how many points will Komodo score vs. GM Lenderman?

[Chessqueen]

[quote=Chessqueen post_id=844357 time=1590171123 user_id=10732]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that with the position closed, if it remains closed against Komodo by GM Lenderman, he can guarantee a draw.

[lkaufman]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that opening the position against Komodo wad a mistake by GM Lenderman, since with the position closed at least he can guarantee a draw.

No one else was playing, I could find nothing wrong. It forfeited on time at the end which was strange, but perhaps it has to do with MCTS seeing it is getting mated, which hasn't happened until this game. I made some small adjustments for the current game. I think he just got an easy position to play in the first game today.

[Chessqueen]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that opening the position against Komodo wad a mistake by GM Lenderman, since with the position closed at least he can guarantee a draw.

No one else was playing, I could find nothing wrong. It forfeited on time at the end which was strange, but perhaps it has to do with MCTS seeing it is getting mated, which hasn't happened until this game. I made some small adjustments for the current game. I think he just got an easy position to play in the first game today.

Knight odds without Lenderman removing one of his pawns or Knight odds in chess960 in certain positions could be horrible for an engine against Lenderman it is my believe, but according to BrendanJNorman only player stronger than Lenderman can see this, or BrendanJNorman believes that he is a Super GM making all kinds of videos on Youtube. I just wish that BrendanJNorman can show us how he can win vs Komodo 14 with contemp set up to 140 with a knight odds plus a pawn removed using the MCTS version, but again I am just a patzer. I only wish that he can set up Zoom.us. com so I can play a few Rapid games to see who is the better patzer.

This is what he wrote to me " Sorry, but when I see patzers authoritatively talking about what a strong GM "didn't understand", it irks me somewhat.

Especially the hyperbole about "Giant Anaconda Snake from the Amazon..."

Look...If Lenderman has no idea what he did wrong, you definitely don't "

WOW how much chess knowledge this guy has, he is Amazing. lets take NOTES all of US and learn from Him. Definitively he can talk a lot, and kept talking about ratings so on and so on, that is all that he is concerned about

[Chessqueen]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that opening the position against Komodo wad a mistake by GM Lenderman, since with the position closed at least he can guarantee a draw.

No one else was playing, I could find nothing wrong. It forfeited on time at the end which was strange, but perhaps it has to do with MCTS seeing it is getting mated, which hasn't happened until this game. I made some small adjustments for the current game. I think he just got an easy position to play in the first game today.

Out of all the people who predicted how many points Komodo 14 will score, I came the closest in predicting even if Komodo 14 draws the last game. Now I have something to celebrate . My prediction was 8 to 8, but nobody came closer than me, and Komodo will get at least 7.0 points, but if it started with the MCTS version as I suggested from the beginning it would had score at least 8.5 points

[Chessqueen]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that opening the position against Komodo wad a mistake by GM Lenderman, since with the position closed at least he can guarantee a draw.

No one else was playing, I could find nothing wrong. It forfeited on time at the end which was strange, but perhaps it has to do with MCTS seeing it is getting mated, which hasn't happened until this game. I made some small adjustments for the current game. I think he just got an easy position to play in the first game today.

Out of all the people who predicted how many points Komodo 14 will score, I came the closest in predicting even if Komodo 14 draws the last game. Now I have something to celebrate . My prediction was 8 to 8, but nobody came closer than me, and Komodo will get at least 7.0 points, but if it started with the MCTS version as I suggested from the beginning it would had score at least 8.5 points

Well now we know, that the MCTS version is the best with odds, and I do NOT believe that Stockfish come even close to give GM like Lenderman Knight odds in Chess960 or Knight for pawn and do as great as Komodo 14 did.

PS: Since Nakamura was the Aegon Chess960 Champion over 11 years ago I believe,The Aegon at Mainz Chess Classic: Nakamura won 960 World Championship. 7/31/2009 – On the first day of the Chess960 World Rapid Chess Championship in Mainz Levon Aronian had dominated, winning all three of his games. On the second day US GM Hikaru Nakamura had won all of his to get into the finals. probably Komodo 14 can give Knight odds and Nakamura removing one of his pawn with a 10'+5" time control. But of course GM Wesly SO is the Current Chess960 champion who beat Carlsen last year, and So deserves to play versus Komodo very soon

[lkaufman]

PS: Something completely different. This Question is for Mr. Larry Kaufman, what is happening on the fist game, is Komodo sharing the same 32 cores system again by playing simultaneously with GM Nakamura and therefore, lowering your 32 cores with Komodo 14 Strength by at least 50 Elo?

I might be a patzer and we all are unless GM Carlsen or Caruana are here on this Forum, but I believe that opening the position against Komodo wad a mistake by GM Lenderman, since with the position closed at least he can guarantee a draw.

No one else was playing, I could find nothing wrong. It forfeited on time at the end which was strange, but perhaps it has to do with MCTS seeing it is getting mated, which hasn't happened until this game. I made some small adjustments for the current game. I think he just got an easy position to play in the first game today.

Out of all the people who predicted how many points Komodo 14 will score, I came the closest in predicting even if Komodo 14 draws the last game. Now I have something to celebrate . My prediction was 8 to 8, but nobody came closer than me, and Komodo will get at least 7.0 points, but if it started with the MCTS version as I suggested from the beginning it would had score at least 8.5 points

Well now we know, that the MCTS version is the best with odds, and I do NOT believe that Stockfish come even close to give GM like Lenderman Knight odds in Chess960 or Knight for pawn and do as great as Komodo 14 did.

PS: Since Nakamura was the Aegon Chess960 Champion over 11 years ago I believe,The Aegon at Mainz Chess Classic: Nakamura won 960 World Championship. 7/31/2009 – On the first day of the Chess960 World Rapid Chess Championship in Mainz Levon Aronian had dominated, winning all three of his games. On the second day US GM Hikaru Nakamura had won all of his to get into the finals. probably Komodo 14 can give Knight odds and Nakamura removing one of his pawn with a 10'+5" time control. But of course GM Wesly SO is the Current Chess960 champion who beat Carlsen last year, and So deserves to play versus Komodo very soon

So although Komodo lost 7 to 9 overall, the MCTS version did get a 6.5 to 5.5 victory on its own. Regarding chess960, it isn't very suitable for knight for pawn odds, because too much depends on the choice of pawn in the precise position. Knight odds Armageddon, meaning the human must win as Black with a White knight removed, makes more sense, but probably Nakamura or Wesley So would be a bit too strong for any engine at this handicap right now at 15' + 10". It might be fair at a faster tc as you suggest. Note that the Lenderman games, though "FRC", were all with kings and rooks on normal squares so we could still call it normal chess (normal castling rules); true FRC might be a bit more favorable for the engine.

[Cornfed]

"Well now we know, that the MCTS version is the best with odds"

But, why?

[lkaufman]

"Well now we know, that the MCTS version is the best with odds"

But, why?

It is better because it does not assume "perfect" play by the opponent, but chooses its move based on what it thinks will score best against good but imperfect play, just like Alpha Zero and Lc0 do, but without the Neural Network. Standard Komodo may avoid moves for reasons that no human would ever imagine.

[duncan]

I pick 10.5 to 11.
Lenderman will not be able to handle the pressure.

If by "pressure" you mean "time pressure", then you might be right. I don't think he'll have any problem with psychological pressure due to the show or anything emotional, he's a well-balanced professional and very rational and realistic. He won't let a loss affect him in the next game. But no human is immune to the pressure of the clock.

I noticed that, a few minutes ago. How much time handicap can Komodo 14 actually give MVL or Nakamura on your 32 cores? LIke 15 Minutes to 15 Seconds for Komodo for the entire game? Today that it is a rest day Komodo 14 on your system should be challenging GM on Chess.com with time handicap and see who will dare to accept

PS: This was a horribly one sided game in favour of Komodo against Jorge Sammour http://talkchess.com/forum3/viewtopic.php?f=2&t=73976

Chess.com isn't set up for time odds now. But anyone, GM or patzer, can play Komodo level 24 anytime and that is roughly the same as playing Komodo on one core at game in three seconds! Even under those conditions, I think Naka would need serious time, like 15' + 10", to be competitive. Most GMs even had a tough time with what is now level 23, which is much weaker than 24, a full class I would say.

Can you or your son play a time handicap against your Komodo 14 at a time that you believe that you could at least draw and post it here, such as 30'+ 5"' and to Komodo 15 Seconds? You can use Arena to set up different time by right clicking the time and either raise it up for you and lower down to Komodo. Once you Righ click on the time under arena, it will turn RED and it will allow you to adjust different times to both sides.

I don't think even Carlsen or Naka can break even with Komodo at those time odds, I certainly have no chance. I'm pretty even playing rapid with Komodo at about level 20 or so, which is way faster than game in one second.

Poor Lenderman, he does not know what he did wrong or how Komodo outplayed him, he kept saying I did NOT blunder, what he does NOT understand is that Komodo squeezed him little by little like a Giant Anaconda Snake from the Amazon and the calculating power of the 32 cores plus MCTS choke his position and made most of his pieces useless on the back rank

Komodo also won the third game today, making the match score even, totally recovering from the 3 game deficit before we swtiched to MCTS. I guess there's no more doubt as to which version is best for such matches.

If Mcts is handicapped, will it get a better score against Komodo than if it was not ?
Also can the matches be replayed with the commentary ?

If you are asking whether Komodo MCTS would score more points starting from a losing position against normal Komodo than it would from the normal one, of course not, it will lose almost every game. But if you want to know whether Komodo MCTS would score more points than Komodo standard if each is giving a handicap to a weaker version of Komodo, that is a good question that I have tested. So far all my tests show that standard Komodo does better in this scenario. It seems that a weakened Komodo is not a good simulation of a human grandmaster, even if the elo is similar. Maybe using a crippled NN would work better for this, I haven't done much testing on that yet. I don't know how to replay twitch chess.com event shows from previous days, if no one else knows I can ask.

I meant the latter. It was wrongly worded. Thanks for that.

[duncan]

"Well now we know, that the MCTS version is the best with odds"

But, why?

It is better because it does not assume "perfect" play by the opponent, but chooses its move based on what it thinks will score best against good but imperfect play, just like Alpha Zero and Lc0 do, but without the Neural Network. Standard Komodo may avoid moves for reasons that no human would ever imagine.

How does the MCTS version decide what move is most likely to cause human problems, even though it is not objectively the best?

Also does MCTS version score better than regular version against a human in non handicapped chess and if the advantage of MCTS version against humans is bigger in handicapped chess than regular chess why is this ?