Register
to add
comments.
Game number: Main Page
White: pete99    Black: antony
twixt.ch.25.2.2 (LG) | This game (LG) | Download JTwixt file
On 2010-07-28 at 00:56, pete99* (info) said:
At antony's request, I have prepared an analysis of where black's game went off track after a strong opening. Other comments welcome. Pete99

On 2010-07-28 at 01:10, pete99* (info) said:
In the game as played, white made some simple opening responses while black carefully built a nice pattern, and white was quickly in trouble -- essentially surrounded by black. White played 7.m12 to try to force a connection to the q13 peg. Black's 8.m16 response was difficult for white to respond to, but it left an opening between m13 and m16 for white to work against. White eventually built a line up the right side 17.r10 by continually threatening to connect to the center white links, until the right side line could stand on its own.

On 2010-07-28 at 02:48, pete99* (info) said:
The best alternative move for black I can see would be |8.n15. If |8.n15 9.l14, then 10.k16 11.k17 12.l18 13.j15 14.j17 15.f16 16.g20 for a black win on the left side, and 17.o13 18.r15 on the right. Alternately, if |8.n15 9.l14 then 10.k16 11.k15 12.j14 13.j13 14.m17 15.i15 16.i17 17.e17 18.h15 19.d15 20.e12 21.e13 22.f14 23.d11 24.d10 wins for black on the left.

On 2010-07-28 at 02:50, pete99* (info) said:
The |8.n15 response creates a link that solidifies black's initial 6.p10 structure.

On 2010-07-28 at 23:12, Alan Hensel (info) said:
Antony might have been in good shape as late as move 15: |16.o8, because of the strong 2-5 gap across the top, threatening, for example, 17.o11 18.k8. That's not what would have actually happened, however, because what |16.o8 actually says is that the o8 group is connected to the right side, and as for the left side, there's a double threat (k8 to go across the top, or q11 to go across the bottom). I'm sure Pete would have responded to the double threat with a defensive move on the left, though it is messy enough that I cannot predict what it would have been or whether it would have worked. For example, 17.g12. Then maybe 18.j13 19.h10 20.i8, and there's a big smiley face for Black :-)

On 2010-07-30 at 20:41, antony (info) said:
Thanks for the analysis! Alan's move 16.o8, in particular, is the kind of move I still have much difficulty to see at my level. However, one thing I can't understand is that you seem to consider that the right group is connected to the top after 17.r10, but it seemed to me that twixt.wetpaint.com's page on the Achilles defense said otherwise (the diagram says that r10 is one of the last points that can be blocked with an Achilles defense), so that was what I was planning to do, until I discovered that it was too late. So where did I get it wrong?

On 2010-07-31 at 01:46, Alan Hensel (info) said:
Well, you might have still been able to win after |17.r10, but it's murkier. I didn't mean to imply that move 16 was your last chance; it's just the last one that I see.

The analysis of the Achilles defense on Wetpaint is better for opening and mid-game moves, or, that is to say, moves when large parts of the board are blank. Surrounding pegs change the analysis. For example, starting with the Achilles defense |18.q6, here are some possible future lines:

19.q8 20.o7 21.u6 22.s7 23.v8 24.t9 25.u11 26.u12 27.v13 28.t10 29.u10 - White win on account of R10-R16 gauntlet. (But: 22.r8 23.t9 24.v6 - win for Black.)

19.o7 20.o5 21.q8 22.n7 23.m8 24.l6 25.j9 26.j7 27.i7 28.h6 29.g6 30.f5 31.e5 32.d4 33.d5 34.e6 35.e7 - White wins because of the c3 peg.

19.p9 20.o7 21.n8 22.m8 23.k8 24.n10 25.o11 - White win on the strength of the K10-K14 group.

And there are many other things that could have happened. So, you have to be careful about applying the Achilles defense analysis from Wetpaint to the endgame.

So what do you do if there are interfering pegs? Feel it out with some possible lines. Since you're facing down a combinatorial explosion, it's best to know the common patterns, so you don't waste time considering bad possibilities. Here, after |18.q6, the first thing I'd look at is 19.p9 -- that's probably the strongest response 90% of the time. I'd only look at other possibilities if that one didn't work.

On 2010-07-31 at 02:48, pete99* (info) said:
I would have thought |18.q6 19.o7 would likely be followed by either 20.q8 21.s8 22.s5 23.q7 24.o5 25.p9 26.n7 27.m8 28.l8 29.k8 30.j7 31.i7 32.h6 33.g6 34.f5 35.e5 36.d4 37.d5 38.e6 39.e7 for a white win, or 20.r9 21.s8 22.s5 23.t6 24.u6 for a black win.

After 17.r10, then 18.r5, and black secures the right side connection.

For remembering to consider link-back moves like 16.o8, it may help to keep in mind the principle that before 16.o8, white had a large and wide open playing space from w2 to b9. However, after 16.o8, white was restricted to having two separate and much smaller playing areas (w2 to q10; b2 to n9) to work with -- therefore more difficult to navigate through -- and furthermore, black had an extra advantage within each one by virtue of having the o8-p10 link in place along one edge of the reduced playing area, no matter which one white might choose to attack.

A four-away drop-back move such as 16.p6 might have been equally useful to black in that spot.

On 2010-07-31 at 04:04, David J Bush (info) said:
At first I thought black tossed the game on his last move but after

|20.s7 21.q8 22.q6 23.p6 24.o6 25.t4 26.p4 27.t9 OR

|20.s7 21.q8 22.r5 23.o7 24.n5 25.n6 26.m7 27.m8 OR

|20.s7 21.q8 22.o6 23.s5 white still wins.

So, backing up one more notch, how about

|18.q6 19.q8 20.p4 OR

|18.q6 19.q7 20.r8 21.p9 22.o7 OR

|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5 25.m6 26.p6 OR ... what?

On 2010-07-31 at 04:11, David J Bush (info) said:
|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5 25.p9 26.j5

On 2010-07-31 at 04:15, David J Bush (info) said:
|18.q6 19.o7 20.r9 21.s8 22.s5 23.t6 24.u6 25.q9 26.q8 27.q7 28.o5 29.p9 30.j5 looks about the same.

On 2010-07-31 at 12:56, Alan Hensel (info) said:
You'd still reply to |18.q6 19.o7 with 20.q8 instead of 20.o5, David? I think the latter response is almost always better...

Anyway, don't forget the latent N10 threat. So, in reply to your OR ... what? How about this:
|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5, like you started out, but then 25.n10 26.o10 27.o8 28.j5 29.l5 30.m6 31.j6 32.i7 33.j4, and it looks like a White win to me. Without N10, it looks like it ends well for Black: 25.o8 26.j5 27.l5 28.m6 29.j6 30.i7 31.j4 32.l8 33.k8 34.n9.

On 2010-07-31 at 20:39, David J Bush (info) said:
I didn't see N10. It looks like my J5 idea doesn't work, for example

|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5 25.n10 26.j5 27.l5 28.n7 29.j9 30.k3 31.g6

But how about

|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5 25.n10 26.l6 27.j9 28.h6

On 2010-08-01 at 01:58, Alan Hensel (info) said:
|18.q6 19.o7 20.q8 21.s8 22.s5 23.q7 24.o5 25.n10 26.l6 27.j9 28.h6 and then 29.i7.

If you're looking for a victory for Black after |18.q6 19.o7, how about: 20.o5 21.q8 22.n7 23.m8 24.j5 ?

You must be logged in to add comments.