New opponent, new handicaps for Komodo.

[lkaufman]

This Saturday (July 1) Komodo will take on a new opponent, Fide Master Andrew Samuelson, in a handicap match. Andy just became an FM this year, so no one can say he's no longer of FM strength. He is 35 years old, near prime age, currently rated 2282 FIDE. I may have played more games against him myself than against any other living player; I had a huge plus score until a couple years ago, but since then he's been winning.
Three games planned, at noon, 3 pm, and 6 pm Eastern Daylight time. As usual, 45 min plus 15 sec each. Broadcast from our Komodochess site. Three different handicaps:
1. Knight (g1 removed) for pawn (f7 removed), Komodo White.
2. Two pawns and move: Komodo plays Black, Andy gets to choose which two Black pawns to remove other than the f7,g7 pair. Komodo thinks f7 and c7 is the best choice. It's probably true since as White the only two pawn handicap that gave Komodo a losing score against GMs was f2,c2.
3. Komodo plays Black with his king starting on e4 and pawn on f6, White to move!! The pawn on f6 makes the position legal. This looks like it should be a simple forced mate, but Komodo doesn't see one. Andy intends to prepare for these games, so I hope he doesn't find a clear win before the game.

We have played somewhat similar handicaps successfully in the past against FMs John Meyer and Larry Gilden, but they were in their 70s and rated about 150 elo lower than Andy. He will be a much tougher opponent.

[JJJ]

I don't see a FM winning game 1 and 3. Game 1 because knight for weakness doesn't seems enough. Also King in middle is well played by computer and not by human. He ll probably miss his chance. So, his best shot is white and two pawn removed. He might draw or win this one. I think he will loose the 2 others game. So I think Komodo will score 2/3 at least.

[lkaufman]

I don't see a FM winning game 1 and 3. Game 1 because knight for weakness doesn't seems enough. Also King in middle is well played by computer and not by human. He ll probably miss his chance. So, his best shot is white and two pawn removed. He might draw or win this one. I think he will loose the 2 others game. So I think Komodo will score 2/3 at least.

The two pawns and move is the largest handicap, as Black gets a bad position on top of the material deficit. The king in the middle handicap lends itself well to preparation, so I won't be surprised if Andy rattles off a dozen moves or so before having to think. This handicap is included more for entertainment than as a genuine test of strength, it's more a test of preparation skill. Knight for pawn may be the hardest handicap for the FM, but if you keep in mind that a knight is worth about four pawns in the opening then you can say that the average handicap of the two material handicaps is 5/2 pawns, with each side having to defend an otherwise somewhat inferior position once. This seems fair enough to me, in view of our past poor results (excluding the chess960 match) trying to give knight odds to much weaker players than this one. But perhaps Komodo has improved enough. We'll see.

[Lyudmil Tsvetkov]

I don't see a FM winning game 1 and 3. Game 1 because knight for weakness doesn't seems enough. Also King in middle is well played by computer and not by human. He ll probably miss his chance. So, his best shot is white and two pawn removed. He might draw or win this one. I think he will loose the 2 others game. So I think Komodo will score 2/3 at least.

The two pawns and move is the largest handicap, as Black gets a bad position on top of the material deficit. The king in the middle handicap lends itself well to preparation, so I won't be surprised if Andy rattles off a dozen moves or so before having to think. This handicap is included more for entertainment than as a genuine test of strength, it's more a test of preparation skill. Knight for pawn may be the hardest handicap for the FM, but if you keep in mind that a knight is worth about four pawns in the opening then you can say that the average handicap of the two material handicaps is 5/2 pawns, with each side having to defend an otherwise somewhat inferior position once. This seems fair enough to me, in view of our past poor results (excluding the chess960 match) trying to give knight odds to much weaker players than this one. But perhaps Komodo has improved enough. We'll see.

[d]rnbqkbnr/ppppp1pp/5p2/8/4K3/8/PPPPPPPP/RNBQ1BNR w kq - 0 1

this one is certainly the hardest for the stronger side, if winnable at all.

of course, at LTC black should win, but at quick shootouts, I have 6 black wins, 1 white win, and 13 draws.

for the other other 2, the record is pretty much 100%.

[Kotlov]

I don't see a FM winning game 1 and 3. Game 1 because knight for weakness doesn't seems enough. Also King in middle is well played by computer and not by human. He ll probably miss his chance. So, his best shot is white and two pawn removed. He might draw or win this one. I think he will loose the 2 others game. So I think Komodo will score 2/3 at least.

The two pawns and move is the largest handicap, as Black gets a bad position on top of the material deficit. The king in the middle handicap lends itself well to preparation, so I won't be surprised if Andy rattles off a dozen moves or so before having to think. This handicap is included more for entertainment than as a genuine test of strength, it's more a test of preparation skill. Knight for pawn may be the hardest handicap for the FM, but if you keep in mind that a knight is worth about four pawns in the opening then you can say that the average handicap of the two material handicaps is 5/2 pawns, with each side having to defend an otherwise somewhat inferior position once. This seems fair enough to me, in view of our past poor results (excluding the chess960 match) trying to give knight odds to much weaker players than this one. But perhaps Komodo has improved enough. We'll see.

[d]rnbqkbnr/ppppp1pp/5p2/8/4K3/8/PPPPPPPP/RNBQ1BNR w kq - 0 1

this one is certainly the hardest for the stronger side, if winnable at all.

of course, at LTC black should win, but at quick shootouts, I have 6 black wins, 1 white win, and 13 draws.

for the other other 2, the record is pretty much 100%.

[d]rnbq1bnr/ppppp1pp/5p2/8/4k3/8/PPPPPPPP/RNBQKBNR w KQ - 0 1

[lkaufman]

Lyudmil, since you had the start position wrong (it's the BLACK king on e4!) does that change your opinion? I can't imagine anyone doubting that White is winning. Only question is how easily?

[JJJ]

Probably not easily. I made Stockfish play against itself yesterday, and it took it around 40/50 move to win and the eval didn't change before a long time. So the human has to be accurate with many move, and so far, I didn't see any human be accurate enough to win when he just has a good starting position without pieces advantages.

[Nordlandia]

Probably not easily. I made Stockfish play against itself yesterday, and it took it around 40/50 move to win and the eval didn't change before a long time. So the human has to be accurate with many move, and so far, I didn't see any human be accurate enough to win when he just has a good starting position without pieces advantages.

Another way to lessen the enormous strength difference is cpu odds.

[JJJ]

I also made Fizbo against Stockfish at 1 min +1 sec, Stockfish managed to draw twice with black.
So I really don't think human stand a chance at this handicap. Human needs material odds to win, not positionnal odds. Even Nakamura lost with positionnal odds after be able to get 3 draw against Komodo.

[lkaufman]

I also made Fizbo against Stockfish at 1 min +1 sec, Stockfish managed to draw twice with black.
So I really don't think human stand a chance at this handicap. Human needs material odds to win, not positionnal odds. Even Nakamura lost with positionnal odds after be able to get 3 draw against Komodo.

One way to look at this handicap is to imagine that White just develops while Black returns his king to e8. Then it would be like a five moves plus castling handicap. But it's probably worse than that for Komodo as White can at least force ...e6 by putting his bishop on c4. I admit it won't be easy for human to win, but I think Komodo will be fighting to draw a poor position. Note that FM Gilden, rated about 170 elo lower, drew with Black king starting on d5 but no ...f6 move, probably a slightly smaller handicap. And he didn't prepare, whereas Samuelson will.