to add
Game number: Main Page
White: EvetsATL    Black: bob440
This game (LG) | Download JTwixt file
On 2020-09-14 at 03:13, EvetsATL (info) said:
What should I have done differently?

On 2020-09-14 at 10:57, bob440 (info) said:
After my 32.h6

|33.e7 34.f5 35.d4 36.c4 37.b5 38.e3 39.c3 40.c2

On 2020-09-14 at 11:46, Alan Hensel (info) said:
|1.f19 - White chooses a very strong opening, inviting a swap
2.j15 - Black lets White keep the advantage
3.m12 - White stands aside, handing the advantage to Black
4.o17 5.m16 6.l16 7.f15 8.g12 9.q17 10.q16 11.r15 12.r14 13.d10 - Black builds his line, and White is helpless to stop him. |14.e11 or |14.f10 would have made victory easy for White, but |14.c12 has one flaw, and White exploits it: 15.g13. A little misadventure to right side, and then White closes in on the upper left, but misses his |35.e7 36.f5 37.d5 shot and hands victory back to Black.

On 2020-09-14 at 13:03, bob440 (info) said:
yeah, |14.c12 was a horrible blunder; looks like |14.e13 would have pretty much sewn it up

but (particularly in retrospect), |18.e8 looks a bit sketchy, too. I think |18.e6 would have been better

the table shows a 54% edge for |1.f19 -- it's an advantage, sure, but very strong?

On 2020-09-16 at 03:34, EvetsATL (info) said:
Bob, when you say the "table shows a 54% edge" what table are you talking about?

On 2020-09-16 at 03:38, EvetsATL (info) said:
Alan or Bob, what should have been White's second move? Why was 3.m12 such a bad move?

On 2020-09-16 at 11:42, bob440 (info) said:
the table is the Twixt First Move Statistics compiled by Alan Hensel -- see his player info page at

well, Alan might (and likely does) have a different idea, but I think you need something more directly between my 1st move and the right border -- N14, O14 and Q14 all look playable

On 2020-09-16 at 22:42, Alan Hensel (info) said:
I probably shouldn't have commented on moves 1 and 2. It's hard to say anything definitive about the first 2 moves. The statistics were compiled many years ago. I stopped when I became disenchanted. The numbers kept jumping around. I realized that what constitutes a good first move depends a lot on what you can expect for a 2nd move. And what constitutes a good 2nd move depends a lot on what you can expect for a 3rd move, and so on. It's no accident that TwixtBot's idea of what's swappable is very different. TwixtBot doesn't swap F19. It thinks it's weak. But then, TwixtBot's ideas of a good 2nd move are very strange.

Bob has the right 3rd move idea. |3.m12 does nothing to block the J15 peg. It lets Black form 2-5 gap (very strong) across the bottom with |4.o17. You can think of a good blocking distance as starting from the 0-2 gap: |3.L15, for example, and then spots proceeding away from that: |3.m15, |3.n15, but with some preference for moving away from the nearest edge, so |3.n14, but usually within or close to the triangle with a 1/2 slope (so n14, p13) - approximately, of course. With experience you'll be able to judge that better. I don't think I would actually do p13, but m14 might be okay. There may be enough space on the other side, too, with the help of the strong F19 peg: |3.h14 happens to be a 2-5 gap from F19. That looks like something TwixtBot would do.

On 2020-09-18 at 12:15, Peyrol (info) said:
After |35.e7 black could try a tricky move 36.d6 which threatens to double link at F5. But then white can play a tricky move right back with 37.d5 which threatens to connect around black's D6-E8 group with either F9 or C7. This kind of sequence happens frequently. I call it a trickyback sequence, where each player extends their group to threaten to connect around the other group. Actually black is not even threatening to connect E8 to H6 here. White didn't even have to play D5. after 37.h4 white threatens to connect at H8 or at F9 and black cannot block both of those. So this is not the best example of trickyback.

You must be logged in to add comments.