how will Leela fare at the WCCC?

[Werewolf]

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.

Mark

Are you suggesting there was a possibility of Leela playing that ending so badly Komodo might have won it? Seriously?

[IanO]

Humans should not be allowed to execute moves on the board, and hereby affect the outcome of the game. Let the engines play on their own, and if it is draw, it is draw...

Well that's how WCCC's are being played. It's human plus the engine.
Remember the 2003 Shredder-Jonny event? The good gesture of Johannes gave Shredder the WCCC(along with Junior or Fritz).
If Johannes was the opponent of Leela today for example, he would gladly accept the draw offer i guess, so standings would be totally different. And it's a 7 rounds tournament anyway so every half point counts.

Thank you for reminding us of that bit of history: Shredder had some sort of bug, allowing a three-fold repetition during a winning attack. If Johannes -- participating with Jonny in the WCCC for the first time -- had rightfully claimed his draw in the final round, then Fritz would have been undisputed champion that year instead of losing the playoff.

Personally, this Leela-Komodo result leaves a sour taste in my mouth, unfitting for Don Dailey's legacy.

On the other hand, this is what tournament competition is all about. I can't count the number of games decided due to stupid bugs as well as operator error. "Welcome to the big leagues, kid."

[yanquis1972]

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.

Mark

Are you suggesting there was a possibility of Leela playing that ending so badly Komodo might have won it? Seriously?

yes, but why do you doubt his honesty? unless this is in retrospect decisive in the championship, it's very difficult to assign any rational significance to it beyond a learning experience for a new entrant. and if it should ultimately have that significance, is leela going to die because it wasn't the 2018 WCCC??

[Werewolf]

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.

Mark

Are you suggesting there was a possibility of Leela playing that ending so badly Komodo might have won it? Seriously?

yes, but why do you doubt his honesty? unless this is in retrospect decisive in the championship, it's very difficult to assign any rational significance to it beyond a learning experience for a new entrant. and if it should ultimately have that significance, is leela going to die because it wasn't the 2018 WCCC??

Well if you don't care about the results of the WCCC that's up to you - don't follow it. Some of us are interested.

His statement implies a chance of Komodo winning. The board position implies none whatsoever. That's the contrast.

The operator would have known there was no serious winning chance and a draw was the fairest result. If I thought Komodo could have even a 1% chance of winning it would be worth playing on - but not in that position! Even a 1400 elo human could draw that for Leela!

[yanquis1972]

why should 'not caring about the results' (which isn't actually true this year; i have a rooting interest in leela) mean i shouldn't follow the games??

you say the board position implied no chance of a win, but that isn't how leela interpreted it. it makes sense to see if leela can draw against perfect play. this is the same thing an impartial operator would do out of sheer curiosity.

[Werewolf]

You're muddling things up. Leela never interpreted the position as having any chance of KOMODO winning. Obviously we're not discussing whether LEELA could win or not - that's not the point. It wasn't the Leela team who turned a draw down.

[Robert Pope]

Do we know for sure that Leela could have drawn that game, and wouldn't have blundered it away?

[Werewolf]

Do we know for sure that Leela could have drawn that game, and wouldn't have blundered it away?

I'm not trying to turn this into some massive argument but surely this can't be possible.

Leela played one of the strongest chess entities on planet Earth: Komodo on a 60 core high-end workstation. It played 133 moves of such good chess it finished an exchange up against it. Alas, this wasn't enough to win.

To blunder it all away would require worse chess than any computer I've ever owned, even from the 1980's. It just wouldn't happen. Surely.

[jorose]

Do we know for sure that Leela could have drawn that game, and wouldn't have blundered it away?

There is no way in hell Leela loses that outside of somehow getting disqualified or losing on time and it seems disingenuous to even suggest it. In a human game I would think white has some limited chances though that is obviously not the case if black has tablebases. I don't think there can be a question, the Komodo team did not think they were ever going to win the game based on something on the board. Whether or not that is somehow morally reprehensible or not is another question, but based on the board position they are either idiots or they were hoping to win some other way and I'd argue its clearly the latter.

[Nay Lin Tun]

Oh man,the first time I see those photos where people are manually entering moves. no IT professionals to set up two computers via LAN? I can even do that!!