mersenneforum.org  

Go Back   mersenneforum.org > Extra Stuff > Hobbies > Chess > Other Chess Games

Reply
 
Thread Tools
Old 2016-10-28, 16:45   #1
MooMoo2
 
MooMoo2's Avatar
 
Aug 2010

659 Posts
Default Engine-assisted game vs Stockfish, move 10 discussion

The most popular move on the move 9 poll was ...b5. Stockfish played 10. Bxf6, and the FEN is:
r1bqkb1r/5ppp/p1np1B2/1p2p3/4P3/N1N5/PPP2PPP/R2QKB1R b KQkq - 0 10

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

Depth: 51 ply. Nodes searched: 2,954,025,368,677 (~2954 G)
Attached Thumbnails
Click image for larger version

Name:	position.png
Views:	83
Size:	14.5 KB
ID:	15074  
MooMoo2 is offline   Reply With Quote
Old 2016-10-28, 17:07   #2
firejuggler
 
firejuggler's Avatar
 
Apr 2010
Over the rainbow

A5816 Posts
Default

As I said earlier, I'm a poor chess player.
My very first reaction would be pxf6. or even qxf6.
Or you can "protect" the queen with either be7 or ne7.

Or move the queen away to a5, effectivelly locking the white knight.
but it is seen as a grossly bad move.

Last fiddled with by firejuggler on 2016-10-28 at 17:36
firejuggler is offline   Reply With Quote
Old 2016-10-28, 18:11   #3
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
Jun 2011
Thailand

263016 Posts
Default

We can't run away with the queen or block, in that case we just lost a knight in vain, we will most probably lose the pawn too, or exchange the bishops, and the game is lost in any case, no need SF for that, from such an advantage Moomoo can win without SF.

We must take the bishop now! And taking it with the queen is a weak move due to the sacrifice Bxb5. So, after 10...Qxf6 11.Nd5 we must go back, ...Qd8 (no other place, due to the check in c7 and we lose the rook, and waste the castling, very bad for us) then as said, 12. Bxb5 axb5 13. Nxb5 then we should consider ourselves extremely lucky if we get a draw from here in spite of the fact that we play with two bishops against a horse and two pawns, but our queen's flank is empty, white has 3 free pawns there, and I see two white queens on the board in the far future there... We may continue like ...Qa5+ 14.c3 then we can not take the horse because we lose the queen with Nc7+, therefore 14...Ra7 15. Nxa7, again bad for us. Note that we can not save the rook, and Rb8 we lose the queen after 15 b4.

So, the only reasonable move is 10...gxf6 (edit: to which he will still reply 11. Nd5, so we can think to black move 11 already)

Last fiddled with by LaurV on 2016-10-28 at 18:33
LaurV is offline   Reply With Quote
Old 2016-10-28, 23:33   #4
WMHalsdorf
 
WMHalsdorf's Avatar
 
Feb 2005
Bristol, CT

33×19 Posts
Default

Quote:
Originally Posted by LaurV View Post
So, the only reasonable move is 10...gxf6 (edit: to which he will still reply 11. Nd5, so we can think to black move 11 already)
For move 11 the only moves I would consider are Be6, Bg7,Rb8 and f5
WMHalsdorf is offline   Reply With Quote
Old 2016-10-29, 07:38   #5
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

1052010 Posts
Default

10...gxf6 is the only reasonable move.

Komodo 10 says that 11. Nd5 is +0.3 pawns better than any other move by white. The opening databases that I'm looking at show that white plays this move over 99% of the time. So I think we can safely assume that is what SF will play.

In response for our move 11, both the opening DB's and Komodo seem to agree that 11...Bg7 and f5 are quite a bit better than anything else. These two moves are played a combined nearly 95% of the time. 11...f5 is played almost twice as frequently but both moves seem to have a similar winning and drawing percentage in the DB.

I will focus Komodo 10 on our move 11 for the next couple of days with multi-PV set to 2. That way it will look most extensively at only the top 2 moves. I suspect that it will determine that the two moves will be very close in strength so it will be up to us to determine what we like the best.
gd_barnes is offline   Reply With Quote
Old 2016-11-01, 04:48   #6
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

244308 Posts
Default

I let Komodo 10 run on our 11th move for about 30 hours. It reached 42 ply. In response to 10...gxf6 11. Nd5 it's suggestions are:

11...Bg7 -0.33
11...f5 -0.34

Based on the games showed in the DB, I also like 11...Bg7. We can discuss it more in a future 11th move thread.

Last fiddled with by gd_barnes on 2016-11-01 at 04:50
gd_barnes is offline   Reply With Quote
Old 2016-11-01, 06:18   #7
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
Jun 2011
Thailand

24·13·47 Posts
Default

You spoil all the fun

Advice: do not post all the details of your analysis. Especially scores for moves. Try to give it your pick first, then let the engine do the heavy job, but don't post like that.
We know you use this engine, as everybody know I use rybka, the moves can later be replayed if somebody wants to analyze something, learn, or for fun. But if you post the first in the tread (not the case here, this is just an example) for every move and say komodo have this and this scores after hundred of hours and cores, who will argue against it?

How can one bet against the strongest engine in the world, having ran for 30 hours

So, now if I have a different opinion, I have to keep quiet...
Kidding you. (Partially.... Or not? hehe...)* Me and Rybka were slowly going toward 11...f5, too. But now I can't argue anymore...

----------
* Raman-style white font...
LaurV is offline   Reply With Quote
Old 2016-11-01, 09:31   #8
axn
 
axn's Avatar
 
Jun 2003

19×271 Posts
Default

Quote:
Originally Posted by LaurV View Post
Me and Rybka were slowly going toward 11...f5, too. But now I can't argue anymore...
Sure you can. What is the PV for 11. ... Bg7 vs 11. ... f5 ? Does the PV look sane? What if you don't agree with some moves in the PV (especially ones that are deeper in the tree)?

There are times when the machine misses a proper continuation because of aggressive pruning. This is where a (experienced) human can contribute and get the machine to analyze line it would not have done on its own.
axn is offline   Reply With Quote
Old 2016-11-01, 16:06   #9
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
Jun 2011
Thailand

24·13·47 Posts
Default

So, you still want scores... Hm... Ok, I let rybka 20 ply, it says +0.35 for f5 and +0.37 for Bg7. Which means in positives, so it considers white being better here (I don't! But I am an optimist ). So, in its conception f5 is better. Here we agree, mainly because its best reply to Bg7 is c4, and I don't like how the game evolves after, that is why I would go for f5, not Bg7. But I can't much argue against komodo, which is much better! I still try to understand the game and bring my own "intelligence" into it, beside the engine, and not for that 0.02 difference.

Edit: Rybka also has a database that extends behind of the current position, where Bg7 was played about 5000 times and comes with a score of +0.48, and f5 was played about 8000 times and it comes with a (worse) score of +0.52. In both cases the ELO of the players who played the move (I don't know how that is computed, it seems too high to be an average) is 2845. The database is from 2010. Other moves that come with better score in this DB, but were played only few times, with "ELOs" around 2000, are Bd7 and Bh6, with scores around +0.45, but I think these moves are bad. Interesting that Rb8, which I still consider a very good move here, comes very low in the DB (+0.85)

On my side, I am very happy to see komodo going into negative scores, in Gary's post, this gives me a lot more hope! But I still consider we should not post scores. It kills all the fun. Moomoo didn't post scores in the other game. Imagine seeing at every move how your position deteriorates, grrr,... hahaha

Last fiddled with by LaurV on 2016-11-01 at 16:25
LaurV is offline   Reply With Quote
Old 2016-11-01, 16:50   #10
unconnected
 
unconnected's Avatar
 
May 2009
Russia, Moscow

7×379 Posts
Default

I think it is no need to run power computers in this well-known position for hours or even days. Both f5 and Bg7 are good moves and there is matter of taste what to prefer. I vote for f5, we can get more dynamic positions with it.
unconnected is offline   Reply With Quote
Old 2016-11-01, 23:54   #11
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

23×5×263 Posts
Default

Sorry to spoil the party. I felt like it was a good idea to run the top-rated program on a position where the top moves were fairly even to see what it came up with. Clearly it agrees that the two moves are effectively equal.
gd_barnes is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Engine-assisted game vs Stockfish, move 36 discussion. Drawn game? MooMoo2 Other Chess Games 10 2017-04-29 17:39
Engine-assisted game vs Stockfish, move 25 discussion MooMoo2 Other Chess Games 6 2017-02-14 00:30
Engine-assisted game vs Stockfish, move 20 discussion MooMoo2 Other Chess Games 6 2017-01-02 21:49
Engine-assisted game vs Stockfish, move 16 discussion MooMoo2 Other Chess Games 4 2016-12-10 16:50
Engine-assisted game vs Stockfish, move 15 discussion MooMoo2 Other Chess Games 5 2016-12-06 05:27

All times are UTC. The time now is 11:33.


Tue Oct 19 11:33:41 UTC 2021 up 88 days, 6:02, 0 users, load averages: 0.89, 1.17, 1.25

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.