Can you find Fischer's ...19 Bg4 !!

[emerson4301972]

[D] r3r1k1/pp1q1p2/2p2npb/PPPp1bnp/3PpN2/2N1P1PP/1R1B1PBK/3Q1R2 b - - 0 19

This is a game Nikolic E. vs Fischer 1968. Fischer played 19 ...Bg4 !! The move seems to be correct because I used this position to play engine vs engine several times and all engines view white as winning, but after several moves black will be winning.

The move is so deep with no forcing line that it might still be too far from the engines horizon.

I don't have a fast hardware. Maybe those who have one can try to find the move.

[AdminX]

You may be right about the horizon affect. 8 cores may still have a problem with this one. I am testing it right now. Below is the complete game.

[Event "Vinkovci"]
[Site "Vinkovci"]
[Date "1968.??.??"]
[Round "9"]
[White "Nikolic, Emil"]
[Black "Fischer, Robert James"]
[Result "0-1"]
[ECO "E61"]
[PlyCount "62"]
[EventDate "1968.09.??"]
[EventType "tourn"]
[EventRounds "13"]
[EventCountry "YUG"]
[Source "ChessBase"]
[SourceDate "1999.07.01"]

1. c4 g6 2. Nc3 Bg7 3. g3 e5 4. Bg2 d6 5. e3 Nf6 6. Nge2 O-O 7. O-O c6 8. d4
Re8 9. Rb1 e4 10. b4 Bf5 11. h3 h5 12. Nf4 Nbd7 13. a4 Nf8 14. c5 d5 15. b5
N8h7 16. Bd2 Ng5 17. Rb2 Qd7 18. Kh2 Bh6 19. a5 Bg4 20. hxg4 hxg4 21. Rh1 Nf3+
22. Bxf3 gxf3 23. Kg1 Bxf4 24. exf4 Kg7 25. f5 Rh8 26. Bh6+ Rxh6 27. Rxh6 Kxh6
28. Qd2+ g5 29. bxc6 Qxf5 30. Nd1 Qh3 31. Ne3 Kg6 0-1

[AdminX]

[D] r3r1k1/pp1q1p2/2p2npb/PPPp1bnp/3PpN2/2N1P1PP/1R1B1PBK/3Q1R2 b - - 0 19

The move is not clearly winning, and there does not appear to be any move that is clearly winning in this position. I let Rybka 3 analyze the position using IDeA for over 8 hours on my Q6700 and it did not locate the move in that time.



I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4 displayed. After 4189 sample games the result was as is posted below.

[D]r3r1k1/pp1q1p2/2p2npb/PPPp2np/3PpNb1/2N1P1PP/1R1B1PBK/3Q1R2 w - - 0 20

[smirobth]

The move is not clearly winning, and there does not appear to be any move that is clearly winning in this position.

I looked at the position a little bit too, but came to the conclusion that 19.... Bg4 probably is winning.

I let Rybka 3 analyze the position using IDeA for over 8 hours on my Q6700 and it did not locate the move in that time.

I don't think having an engine analyze from the root position has any chance of telling us if Bg4 wins. The lines are just way too deep.

I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4

I would be interested to know if any of the games varried from 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3. If so, was it White or Black who varried? If Black, then I think you might want to run a new Monte Carlo from the position after Black's move 22.

[AdminX]

I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4

I would be interested to know if any of the games varried from 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3. If so, was it White or Black who varried? If Black, then I think you might want to run a new Monte Carlo from the position after Black's move 22.

Hi Robin,

Chess is a battle of ideas, so you have your opinion and I have mine, and thats where the real fun starts! The reason I used Monte Carlo Analysis from the "Bishop G4" move was because I wanted to see the Win / Lose ratio of the move in question since the author stated that Black won in alot of engine versus engine matches that he had played.

I do think that it is a hard move to spot for the human eye and clearly hard for Rybka to spot on my quad , however I don't think it is any better or worst than the other moves for black and I don't think it leads to a forced win. So far the numbers from Monte Carlo seem to back me up here.

[Dr.Ex]

Monte Carlo is rubbish.

[smirobth]

I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4

I would be interested to know if any of the games varried from 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3. If so, was it White or Black who varried? If Black, then I think you might want to run a new Monte Carlo from the position after Black's move 22.

Hi Robin,

Chess is a battle of ideas, so you have your opinion and I have mine, and thats where the real fun starts! The reason I used Monte Carlo Analysis from the "Bishop G4" move was because I wanted to see the Win / Lose ratio of the move in question since the author stated that Black won in alot of engine versus engine matches that he had played.

I do think that it is a hard move to spot for the human eye and clearly hard for Rybka to spot on my quad , however I don't think it is any better or worst than the other moves for black and I don't think it leads to a forced win. So far the numbers from Monte Carlo seem to back me up here.

Hi Ted,
Monte Carlo analysis is a cool idea, and I think it can be very useful in some situations. However in this case if there is only one winning plan for Black (let's say 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3), and if in the Monte Carlo games Black does not follow that plan, then the Monte Carlo analysis will be mostly useless. Which is why I asked if Black or White varied from this line in your Monte Carlo games. Can you tell us please?

[chrisw]

I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4

I would be interested to know if any of the games varried from 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3. If so, was it White or Black who varried? If Black, then I think you might want to run a new Monte Carlo from the position after Black's move 22.

Hi Robin,

Chess is a battle of ideas, so you have your opinion and I have mine, and thats where the real fun starts! The reason I used Monte Carlo Analysis from the "Bishop G4" move was because I wanted to see the Win / Lose ratio of the move in question since the author stated that Black won in alot of engine versus engine matches that he had played.

I do think that it is a hard move to spot for the human eye and clearly hard for Rybka to spot on my quad , however I don't think it is any better or worst than the other moves for black and I don't think it leads to a forced win. So far the numbers from Monte Carlo seem to back me up here.

Hi Ted,
Monte Carlo analysis is a cool idea, and I think it can be very useful in some situations. However in this case if there is only one winning plan for Black (let's say 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3), and if in the Monte Carlo games Black does not follow that plan, then the Monte Carlo analysis will be mostly useless. Which is why I asked if Black or White varied from this line in your Monte Carlo games. Can you tell us please?

That's the problem with Monte Carlo for chess. It is ro can be a very unstabel game. Whether soem insignificant pawn is on b2 or b3 might be the complete turn of the position some moves later.

Chess can be like picking a path through a minefield. One wrong turn and bang.

If no minefield, then Monte Carlo has points in its favour.

Basically you have to decide whether it is sensible to use or not dependent on position. If everything is very quiet and positonal and will remain so, then MC might tell you something. If tactical and wild, that 1% of games that MC tells you white wins, with 99% lost, might be bad info, you should go down the path of the 1% if it exists tactically.

[AdminX]

I then let Monte Carlo Analysis have a go at the position below, which has the key move 19. ... Bg4

I would be interested to know if any of the games varried from 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3. If so, was it White or Black who varried? If Black, then I think you might want to run a new Monte Carlo from the position after Black's move 22.

Hi Robin,

Chess is a battle of ideas, so you have your opinion and I have mine, and thats where the real fun starts! The reason I used Monte Carlo Analysis from the "Bishop G4" move was because I wanted to see the Win / Lose ratio of the move in question since the author stated that Black won in alot of engine versus engine matches that he had played.

I do think that it is a hard move to spot for the human eye and clearly hard for Rybka to spot on my quad , however I don't think it is any better or worst than the other moves for black and I don't think it leads to a forced win. So far the numbers from Monte Carlo seem to back me up here.

Hi Ted,
Monte Carlo analysis is a cool idea, and I think it can be very useful in some situations. However in this case if there is only one winning plan for Black (let's say 19... Bg4 20. hxg4 hxg4 21. Rh1 Nf3+ 22. Bxf3 gxf3), and if in the Monte Carlo games Black does not follow that plan, then the Monte Carlo analysis will be mostly useless. Which is why I asked if Black or White varied from this line in your Monte Carlo games. Can you tell us please?

Ahhh ..., I see your point here Robin.

Thanks

[Kurt Utzinger]

I very much doubt that Fischer's 19...Bg4 is winning the game.
A critical position is the following:

[D]r3r1k1/pp1q1p2/2p2npb/PPPp4/3PpN2/2N1PpP1/1R1B1P1K/3Q3R w - - 0 23

and here White should perhaps have played 21.Nh3 to avoid
Black's Bxf4 allowing to open the h-file with tempo.

Regards
Kurt