@David: I did not consider the upper parts of the board yet. But we definetly should not play 4.M14 !!! My favoured move is 4.Q14.
On 2011-06-17 at 01:21, upendra said:
@Klaus To make sure I follow, after 1.g42.j153.m144.o15 white is forced to do 5.n16 (or 5.l16) as black's two pegs are connected to the left side and right side, and so white is forced to break down the middle, and in the process gives black the opportunity to do a form of mousetrap?
On 2011-06-17 at 23:20, maaaraca said:
@upendra Yes that's it. It's not an 'original' mousetrap, but if you define a mousetrap as "reducing the territory of the opponent" then it sure is, because it reduces white's southern border to the single peg M14 (there is some room for an attack to go left around J15, but as far as I can see none of them has a chance to win). The only answer as Klaus said is |1.g42.j153.m144.o155.p13 and it is surprisingly strong, I couldn't see any easy win for black. example continuation: 6.n177.s168.r209.r1910.q1611.r1412.s1813.m19 (threatens L16) and white would probably win 14.l1615.q17
To make sure I understand: Why would black stop the ladder when it does in each variation? Is it because it's at that point that if black did continue the ladder, white could opt not to play along and go elsehwere, and should be able to connect to the south edge even if black played again?
On 2011-06-27 at 05:04, David J Bush said:
Well every single variation we looked at here has begun |1.g42.j153.m14
But Klaus said we should not play m14. He labeled the move number 4, probably counting the swap move, but there is no way to make a swap move here, so m14 is move 3. He recommended q14 instead for move 4 by which I believe he meant move 3 here.
So, Klaus, I am willing to go with your suggestion, but perhaps you could show some variations? Also please use the preview button to make sure your posted moves click correctly. Include the whole variation from move 1, and include the vertical bar at the start.
Thank you for your analysis. I agree that this move must be answered and that there are two extreme ways to to so. What I'm wondering about is which is the extreme move which maximize the power on the right side.
Wouldn't be 8.n20 enough ?
By the way, I haven't completely understood your technique to analyse where they/we need a peg in the right side struggle. Could you illustrate it with an example ?
On 2011-08-16 at 06:19, Thierry Pertuy said:
After this night I was wondering if white was eligible to an Achille defense after 9.f1610.h1611.d15 :
First you 'pass', this means playing at a useless spot, in this game I've chosen 8.a2 Then you play out a variation on the right side, for example: 9.s1310.q1911.q1812.r1713.p1614.o1815.l18 Now you see, white can probably win with 16.m1717.k1618.k1319.i1520.g1321.k1222.l1123.g1424.j1025.g10 So you go back to 14.o18, 'pass a move' and look how you have to play out the left side that this doesn't happen, some examples: 15.b116.k1817.h1618.g18 15.b116.k1817.f1618.h1419.g1420.f18 15.b116.i1717.l1618.k18 15.b116.k1717.f1618.g2019.i1920.j19 The idea is to get a feeling how each move is working in different situations. Now you can play out another right side and look how your favoured moves behave in the next situation. The move that does best overall is then the move to play. I often observed that good moves somehow accumulate with this technique (reminds me somehow at Monte Carlo simulation, although it isn't really random). Well so much for theory, I'm often too lazy to do all that, but if I did, the results were usually good and besides that you get a good feeling of the current game.
In the case white would answer 9.u14 then 10.q19 is not ok : 11.s19 It forces black to go very low, opening a path toward the center at the bottom : 10.s2011.q18
So ladder breakers doesn't seem to be such a bad idea in this game. Maybe we should now focus on some possible playouts on the left side. We know now that for the variation above white needs a peg on or behind the diagonal.
On 2011-08-18 at 00:37, maaaraca said:
somehow there is stuff missing in my comment, i know it was there. 2 times the same variation was probably my mistake, sorry, i hope you get it anyway what i'm trying to say. When I said 'it doesn't help much' I was speaking about extending to the southern border for supporting the right side. That's why I thought trying to reduce the ladder could be a good alternative to the above analyzed move K17.
White could have played 33.f14 for further support but they need to play h15 to make it solid. So that we have to answer a way so that we block them and they cannot make a threat with h15.
I wouldn't recommend directly 8.g12 any more, this move gives away the initiative, so it's very hard to predict what happens next. I'm really starting to like i17.
On 2011-08-19 at 14:21, maaaraca said:
the k17 example was not good, it happens to be a special case behind the diagonal for which we still can win, but if white plays like this then not: |1.g42.j153.q144.p115.l126.n157.i138.k179.h1510.g1811.g1712.h2013.e16 threatening to play d20 next, so we'll have to answer.
I agree with the need of a solid link with i13.
I'm trying to summarize : k17 is not enough to prevent white to play a move in a way it threats to connect to the bottom AND it will support the ladder. i17 is enough.
What bothers me in this whole story is that it depends on the reliability of a very long series of moves. You have got much more experience than I do (I'm still a newbie who plays Twixt for only 7 months who has played less than 200 games). I'm confident in the line starting from the Achille, but would I play by myself I would be unsure of the start of the sequence wich drives to the Achille.
Now I would say we could have avoided all this if we would have played |1.g42.j153.q144.q12, afterwards you're always wiser. On a move like L13 we have the strong L10. I think white would really be in trouble with 4.q12, but with 4.p11 we only have the Achilles and not the direct connection threat. Of course I can't prove it, move 4 is still very early in the game, although TwixT is a very fast game.
On 2011-08-20 at 15:35, Thierry Pertuy said:
I understand what you mean about things which tend to happen whatever you play.
It seems that the Achille depends on the f16-h17-j18 link. I haven't got all the automatisms you have, so itr takes me some time to see if f18 is the only possible answer to f16 and if this is possible to avoid this kind of mousetrap.
So far, I agree with your analysis on the importance of th i13 solid link for white defense on the Achille and I agree that i17 seems to be the best answer. I'm curious to see where the game will go now. Because they are likely to try to avoid the Achille.
For example, I suppose that if we answer 10.h16 white will play 11.d15 which I suppose is enough to counter an Achille defense 12.g2013.e20 Then black can only play 14.h19 and we come back to the line driving to the Achille in the upper right corner 15.j1816.j2017.l19
I've made an error in my previous post / last sentence. Please read :
"Then black can only play 12.f20 and we come back to the line driving to the Achille in the upper right corner 13.h1714.h1915.j1816.j2017.l19 "
We could use this board for the team game:
http://www.littlegolem.net/jsp/forum/topic2.jsp?forum=55&topic=255
|1.g4 2.j10 3.n11 and it will be very difficult for black to stop N11 from reaching the top.
|1.g4 2.j15 3.m14 4.m11 5.h12 6.j19 7.q14
Will somebody who isn't me please add a variation?
These 3 variations look good:
|1.g4 2.j15 3.m14 4.m16 5.q15
|1.g4 2.j15 3.m14 4.m18 5.q17
|1.g4 2.j15 3.m14 4.l14 5.o13 6.n15 7.p15
but 4.O15 is difficult: 5.N16- or 5.L16- are bad answers, and 5.P13 is very difficult to play and complex
Why are (3.m14 4.o15) 5.n16 and 5.l16 bad ? Because of 6.m12 ?
|1.g4 2.j15 3.m14 4.o15 5.n16 6.n17 7.m18 8.m19 9.l20 10.l10 or
|1.g4 2.j15 3.m14 4.o15 5.l16 6.k17 7.k18 8.m12
@David: I did not consider the upper parts of the board yet. But we definetly should not play 4.M14 !!! My favoured move is 4.Q14.
The only answer as Klaus said is |1.g4 2.j15 3.m14 4.o15 5.p13 and it is surprisingly strong, I couldn't see any easy win for black. example continuation: 6.n17 7.s16 8.r20 9.r19 10.q16 11.r14 12.s18 13.m19 (threatens L16) and white would probably win 14.l16 15.q17
Thank you for your answer. Why not pushing the ladder one step further ? : |1.g4 2.j15 3.m14 4.o15 5.n16 6.n17 7.m18 8.m19 9.l20 10.l21 11.k22 12.l10
...and for my understanding, what answer to 13.p11 ?
13.p11 14.o7 15.s8
13.p11 14.q8 15.q9
13.p11 14.p9 15.o9
13.p11 14.p7 15.r7 16.r8 17.s9 18.s10 19.r12 20.u11 21.m8 22.o11 23.n10
I stop my commentsof this variation... ;-)
|1.g4 2.j15 3.m14 4.o15 5.n16 6.n17 7.m18 8.m19 9.l20 10.l10 or
|1.g4 2.j15 3.m14 4.o15 5.l16 6.k17 7.k18 8.m12
To make sure I understand: Why would black stop the ladder when it does in each variation? Is it because it's at that point that if black did continue the ladder, white could opt not to play along and go elsehwere, and should be able to connect to the south edge even if black played again?
But Klaus said we should not play m14. He labeled the move number 4, probably counting the swap move, but there is no way to make a swap move here, so m14 is move 3. He recommended q14 instead for move 4 by which I believe he meant move 3 here.
So, Klaus, I am willing to go with your suggestion, but perhaps you could show some variations? Also please use the preview button to make sure your posted moves click correctly. Include the whole variation from move 1, and include the vertical bar at the start.
Perhaps you do not like |1.g4 2.j15 3.m14 4.o15
But what about |1.g4 2.j15 3.q14 4.p10
Klaus showed me the line
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.j19 9.s13 10.q19 11.h16 12.h18 13.l18 14.k17 15.m16 16.m13 17.o15 18.n11 19.r10 20.q6 21.k12
But now it looks like 22.m9 would be good for black.
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.j19 9.s13 10.q19 11.h16 12.h18 13.l18 14.k17 15.r17
So 7.i13 looks okay to me.
Hi Maraca,
Thank you for your analysis. I agree that this move must be answered and that there are two extreme ways to to so. What I'm wondering about is which is the extreme move which maximize the power on the right side.
Wouldn't be 8.n20 enough ?
On the left side :
9.f16 10.h16 11.g14 12.g20
9.h15 10.h19 11.g18 12.g21 13.i19 14.j20
9.h16 10.h19 11.g18 12.f20 13.i19 14.j20
On the right side :
9.s13 10.q18 11.s18 12.u18 13.q17 14.s19 15.p19 16.p20
9.u14 10.s18
By the way, I haven't completely understood your technique to analyse where they/we need a peg in the right side struggle. Could you illustrate it with an example ?
12.g20 13.e20 14.e19 15.d18 16.d17 17.c16 18.g14 19.e13 20.d13
but white can pass below and introduce a weakness in black's defense : 12.g20 13.d20 14.c20 15.f19 16.f18 17.g21 18.d17 19.i20 20.j18
...but there may be better defenses. 12.f21 ?
Here's the promised analysis technique example:
First you 'pass', this means playing at a useless spot, in this game I've chosen 8.a2
Then you play out a variation on the right side, for example: 9.s13 10.q19 11.q18 12.r17 13.p16 14.o18 15.l18
Now you see, white can probably win with 16.m17 17.k16 18.k13 19.i15 20.g13 21.k12 22.l11 23.g14 24.j10 25.g10
So you go back to 14.o18, 'pass a move' and look how you have to play out the left side that this doesn't happen, some examples:
15.b1 16.k18 17.h16 18.g18
15.b1 16.k18 17.f16 18.h14 19.g14 20.f18
15.b1 16.i17 17.l16 18.k18
15.b1 16.k17 17.f16 18.g20 19.i19 20.j19
The idea is to get a feeling how each move is working in different situations. Now you can play out another right side and look how your favoured moves behave in the next situation. The move that does best overall is then the move to play. I often observed that good moves somehow accumulate with this technique (reminds me somehow at Monte Carlo simulation, although it isn't really random). Well so much for theory, I'm often too lazy to do all that, but if I did, the results were usually good and besides that you get a good feeling of the current game.
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17 9.f16 10.g12 11.g14 12.j11 13.k14 14.l10 or 14.m8
In the case white would answer 9.u14 then 10.q19 is not ok : 11.s19
It forces black to go very low, opening a path toward the center at the bottom : 10.s20 11.q18
...or to try to use the influence of the center pegs such as 10.q19 11.s19 12.s18 13.t17 14.t16 15.v16 16.r12
...but I'm not sure it works 17.s13 18.o13 19.t11 and this is not very comfortable.
Any other ideas on how to answer to 9.u14 ?
I like 8.k17 and also 8.n20, but yes it seems like opening a new adjacent path south.
@Thierry
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17
Maybe after 9.u14 10.q18 with eg. 11.r17 12.r16 13.t17 14.q13 15.o13 16.s14
@Jeff K.
If 8.k17 9.u14 10.q18 then I think that white would probably answer 11.s18 and I cannot see a way to block them :
12.u18 13.q17
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17 9.u14 10.q19 11.s19 12.s18 13.t17 14.t16 15.v16 16.t12 17.s13 18.o13
Now T12 prevents white from playing S9, S8 or R10. Possible continuations:
19.r11 Achilles now doesn't work, so we gave to go 1 step further: 20.q9 21.s9 22.r5 23.q8 24.p6 25.o9 26.m8 27.m10 28.i8 29.e8 30.f11 31.f10 32.h10
If white wants to win this variation they probably had to play a solid link before, because we've got I12.
So either 33.h15 which could have been played out like this before: 33.h15 34.g18 35.g17 36.h20 37.f19 38.f21 39.d20 40.d22 41.b21 42.b23 and we would have right enough space to win, it's extremely close.
Or this could work too 33.h15 34.g18 35.g17 36.h20 37.f19 38.f21 39.d20 40.d22 41.b21 42.b23 because I think now they can't play D15, because we don't have to answer it.
But this would lose 33.h15 34.g17 35.g18 36.h19 37.f16 38.f20 39.e19 40.d21 41.c20 42.c2 on the other hand we could maybe answer something else instead of Q19, because we've got more support now.
The other solid answer for white: 33.g14 34.f17 35.f16 36.g19 37.e18 38.f21 39.c19 40.d22
Can't come up with a reasonable line now, but it looks like it doesn't help too much.
So ladder breakers doesn't seem to be such a bad idea in this game. Maybe we should now focus on some possible playouts on the left side. We know now that for the variation above white needs a peg on or behind the diagonal.
Yes, I think that I get what you mean except when you say that we would lose with this line : 33.h15 34.g17 35.g18 36.h19 37.f16 38.f20 39.e19 40.d21 41.c20
...maybe the missing stuff ?
I don't see any flaw in your line.
White could have played 33.f14 for further support but they need to play h15 to make it solid. So that we have to answer a way so that we block them and they cannot make a threat with h15.
For clarification, here is it in the correct order :
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17
Good for white : 9.f14 10.f19 11.h15 because of h19 threat : 12.c2 13.h19 14.j19 15.f18 and then black must answer : 12.h18 for example, then follows your line with 13.u14 14.q19 15.s19 16.s18 17.t17 18.t16 19.v16 20.t12 21.s13 22.o13 23.r11 24.q9 25.s9 26.r5 27.q8 28.p6 29.o9 30.m8 31.m10 32.i8 33.e8 34.f11 35.f10
Good for black : 9.f14 10.g17 because we don't need to answer to 11.h15 12.c2 13.g18 14.h19 15.e17 16.f20 17.d19 18.d21
I cannot see any other white move which both threats to connect on the bottom left and provide support enough to f10 in your line.
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17 9.u14 10.q19 11.s19 12.s18 13.t17 14.t16 15.v16 16.t12 17.s13 18.o13 19.r11 20.q9 21.s9 22.r5 23.q8 24.p6 25.o9 26.m8 27.m10 28.i8 29.e8 30.f11 31.f10 32.h10 33.h15 34.g17 35.g18 36.h19 37.f16 38.f20 39.e19 40.d21 41.c20 42.b20 43.e12 44.e13 45.d14 46.d15 47.c16 48.c17 49.b18
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17 9.u14 10.q19 11.s19 12.s18 13.t17 14.t16 15.v16 16.t12 17.s13 18.o13 19.r11 20.q9 21.s9 22.r5 23.q8 24.p6 25.o9 26.m8 27.m10 28.i8 29.e8 30.f11 31.f10 32.h10 33.h15 34.g17 35.g18 36.h19 37.f16 38.f20 39.e19 40.d21 41.c20 42.b20 43.e12 44.e13 45.d14 46.d15 47.c16 48.c17 49.b18
Some left side playouts with solid link, if they make a gap the ladder doesn't support enough, so I didn't look at those lines:
8.i17 9.h15 10.f20 11.g17 12.g18 13.e18 14.d21
8.i17 9.g14 10.f20
8.i17 9.k14 10.o17 11.l16 12.l20
On the other hand K17 allows white to go behind the diagonal. only one example, I think it's possible in any case:
8.k17 9.h15 10.g18 11.g17 12.h20 13.e18 14.d20 15.d21 16.e22 17.c19 18.c21
I wouldn't recommend directly 8.g12 any more, this move gives away the initiative, so it's very hard to predict what happens next.
I'm really starting to like i17.
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.k17 9.h15 10.g18 11.g17 12.h20 13.e16 threatening to play d20 next, so we'll have to answer.
I agree with the need of a solid link with i13.
I'm trying to summarize : k17 is not enough to prevent white to play a move in a way it threats to connect to the bottom AND it will support the ladder. i17 is enough.
What bothers me in this whole story is that it depends on the reliability of a very long series of moves. You have got much more experience than I do (I'm still a newbie who plays Twixt for only 7 months who has played less than 200 games). I'm confident in the line starting from the Achille, but would I play by myself I would be unsure of the start of the sequence wich drives to the Achille.
What about 11.o18 for example? 12.p16 13.r17 14.s18 15.t18
I still like n20, and I don't have enough experience to feel it as a "don't do it". Actually, I feel it as "do it"...
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.n20 9.f16 10.f18 11.h17 12.h19 13.j18 14.j20 15.l19 16.l21 17.d17 18.d19 19.b18 20.b20 21.p19 22.o18 23.o17 24.n16 25.s13 26.p17 27.s18 28.r16 29.t16 30.o13 31.r11 32.q9 33.s9 34.r5 35.q8 36.p6 37.o9 38.m8 39.m10 40.i8 41.g8 42.h10 43.f10 44.i12 45.k14 46.h14 47.j16 48.d12 49.d11 50.e14 51.c13 52.c15 53.b15 54.a16
This time the ladder was more behind the diagonal, but we could use the fact, that I13 didn't have a solid connection... this could also be different.
|1.g4 2.j15 3.q14 4.p11 5.l12 6.n15 7.i13 8.i17 9.u14 10.q19 11.o18 12.p16 13.r17 14.s17 It's not pretty, but I think this would be the correct move to win for black in your example.
It seems that the Achille depends on the f16-h17-j18 link. I haven't got all the automatisms you have, so itr takes me some time to see if f18 is the only possible answer to f16 and if this is possible to avoid this kind of mousetrap.
So far, I agree with your analysis on the importance of th i13 solid link for white defense on the Achille and I agree that i17 seems to be the best answer. I'm curious to see where the game will go now. Because they are likely to try to avoid the Achille.
For example, I suppose that if we answer 10.h16 white will play 11.d15 which I suppose is enough to counter an Achille defense 12.g20 13.e20
Then black can only play 14.h19 and we come back to the line driving to the Achille in the upper right corner 15.j18 16.j20 17.l19
I've made an error in my previous post / last sentence. Please read :
"Then black can only play 12.f20 and we come back to the line driving to the Achille in the upper right corner 13.h17 14.h19 15.j18 16.j20 17.l19 "
Actually 12.f20 is not enough 13.d20...