Engine-assisted game vs Stockfish, move 19 discussion
 

The only discussed move on the move 18 discussion thread was ...axb5. Stockfish played 19. Bf3, and the FEN is:
2rq1rk1/6bp/2npb3/1p1Npp2/8/2P2B2/1PN2PPP/R2Q1RK1 b - - 1 19

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

Depth: 56 ply. Nodes searched: 2,187,604,159,550 (~2188 G)

Anybody have something better than the obvious 19...e4

Yes. 19... Kh8 and see what white will do. :smile:

This is why 19...e4 looks good to me. (taking advantage of one of the flaws that most chess programs have.)

19...e4 20.Be2 b4 21.Ncxb 22. Nxb4 d5 23.Ra6 Qd7 24.Nc2 Rb8 25.Qc1 f4 26.Nd4 Bxd4 27.cxd4 Kh8 28.Qc6 Qxc6 29.Rxc6 Bf5 30.Re1 Rxb2

So far, 19...e4 is my best guess, too.

You probably meant 21. Ncxb Nxb4 ?
I could imagine that white plays 21. Ndxb

We have been following the line given by Mr. Barnes in post 5 of the move 16 thread:

17. a4 f5 18. axb5 axb5 19. Bf3 Kh8

Hopefully Mr. Barnes or dh1 (Denny?) is running Komodo now on our current position.

I had missed that we skipped the vote on the last move. Beginning on Friday, I'll have Komodo take a good long look at this move.

This game is now just running too fast for me to keep up with. I have been having a lot of problems getting Arena and Komdo to run properly the last couple of weeks. Now it has quit working completely. I even did a complete uninstall and reinstall with only temporary results. With the holidays I am not sure that I will be able to find the time to help with this move. I will keep trying but can give no guarantees. Sorry.

OK I finally got Arena-Komodo working again. It's definitely going back and forth between 19...e4 and Kh8 as the top moves at about 30 ply. I will let it think for about 24 hours. I will post results on Sunday along with some discussion. With the early posting of the poll I'd like to request that we wait to vote until after more discussion.

Below is Komodo's analysis for the top 3 moves. Thinking lasted about 24 hours and 37 ply.

19...e4 20. Be2 Kh8 21. g3 Qg5 .00
19...Kh8 20. Ra6 e4 21. Be2 f4 .00
19...Qg5 20. Be2 Kh8 21. Kh1 e4 -.07

It reached .00 for the score more quickly for Kh8 than for e4 so it initially thought that Kh8 forces a draw slightly more quickly. But we are not looking for a draw. I think I like 19...e4 somewhat better because of the line that Halsdorf posted because it is more radical. Because it is a little more radical, I will have Komodo look at:

19...e4 20. Be2 b4

What we are looking for is a situation where Komodo continues to improve it's score of the position in our favor as the plies get deeper even if it initially looks like the line is not as good for us. I'll post that analysis sometime late Monday.

[QUOTE=WMHalsdorf;449748]This is why 19...e4 looks good to me. (taking advantage of one of the flaws that most chess programs have.)

19...e4 20.Be2 b4 21.Ncxb 22. Nxb4 d5 23.Ra6 Qd7 24.Nc2 Rb8 25.Qc1 f4 26.Nd4 Bxd4 27.cxd4 Kh8 28.Qc6 Qxc6 29.Rxc6 Bf5 30.Re1 Rxb2[/QUOTE]

After our move 20 in this line, Komodo likes this line for both sides up thru move 24 by white. After that it prefers 24...Kh8 25. Nd4 with an advantage of +0.4 to +0.55 for white. The reason is if we play 24...Rb8 as you state then 25. Nd4 Bf7 26. b4 f4 and white has an advantage of +0.75 to +0.85 with little apparent way for us to improve. The protected passed pawn for white looks ominous. For these reasons I do not think we can play 19...e4 20. Be2 b4.

I still somewhat prefer 19...e4 over 19...Kh8 but in response to 20. Be2 I think we should play 20...Kh8.

Part of the reason that I like 19...e4 now is that white is forced to Be2 right away...not a great place for his bishop. If we play 19...Kh8 then white can play 20. g3 and then if we play 20...e4 then white has a better place for his bishop with 21. Bg2.

Are you using any table bases for your analysis? If not, would you be able to?

[QUOTE=axn;449933]Are you using any table bases for your analysis? If not, would you be able to?[/QUOTE]

I used an opening database through about move 16. Are you aware of any table bases that go this deep in this variation?

No, I meant end-game table bases. K8 supports Syzygy. Using EGTBs will lead to more accurate analysis.

[QUOTE=gd_barnes;449877]This game is now just running too fast for me to keep up with. I have been having a lot of problems getting Arena and Komdo to run properly the last couple of weeks. Now it has quit working completely. I even did a complete uninstall and reinstall with only temporary results. With the holidays I am not sure that I will be able to find the time to help with this move. I will keep trying but can give no guarantees. Sorry.[/QUOTE]
Happy holidays! After voting is done on this move, would everyone like the following system as a replacement to the current 3 days discussion / 3 days vote system?

Sunday morning, ~ 4 AM GMT: Poll is posted with a poll closing time of ~4 AM, Thursday morning GMT
Thursday morning, ~4 AM GMT: Discussion thread is posted with Stockfish's response

If the move is obvious, the poll may be skipped, but the discussion will still be posted on Thursday morning at ~4 AM GMT. It slows the game a bit and hopefully makes the game easier to follow. If I'm out of town on Sunday or Thursday, I'll post the poll/discussion threads early, but I'll give extra time for voting/discussion so that the next poll/discussion post will remain on Sunday/Thursday. Also, Stockfish's response will continue to be posted almost immediately after the poll closes.

I agree with MooMoo's game schedule, in case we play this game to the end. But as it was said before, I don't think we can win anymore, and we (Rybka and me) partially lost the interest.

About the game: Kh8 is sure a draw, we can not improve anything. I played about 12-15 moves against Rybka (I mean both of us playing boths sides at each move, i.e. I was playing/chosing, but with its analysis running in front of me, taking moves back, etc.) which would make about 30-40 moves deep from here (including its analysis' depth) and there is no way for any part to win, unless the other part plays stupid. Rybka still gives a small advantage to white in this position, but that slowly melts as the game progresses.

We can try the other branch (i.e. e4).

Happy holidays to all!

I also like MooMoo's schedule.

[QUOTE=axn;449939]No, I meant end-game table bases. K8 supports Syzygy. Using EGTBs will lead to more accurate analysis.[/QUOTE]

I have a 7-piece end-game analysis site that I have paid for. I guess I'm not familiar with end-game table bases. Can you refer me to Syzygy?

You can refer to komodo 8 readme: [url]https://komodochess.com/store/pages.php?cmsid=14[/url]
Section: Tablebase support

Has links to where you can download Syzygy TBs.