hope everyone else is. (meow)
Against m20 we had k19, against H21 it is i18.
We play on the left to reinforce this part and we threat to win directly this way but we keep a good connection to o19 so that all our pieces seem connected, remember that it was the argument of o19... We played this and not t20 precisely for this reason.
This is the move which makes sense for our game and justify our previous moves.
Of course a move like p17 is interesting but it's like... I think if we play this we have to be 100% it's a quick win or we gonna lose. And I don't think it's a quick win for us.
For these reasons I'm already okay for playing i18, this move is totally obvious to me, when d10 for instance was great but maybe "could" need a preparation.
And now we enter on the TB part, the tactics.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.i18 ...
12. H13 .
Remaining time 132 hours. That's 5 days, 12 hours. mc.
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.i18 12.h13 13.f15:
sth like 14.g15 15.r12 16.p19 17.p17 18.n20 19.h20 20.g19 21.i22 22.l19 23.h16 24.e14 25.e12 26.c15 27.j15
not sure if 13.j15 works: 14.K16 - doesnt look good
maybe 13.j9?
13.j9 14.p14 15.o13 16.q16 17.p17 18.o12 19.l15 20.n15 21.k17
14.l4 15.f5 16.j12 17.k11 18.o13 19.l15 20.p19 21.k17 22.j22 23.n21 24.l19 25.p17 26.n15 27.r11 28.q12 29.q13 30.r10 31.s9 32.r5 33.q8
-16.k16 17.g17 18.o13 19.d14
-16.l21 and? 17.g17?
--18.f14 19.r12 20.p21 21.m20 22.n22 23.k21 24.j22 25.f19
|1.d6 2.g4 3.m12 4.m17 5.r16 6.r20 7.o19 8.l8 9.d10 10.h21 11.i18 12.h13 13.j9 14.l4 15.f5 16.l21 17.g17 18.f14 19.r12 20.p21 21.m20 22.n22 23.k21 24.j22 25.f19 and 26.f22 but white can play 25.d20
-14.p19 15.g17 16.g11 17.d14 18.e10 19.f9 20.c11 21.b11 22.a12 23.r8 24.o17 25.l15
http://twixt-commentator.duckdns.org/game/2142725
OK, I did what was requested: I created a NEW, fake game stub for analysis. The links to the game and analysis page are posted at The Forum ('The World vs. Twixtbot part 3'). Let me try to do it here:
New commentary page: http://twixt-commentator.duckdns.org/game/2142726
as always, the actual game can be found here: http://www.littlegolem.net/jsp/game/game.jsp?gid=2140447
Now, I'm going to try to copy and paste all the recent comments about move 13 from here, over to the new commentary page. However, I'm only grabbing the most recent ones; if there are other lines of value above, you'll have to refer back here.
(meow)