Next Komodo match on chess.com

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.

How will Komodo score against GM Lenderman?

Poll ended at Wed Nov 08, 2017 5:21 am

0 or .5 out of 4
0
No votes
1 out of 4.
1
10%
1.5 out of 4.
0
No votes
2 out of 4.
2
20%
2.5 out of 4.
4
40%
3 out of 4.
2
20%
3.5 or 4 out of 4.
1
10%
 
Total votes: 10

lkaufman
Posts: 3120
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: Next Komodo match on chess.com

Post by lkaufman » Tue Nov 14, 2017 3:19 pm

Uri Blass wrote:
lkaufman wrote:
Nordlandia wrote:
Kappatoo wrote:Is there a site where Komodo's handicap games are collected? I think it would be great to have all the games and results in one place.
Look here ->

https://komodochess.com/store/pages.php?cmsid=17

Larry: any observation/prediction on how much odds increase per year based on engine development. Every new year chess engines get stronger and that mean more odds can be allocated human opponent to offset strength difference.

Considering top engines increase in playing strength each year, human players need considerable odds to have practical chances against strong engines.

Something similar to Moore's law for odds given to human opponents as engines advances :)
The performance of Komodo in handicap games seems to be improving more than Kai predicted. We lost our very first handicap match badly at two pawn (c2,f2) handicap, but in the past year we have scored very highly against higher rated GMs at larger (based on evals) two pawn handicaps. We made a change before the latest match that might be very important in lost positions like handicap games although not significant in normal games. Contempt also makes a big difference (we didn't have it yet at the time of our first match). What I'm saying is that the improvements that matter in these handicap matches are not the same as the ones that matter against near-equal opponents in normal chess. If we doubled our speed humans might hardly be able to tell in handicap games, but Stockfish sure would see the difference in normal games!
I wonder if you test your changes in handicap matches against weaker engines to see if you get significantly better result against them.
That is a good idea, but ideas for even games have higher priority for testing now.
Komodo rules!

leavenfish
Posts: 219
Joined: Mon Sep 02, 2013 6:23 am

Re: Next Komodo match on chess.com

Post by leavenfish » Tue Nov 14, 2017 5:02 pm

Larry - Not so much these handicap matches....but does maybe something like Fischerandom in any way give ideas as to how Komodo might be improved or where weaknesses are?

lkaufman
Posts: 3120
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: Next Komodo match on chess.com

Post by lkaufman » Tue Nov 14, 2017 7:42 pm

leavenfish wrote:Larry - Not so much these handicap matches....but does maybe something like Fischerandom in any way give ideas as to how Komodo might be improved or where weaknesses are?
Fischerandom might possibly show up some problems that don't show normally, but for the most part we don't need to find problems, we need to find solutions that actually help the rating. That's the hard part.
Komodo rules!

Post Reply