how will Leela fare at the WCCC?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
frankp
Posts: 216
Joined: Sun Mar 12, 2006 2:11 pm

Re: how will Leela fare at the WCCC?

Post by frankp » Mon Jul 16, 2018 5:04 pm

jdart wrote:
Mon Jul 16, 2018 4:59 pm
Loss on time happens in human tournaments all the time. Even GMs have been known to lose on time. It is part of the game.

--Jon
Yes.
And if the computer lost on time, fine.
But clearly, particularly for closely matched engines, the operator has a significant input.
Jonny is still playing at 160 moves. I think whoever dies first loses. LOL

Werewolf
Posts: 1192
Joined: Thu Sep 18, 2008 8:24 pm

Re: how will Leela fare at the WCCC?

Post by Werewolf » Mon Jul 16, 2018 5:11 pm

frankp wrote:
Mon Jul 16, 2018 5:04 pm
jdart wrote:
Mon Jul 16, 2018 4:59 pm
Loss on time happens in human tournaments all the time. Even GMs have been known to lose on time. It is part of the game.

--Jon
Yes.
And if the computer lost on time, fine.
Exactly. It's the drawn ending plus the other human using a cheapo to get a point which makes it unjust.

JJJ
Posts: 1285
Joined: Sat Apr 19, 2014 11:47 am

Re: how will Leela fare at the WCCC?

Post by JJJ » Mon Jul 16, 2018 5:15 pm

noobpwnftw wrote:
Mon Jul 16, 2018 4:54 pm
JJJ wrote:
Mon Jul 16, 2018 3:39 pm
I m very surprised the Komodo team refused a draw. To me it is very unfair. Leela is not the favorite here, and even if it was, it would have been unfair.

Now if Komodo win, who's gonna be happy for the team ? Few people and most won't. Game is a draw. First time Leela play, and they refuse draw because the human ( not the program after all , he can insta move ) is low on time. Not fair ! Even if it is not against the rule.
The thing is that the rules were clearly stated and not biased against any participants. In fact the very reason for the high expectations for Leela is it gives more freedom on the hardware.

It is interesting that engines don't fight over a cable but teams meet in person and play on a real chessboard, which is part of the spirit.
There can be human errors like misplacing a piece or even unable to participate due to various reasons, which is also a part of the drama.

If people can't accept the fact that shit happened and not in their favor, then it's no better than withdrawing into a moral victory.


EDIT: Wrong quote, sorry. See below.

It is your opinion. But most people are gonna see this as a stolen win. Maybe Mark or Larry will ask for a draw, because I don't see them tring to win the WCCC by doing that.

User avatar
Laskos
Posts: 9417
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: how will Leela fare at the WCCC?

Post by Laskos » Mon Jul 16, 2018 5:16 pm

Guenther wrote:
Mon Jul 16, 2018 11:59 am
George Tsavdaris wrote:
Mon Jul 16, 2018 11:53 am


It was this position, a dead draw one:
8/8/8/KR6/p3k3/P7/2b5/8 w - - 0 1

Leela's operator was having problems with time and he was running out of time, so he offered a draw 2 moves before the end i think, but Komodo's operator didn't accept(!! really very bad Chess sportsmanship) and he couldn't keep up with executing the moves so he run out of time so 0-1.

So this loss for Leela was because her operator wasn't good enough by playing the moves on a "dead draw" position. This is not Chess of course. They should have accepted the draw in this kind of dead position but it's ok since it's inside the rules of course.
Thanks for confirming what I guessed in my post, which I was writing at the same time, George.

Even if it is covered by the rules you won't make friends, if you refuse a draw after >130 moves and being on
the weaker side
of a dead drawn 6men endgame! Very bad sportsmanship...
We can always keep our own scoresheet. The only thing to consider is that maybe playoffs, if they will be needed, will be played by different engines than our scoresheets show. Two remarks as of now: many draws in general and Leela draws too against strong engines on strong multi-CPUs. I guess Komodo was running on 48-core machine, as in previous editions, right? It seems the opponents haven't considered anti-Leela books. Even I could have prepared one in 2-3 days, as Leela is abysmal on tactical puzzles.

noobpwnftw
Posts: 356
Joined: Sun Nov 08, 2015 10:10 pm

Re: how will Leela fare at the WCCC?

Post by noobpwnftw » Mon Jul 16, 2018 5:22 pm

JJJ wrote:
Mon Jul 16, 2018 5:15 pm
It is your opinion. But most people are gonna see this as a stolen win. Maybe Mark or Larry will ask for a draw, because I don't see them tring to win the WCCC by doing that.
In your views, if the quite opposite happened, are "most people" willing to give up the win instead? I don't think so, because if they have every right to keep the win, why would they think it is "stolen"?

User avatar
George Tsavdaris
Posts: 1575
Joined: Thu Mar 09, 2006 11:35 am

Re: how will Leela fare at the WCCC?

Post by George Tsavdaris » Mon Jul 16, 2018 5:30 pm

Laskos wrote:
Mon Jul 16, 2018 5:16 pm
Two remarks as of now: many draws in general and Leela draws too against strong engines on strong multi-CPUs. I guess Komodo was running on 48-core machine, as in previous editions, right? It seems the opponents haven't considered anti-Leela books. Even I could have prepared one in 2-3 days, as Leela is abysmal on tactical puzzles.
From ICGA:
Komodo uses a 60 core server (58 cores in use) Intel Xeon E7-8890 v2.2 4x15 cores @ 2.8 GHz
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....

lkaufman
Posts: 3685
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: how will Leela fare at the WCCC?

Post by lkaufman » Mon Jul 16, 2018 6:03 pm

Albert Silver wrote:
Mon Jul 16, 2018 1:56 pm
frankp wrote:
Mon Jul 16, 2018 12:19 pm
Allowing the possibility of 'gamesmanship' in a computer tournament is a little perplexing.
Relies on 'good' behaviour', I guess. When money/commercialism is involved, this is probably asking too much of some people.
More than a little perplexing. Leela and operator lost on time through their own setup. There was no power outage, no dizzy spell by the operator, in other words, no mitigating circumstances other than failing to be prepared. The clock is a part of the game. Take ownership of mistakes made, and don't blame the opponent for benefiting from them.
It seems that Leela's operator tried to get by with just five seconds operator time (by setting the increment to just five seconds less than the stated one), which was clearly too small given all the circumstances. If he had set it to a more reasonable ten seconds operator time, it would have been a different game. Personally I would prefer not to have operator time count, but I don't make the rules. Erdo, our operator, feels obliged to do his best for Komodo within the rules, just as others have done in the past. In fact in this very tournament, in the Blitz, Komodo disconnected and lost a couple minutes of clock time resetting. No one offered to restore the time, and Komodo was left with just one minute, but still we got a draw in the end.
Komodo rules!

mjlef
Posts: 1424
Joined: Thu Mar 30, 2006 12:08 pm
Contact:

Re: how will Leela fare at the WCCC?

Post by mjlef » Mon Jul 16, 2018 6:09 pm

[quote="Albert Silver" post_id=768489 time=1531749419 user_id=143]
[quote=frankp post_id=768481 time=1531743556 user_id=586]
Allowing the possibility of 'gamesmanship' in a computer tournament is a little perplexing.
Relies on 'good' behaviour', I guess. When money/commercialism is involved, this is probably asking too much of some people.
[/quote]

More than a little perplexing. Leela and operator lost on time through their own setup. There was no power outage, no dizzy spell by the operator, in other words, no mitigating circumstances other than failing to be prepared. The clock is a part of the game. Take ownership of mistakes made, and don't blame the opponent for benefiting from them.
[/quote]

I posted a reply earlier but it seems gone now. So I will try to repeat it. Some more info below.

The lc0 operator was using some simplified GUI they wrote. Well, it was not graphical but it has things like "wp" for a white pawn (very much what I did when I started chess programming). I told him the writer of the UCI spec was in the tournament and they should write a UCI version and try out some of the many good chess GUIs (this was during the game).
There was no opening book for lc0, so it started thinking on the first move.

I spoke with him after the game and suggested he reduce his increment. Someone else had already suggested this. He had set his increment to, I believe, 10 seconds per move for the first game. For the second game he had already set it to 5 seconds, because 10 seconds (with the actual 15 sec increment in the time control) only allows 5 seconds per move. I also believe moves were entered as e2e4<Return> in the unique GUI, which takes more time than a single or double click in a standard GUI. These things all took up more time. Note he did not seem upset by the time loss.

You might ask why we played on once Komodo reached the 6 piece Syzygy draw. Our experience with MTCS nn engines is they do not play the endgame as well as traditional programs. We also discussed Syzygy with the lc0 operator and he said they had not implement them yet. If they had them, we would have proposed a draw ourselves long before since why play on when we know the result? When you know little about a new opponent, the safest thing to do is play on. If it was say Shredder, then game would have ended a lot sooner.

About two moves before the end of the game game I reminded the lc0 operator that he could alwys request a draw. I was not looking at the clock when I said this. He proposed it, and I went over to our operator to start discussing if we should accept. The clock fell and it lost on time.

Earlier in the game I saw lco was taking a lot less time than Komodo, so I did not keep track of its time. And I did not know he had such a large increment which only allowed 5 seconds (on average) to enter a move. I am also not sure if he tried adjusting the time. Operating a chess engine in a tournament srequire a lot of scanning of everything to make sure all is good.

I spoke with Larry later in the day and he agreed with what we had done. I think nearly all chess programmers learn a lot from the first tournament. I sure did. Preparation and practice is important. I will try to follow the time better and suggest people adjust time if it is getting off.

Mark

yanquis1972
Posts: 1762
Joined: Tue Jun 02, 2009 10:14 pm

Re: how will Leela fare at the WCCC?

Post by yanquis1972 » Mon Jul 16, 2018 6:16 pm

i didn't realize that (according to a previous post) if an operator plays the wrong move, it counts (!?); as opposed to being correctable or equaling a loss.

did any team ever offer kasparov or anand the opportunity to serve as 'operator' back in the day?? and, if the operator is allowed to input the 'incorrect' move into the engine, even any IM/GM well-schooled in engine play, back in the day?

oh crap, from the history section on icga:

Moron 0.01 (Author: "fern"; Operated by Weinstein, G.): 1994 World Computer Chess Champion

frankp
Posts: 216
Joined: Sun Mar 12, 2006 2:11 pm

Re: how will Leela fare at the WCCC?

Post by frankp » Mon Jul 16, 2018 6:18 pm

mjlef wrote:
Mon Jul 16, 2018 6:09 pm
Albert Silver wrote:
Mon Jul 16, 2018 1:56 pm
frankp wrote:
Mon Jul 16, 2018 12:19 pm
Allowing the possibility of 'gamesmanship' in a computer tournament is a little perplexing.
Relies on 'good' behaviour', I guess. When money/commercialism is involved, this is probably asking too much of some people.
More than a little perplexing. Leela and operator lost on time through their own setup. There was no power outage, no dizzy spell by the operator, in other words, no mitigating circumstances other than failing to be prepared. The clock is a part of the game. Take ownership of mistakes made, and don't blame the opponent for benefiting from them.
I posted a reply earlier but it seems gone now. So I will try to repeat it. Some more info below.

The lc0 operator was using some simplified GUI they wrote. Well, it was not graphical but it has things like "wp" for a white pawn (very much what I did when I started chess programming). I told him the writer of the UCI spec was in the tournament and they should write a UCI version and try out some of the many good chess GUIs (this was during the game).
There was no opening book for lc0, so it started thinking on the first move.

I spoke with him after the game and suggested he reduce his increment. Someone else had already suggested this. He had set his increment to, I believe, 10 seconds per move for the first game. For the second game he had already set it to 5 seconds, because 10 seconds (with the actual 15 sec increment in the time control) only allows 5 seconds per move. I also believe moves were entered as e2e4<Return> in the unique GUI, which takes more time than a single or double click in a standard GUI. These things all took up more time. Note he did not seem upset by the time loss.

You might ask why we played on once Komodo reached the 6 piece Syzygy draw. Our experience with MTCS nn engines is they do not play the endgame as well as traditional programs. We also discussed Syzygy with the lc0 operator and he said they had not implement them yet. If they had them, we would have proposed a draw ourselves long before since why play on when we know the result? When you know little about a new opponent, the safest thing to do is play on. If it was say Shredder, then game would have ended a lot sooner.

About two moves before the end of the game game I reminded the lc0 operator that he could alwys request a draw. I was not looking at the clock when I said this. He proposed it, and I went over to our operator to start discussing if we should accept. The clock fell and it lost on time.

Earlier in the game I saw lco was taking a lot less time than Komodo, so I did not keep track of its time. And I did not know he had such a large increment which only allowed 5 seconds (on average) to enter a move. I am also not sure if he tried adjusting the time. Operating a chess engine in a tournament srequire a lot of scanning of everything to make sure all is good.

I spoke with Larry later in the day and he agreed with what we had done. I think nearly all chess programmers learn a lot from the first tournament. I sure did. Preparation and practice is important. I will try to follow the time better and suggest people adjust time if it is getting off.

Mark
You had bishop and wrong colour rook pawn.
Leela had rook and pawn. Pawns blocked on same file. IIRC


Do not misunderstand, I am most definitely not suggesting you broke the rules or did not have the right to play on.
Equally, the primary failing is definitely by the leela team.
But the game had been drawn for some time.

Post Reply