Engine-assisted game vs Stockfish, move 22 discussion
 

The most popular move on the move 21 poll was ...Be5. Stockfish played 22. f4, and the FEN is:
2rq1r1k/7p/2npb3/1p1Nbp2/4pP2/2P5/1PNQB1PP/R4RK1 b - f3 0 22

Game so far: 1.e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 Nc6 6. Ndb5 d6 7. Bf4 e5 8. Bg5 a6 9. Na3 b5 10. Bxf6 gxf6 11. Nd5 f5 12. c3 Bg7 13. Bd3 Be6 14. Nc2 O-O 15. O-O fxe4 16. Bxe4 Rc8 17. a4 f5 18. axb5 axb5 19. Bf3 e4 20. Be2 Kh8 21. Qd2 Be5 22. f4

Depth: 53 ply. Nodes searched: 1,061,117,102,247 (~1061 G)

I am out of town for 9-10 days and do not have access to Komodo. Since SF responded like Komodo thought he would I would suggest going with the line that Komodo came up with for the last move that I posted then; that is 22...Bxd5. If anyone else has some strong engines that they would like to run on the move I'm certainly in favor of that.

22... Bxd5 looks good for me. After that 23. Qxd5 Qb6+ 24. Kh1 Bg7 seems forced and here white have several different moves.

The other 2 options are not all that appealing that is exf3 and Bg7, so Bxd5 line is the only options.

This looks like a lost game.
Btw, what is the appeal of playing with computer assistence?
Anyway, have fun
Till

[QUOTE=Till;450889]This looks like a lost game.
Btw, what is the appeal of playing with computer assistence?
Anyway, have fun
Till[/QUOTE]

It is likely a forced draw.

[QUOTE=Till;450889]This looks like a lost game.
[/QUOTE]
[QUOTE=gd_barnes;450893]It is likely a forced draw.[/QUOTE]
We drew SF as white without engine assistance, but can't beat SF as black with engine help? This is embarrassing. At what move did this game start to go wrong for us?

Anyway, the best engine to use would probably be the latest version of Houdini, not Komodo. It beat both SF and Komodo in TCEC rapid and beat Komodo to get in to the TCEC superfinal.

[QUOTE=The Carnivore;450899]We drew SF as white without engine assistance, but can't beat SF as black with engine help? This is embarrassing. At what move did this game start to go wrong for us?

Anyway, the best engine to use would probably be the latest version of Houdini, not Komodo. It beat both SF and Komodo in TCEC rapid and beat Komodo to get in to the TCEC superfinal.[/QUOTE]

And just who are you and Till to make such comments? You come in here without hardly having participated and you critize our game. We respectfully ask that you keep your negative comments to yourself. Besides there is nothing embarassing about it. At no move did we go wrong. We never made a move that made our position worse. We just weren't able to improve it enough by about move 20 to have a chance to win. It's extremely difficult to beat any engine while playing as black. We got a little lucky as white that SF inadvertantly started with a somewhat drawish line and we did everything we could to simplify and only play for a draw.

If some people want to put multiple cores together and run a strong engine as black against SF as white we might have a chance but the interest here has not been that great. It's just been 2-3 of us running a couple of cores of Komodo for 24-48 hours per move at times as well as running another engine.

[QUOTE=gd_barnes;450869]I am out of town for 9-10 days and do not have access to Komodo.[/QUOTE]
I'll be out of town from next Friday (Jan. 20) to Monday (Jan 23). If there's some agreement, we can take a halftime break to accommodate both of our schedules and get more analysis in. The game would resume with the move 22 poll on Sunday, Jan 29 instead of on Sunday, Jan 15 (yes, that's in <24 hours, sorry for the late proposal).

If not, the game schedule will continue as normal, and I'll try to fit the poll and threads on/around my out of town dates as best as I can. I'll have some Internet access, but it may be spotty.

[QUOTE=The Carnivore;450899]We drew SF as white without engine assistance, but can't beat SF as black with engine help? This is embarrassing. At what move did this game start to go wrong for us?
[/QUOTE]
[QUOTE=gd_barnes;450923]And just who are you and Till to make such comments? You come in here without hardly having participated and you critize our game. We respectfully ask that you keep your negative comments to yourself. Besides there is nothing embarassing about it. At no move did we go wrong. We never made a move that made our position worse. We just weren't able to improve it enough by about move 20 to have a chance to win.[/QUOTE]
There's nothing embarrassing about the forum's performance. First of all, Stockfish was upgraded from version 6 to version 7 (worth ~50 elo points). Add in increased computing power and thinking time (~10-12 ply deeper, ~30 elo points per ply increase), and you've got a monster that's several hundred elo stronger than what the forum played against during the first half of last year.

Second, Gary is right that the forum didn't make any bad moves. I'll post more after the game, but Stockfish's evaluation didn't have any notable swings in favor of either white or black at any move since the start of the game.

Houdini 5 at 40 ply analysis =

0.00 22. ... Bxd5 23.Qxd5 Qb6+ 24.Kh1 Bg7 25.Qxd6 Rcd8
-0.24 22. ... exf3 23.Bxf3 Rg8 24.g3 Rg7 25.Nf4 Bc4
-0.61 22. ... Bg7 23.Ra6

[QUOTE=gd_barnes;450923]And just who are you and Till to make such comments? You come in here without hardly having participated and you critize our game. We respectfully ask that you keep your negative comments to yourself.[/QUOTE]

I am Till and I think that it's a bit harsh to interprete my post as "negative". All I did was posting my personal point of view about the current game position and a question. Admittedly I do not contribute very often. But is this game not open for anybody, even for casual contributors?

Yes, it is open, and we took your post (and Carnivore's) as a good joke. No harm done. We are still struggling, but we are short of time... We would need an additional queen here... :smile:

Edit: agree with the break (we were very busy at job, lots of work, and very tired at home, in the last days, we have to finish 3 projects before we go to Australia on Feb.15).

[QUOTE=LaurV;450954]Yes, it is open, and we took your post (and Carnivore's) as a good joke.[/QUOTE]

Good to hear. Probably you will demonstrate that it was a joke. I'm looking forward to the result :boxer:

[QUOTE=LaurV;450954]Yes, it is open, and we took your post (and Carnivore's) as a good joke. No harm done. We are still struggling, but we are short of time... We would need an additional queen here... :smile:

Edit: agree with the break (we were very busy at job, lots of work, and very tired at home, in the last days, we have to finish 3 projects before we go to Australia on Feb.15).[/QUOTE]
Thanks for the input! The poll will be posted 14 days from now instead of today.

Now let's get some good analysis in :smile:

Excellent. I'm happy to hear about the break.

[QUOTE=gd_barnes;450923]
If some people want to put multiple cores together and run a strong engine as black against SF as white we might have a chance.[/QUOTE]
I fired up Fire 5. Pun intended.

It's the next strongest engine behind SF, Houdini, and Komodo. It reached 44 ply after a day of thinking. The line it favors is:

22...Bxd5 23. Qxd5 Qb6+ 24. Kh1 Bg7 25. Qxb5 Rb8 26. Qxb6 Rxb6 27. Ra2 Rfb8 Score: 0.00

This line and 0.00 score has remained the same from depth 37-44. Previously, at depths 10-25, Bg7 was leading with a score that favored white by 0.1 pawns.

To summarize, here are the evaluations we have so far:

Houdini 5, 40 ply: 22. ... Bxd5 23.Qxd5 Qb6+ 24.Kh1 Bg7 25.Qxd6 Rcd8. Score: 0.00
Fire 5, 44 ply: 22...Bxd5 23. Qxd5 Qb6+ 24. Kh1 Bg7 25. Qxb5 Rb8 26. Qxb6 Rxb6 27. Ra2 Rfb8 Score: 0.00
Komodo 10, 37 ply: 22...Bxd5 23. Qxd5 Qb6+ 24. Kh1 Bg7 25. Qxb5 Rb8 26. Qxb6 Rxb6 27. Ra2 Rfb8 28. b4 Bxc3 29. b5 Nd4 30. Nxd4 Bxd4. Score: 0.00

I subtracted 2 ply from Komodo's eval since the 39 ply depth was for move 21, not move 22. We may want to consider exf3 or Bg7 if Bxd5 is looking to be a forced draw. I recall reading earlier that we should keep the bishop pair to have a chance of winning, but the pair's gone if we play 22...Bxd5 23. Qxd5.

[QUOTE=Till;450953]I am Till and I think that it's a bit harsh to interprete my post as "negative". All I did was posting my personal point of view about the current game position and a question. Admittedly I do not contribute very often. But is this game not open for anybody, even for casual contributors?[/QUOTE]

You stated that "this looks like a lost game" even those the position is at least a draw for black. That makes no sense given all Engine analysis. How else are we to interpret it other than trying to negatively undermine our performance? I'm all for people contributing frequently or infrequently. Just do not state something blatantly false without backing it up with good reasoning.

[QUOTE=The Carnivore;451204]I subtracted 2 ply from Komodo's eval since the 39 ply depth was for move 21, not move 22. We may want to consider exf3 or Bg7 if Bxd5 is looking to be a forced draw. I recall reading earlier that we should keep the bishop pair to have a chance of winning, but the pair's gone if we play 22...Bxd5 23. Qxd5.[/QUOTE]

I feel like we have to play 22...Bxd5. Playing anything else is risking a loss IMHO. SF7 is quite a bit stronger than SF6 and MooMoo is routinely letting it search to > 50 ply.

[QUOTE=gd_barnes;451496]You stated that "this looks like a lost game" even those the position is at least a draw for black. That makes no sense given all Engine analysis. How else are we to interpret it other than trying to negatively undermine our performance? I'm all for people contributing frequently or infrequently. Just do not state something blatantly false without backing it up with good reasoning.[/QUOTE]

Dear Mister Barnes,
I see, I see. You take this game much more serious than I do, And you are right about the following thing: I ignored all the analysis the community did before. Let me explain: I just entered the site, looked at the chessboard for a couple of minutes, and I thought: This looks very difficult. Then I posted.

What I do not like is that you qualify my comment as a kind of negative behaviour. This is truly not the case. It was an assessment, taken in a short time and maybe inaccurate, but my true opinion then.

I am not a fan of "forced draws", but if you manage to grasp a win (and send me your mail address) I will reward you with my andalusian art-work chess-board. (plus a note about the cause). It is not valuable in terms of money, but means something to me.

Have fun
Till