Supranormal Acitivity in Chess

This is the second of three wins against Houdini 3 at game in 60 in the Two Knights' Defense:

Dear Lev, Keith, Alex and Mark:

The questions being addressed are: 1) “Can Houdini 3 be beaten if it plays the Two Knights’ Defense?” 2) “Are all variations of the Two Knights’ Defense better for White?” What I hope to demonstrate is that the Two Knights’ Defense is so good for White that even a super computer can’t equalize against best play by White. One of the practical results I have discovered is that in the middle game, if the evaluation ever rises above +1.2, then it is a win for White.

 What I intend to do is to test every major variation of the Two Knights’ Defense under tournament conditions. I will test under what many players would consider valid i.e. no sight of either the moves or evaluation of the computer as it moves under supervision by a tournament director. Due to time limitations I will play the games at a time limit of 2 hours for the entire game with a 10 second bonus for each move; Houdini will not be allowed to lose on time

This is  the second win in the Ulvestad: This is an unsupervised game but shows the problems facing Black:

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 b5?! (The Ulvestad)

6.Bf1 h6? (Perfectly acceptable according to MCO 15; according to theory Black gets a good game with an early Bc5; as demonstrated here, h6 is a very weak move. Black should play 6…Nd4 and transpose to the Fritz as Berliner did in his classic game against Estrin for the Correspondence World Championship)

7.Nxf7! Kxf7

8.dxc6 a6?!

9.a4 Bc5

10.Be2 Ne4

11.O-O Nxf2

12.d4 Bxd4

13.Rxf2 Bxf2ch

14.Kxf2 Qxd1

15.Bxd1 Bf5

16.Be3 (Houdini prefers Bd2; this is more active) Be4

17.Nc3 Bxc6

18.Be2 (I don’t see Houdini’s point of Bh5ch; I intend to control all invasion routes by the Rooks to the 1^st, 2^nd, and 3^rd) Rhf8

19.axb5 Kg8ch

20.Kg1 axb5

21.Rxa8 Rxa8

22.Nxb5 (According to my theories White should not even consider Bxb5. The reason is that Nxb5 gains a tempo over Bxb5 i.e. the Bishop can access b5 in one move, thus the move sequence Be2/Bxb5 does not gain time. Since the Knight cannot access b5 in under two moves, Nxb5 is a developing move. While it might not seem so, this is still a time/tempo attack even though it is an endgame.) Rb8?!

23.c4! (I welcome an endgame of two Bishops and pawns versus Rook and pawns. Again, according to my theories c4 gains time) Bxb5

24.cxb5 (Black has a dismal defensive task here) Ra8

25.Bd2 (I regard this as a simple technical win. After Bc3 all the White pawns are defended and it is just a matter of time before White creates a passed pawn on the Queenside) Kf7

26.Kf2 Ke6

27.Ke3 Ra4

28.Bc3 (According to my theories this is still a time/tempo attack i.e. all White’s moves are developing moves) Kd5

29.Bf3ch (Not to beat a dead horse, but this is a developing move gaining time with an attack on the King. Up to this point I have only used a few seconds/move because I am just taking a quick look at tactical shots 1-2 moves ahead and just developing as quickly as possible. It is only after this last move do I start actually “crunching” the position; the previous moves were just following my theories. I know I am winning but don’t have a clue how to win it so I just decide to make a few harmless moves until a plan appears.) Kd6

30.Be4 Ke6

31.h3 Kd6

32.g3 Ra1

33.Bc6 Ke6

34.Ke4 Rd1

35.Bb7 g6

36.Bc8ch Kd6

37.Bg4 Rb1

38.Bxe5ch Kc5

39.Be2 c6? (Houdini was out to a depth of 28 and could see nothing better; now I have a plan---create passed pawns on both wings)

40.bxc6 Kxc6

41.h4 Kd7

42.Bc4 Re1ch

43.Kf4 h5

44.Bg7 Ke8

45.Bf6 Rg1

46.Be5 Re1

47.Bb8 Kf8

48.Bd6ch Kg7

49.b4 Rd1

50.Be5ch Kf8

51.Kg5 Ke7

52.Kxg6 Re1

53.Rf6ch 1-0

White followed a “cook book” recipe out to move 28 and then just had to win a won game:

bump

Keep up the practice. Maybe time to play at a tougher level.

I hate your guts for bumping this.  Oh God why did I still have this tracked for all this time?

Just stumbled on to this thread, and am not going to read the whole thing (would if I had time  ), but I think it's off track on principle to look for "surpranormal" or supernormal causes unless there is actual proof.

On a tangential subject, people like ot point out examples of Fischer glancing at positions and demonstrating the winning line, as if he just "knew" everything, while in fact as great as he was (as someone said early in the thread, "Good grief" - he was a hundred points above the opposition) he had to work as hard at the board as did his fellow GMs.

And despite amazing winning streaks by him and others, saying they were world champions or similar doesn't explain it because they were world champs or world class GMs when they had less spectaculor results too.

This has been an interesting thread. I read it all. I haven't made up my mind about the OP though.

I tried to read this thread from beginning to end. And hearing all thes amazing stories just ........    I have yet to have anything close to these guys in chess. Going to kids chess area and stand next to them and tell them my stories which will sound great to them and make me feel better.

Was the key to eat your vegetables when your a kid and only play chess?

This may not constitute supranormal activity, but it is consistent with my theory of communal blind spots. In the Wilkes-Barre there is almost no coverage either in the literature or in practical terms (In my 2013 data base, I couldn't find even one game played by a strong player, but this variation may be best) of an "obvious" choice. Williams' in "The Real American Wilkes Barre Variation the Two Knights' Defense" devotes only one line of analysis to it.

Estrin in "Wilkes-Barre Variation" devotes only one paragraph and gives a typically rosy assessment for Black. Where Estrin claims, "Black has the initiative for the sacrificed pawn." This is nonsense and may have steered players away from a winning position for White. Here is analysis from Houdini 3:

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 Bc5 5.Bxf7ch Ke7 6.Bc4 d6 7.Nc3 Qf8 (To force White to castle) 8.O-O h6 9.Nf3 g5 10.Nb5 +-

The 5.Nxf7 Bxf2ch lines are equal. 5.d4 is about equal. Here is the detailed analysis of the 6.Bc4 variation. To the best of my knowledge this entire variation has been missing from main stream play and theory. Using Intuitive Iteration, it was possible to find this variation:

One of the surprising findings I have made with Houdini 3 is that what appeared to be a clear cook of the Wilkes-Barre Traxler turned out not to be so clear. While White can get in the slow plan Bxf7ch/Bb3/d3/Nf3/Be3/fxe3/Nbd2/Qe2/O-O-O Black has counterplay. What happens typically is that Black can force h3; then Black plays g5/g4, exchanges off White's h-pawn, then follows with h5/h4/h3 to exchange off White's g-pawn. Now with all the pawns on the Queenside, White doesn't have enough to achieve a clear plus.

What is surprising is that White has a much sounder choice with 6.Bc4. In view of the fact that Houdini's advantage never dropped below +/-, this seems best. The key difference is the ability of White to play Be2 to deny Black the exchange sacrifice on f3 or his ability to liquidate the Kingside. Do you see any way for Black to equalize?

1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. Ng5 Bc5 5. Bxf7+ Ke7 6. Bc4 Rf8 7. Nc3 Na5 8.d3 d6 9. O-O Nxc4 10. dxc4 h6 11. Nf3 Be6 12. Qd3 Kf7 13. b3 Kg8 14. Be3 Bxe3 15. Qxe3 Nh5 16. Ne1 Nf4 17. Nd3 Qg5 18. Qg3 a6 19. Qxg5 hxg5 20. Nd5 Ne2+ 21.Kh1 Rf7 22. f3 a5 23. Rf2 Nf4 24. N5xf4 gxf4 25. c5 b6 26. cxb6 cxb6 27. Rd1 b5 28. Rfd2 Ra6 29. Nc1 Rc7 30. g3

2. 6. Bc4 Qe8 7. Nc3 Nd4 8.Na4 Qh5 9. d3 Qxd1+ 10. Kxd1 Bd6 11. Nc3 c6 12. Ne2 Nxe2 13. Kxe2 Bc7 14. Be3d5 15. exd5 cxd5 16. Bb3 Bg4+ 17. f3 Bh5 18. Nh3 a5 19. a4 Bg6 20. Nf2 Kd7 21.Ng4 Nxg4 22. fxg4

3.  6. Bc4 Qe8 7. Nc3 h6 8.Nf3 d6 9. O-O Kd8 10. Be2 a6 11. d3 Be6 12. Be3 Bd4 13. Bxd4 Nxd4 14. Qd2 c515. Nxd4 cxd4 16. Nd5 Bxd5 17. exd5 Nxd5 18. Qa5+ Nc7 19. c4 Rb8 20. Qb4 Qc621. f4 e4 22. Qe1 Re8 23. Qh4+ Re7 24. dxe4 Qxe4 25. Bf3 Qf5 26. Rae1 Qf6 27.Qxf6 gxf6 28. Rxe7 Kxe7 29. Re1+ Kd7 30. Rd1

4. 6. Bc4 Qe8 7. Nc3 d6 8.Nd5+ Nxd5 9. exd5 Qg6 10. dxc6 Qxg5 11. O-O Qh4 12.Bd5 Bg4 13. Qe1 bxc6 14.Bxc6 Rab8 *

We now enter the main line where White missed a half dozen wins in the late middlegame in Game

Moody -Houdini 3 

I had a computer specialist build the computer to the specifications recommended by the manufacturer. According to a GM I contacted Houdini is playing at its suggested strength. Unfortunately, I couldn't get the game editor  to work. 

5.Bxf7ch Ke7 6.Bc4! (There is almost no analysis of this variation and no master games where this was played except for one crush by White reported by Estrin in Wilkes-Barre Variation Two Knights Defense. Williams, The Real American Wilkes-Barre, offers almost no analysis; MCO 13, 14, or 15, offer no analysis, nor does BCO 2 ---White anticipates an exchange sacrifice on f3 which can be neutralized if White can play Be2) Na5?! (One way to meet the threat of Nf7, the point being if Nf7 Qe8 8.Nxh8 Nxc4 and the Knight is dead---Qe8 is equivalent, +/- 6...Rf8! +/=) 7.Bd3!! White defends e4 in anticipation of h6 8.Nf3 Nxe4; true, White has an advantage after Nxe4 but this seems stronger taking away any counterplay) h6 8.Nf3 d6 9.Nc3 Be6 10.Be2 (I have been able to draw with Na4, but decided to try something different; rather than play tactically against the computer, I try quiet, positional play) Nc6 (Demonstrating that Na5 is a bad idea) 11.d3 a6 12.Na4 (To free the c-pawn) Ba7 13.O-O (I considered b3 with the idea of Nb2) Kf7 14.h3 (Luft and taking away g4) Rf8 15.b3 (Anticipating an isolated pawn on e5 after Black plays d5) Kg8 16.Be3 Bd4 17.c3 (Houdini doesn't like this move; the point is that Nb2 defends the d3 square. White can try c4 but has fewer winning chances but fewer losing chances) Bxe3 18.fxe3 b5 19.Nb2 d5 20.exd5 Nxd5 21.Qc1 (White holds c3 and e3) Qd6 (White is better) 22.Nd1! (This Knight is needed on the Kingside. This defends e3 and c3 and prepares Nf2 holding h3) Rab8!? (Black anticipates opening the b-file) 23.c4 (Now Black must exchange a critical pawn and give White a solid pawn majority in the center) bxc4 24.bxc4 Nf6 25.Nf2 Qc5 26.Rd1 (Preparing d4) Qa5 27.Rd2 Bf5 28.d4 e4 29.Nh2 (Intending to recycle the Knight to the f1 square) Kh8 30.Nf1 (I'his obeys the edict of Nimzowitch to overprotect the base of the pawn chain) Qb4 31.Nd1 Qd6 32.c5 Qe7 33.Rb2 Rxb2 34.Qxb2 (Each exchange brings White closer to victory) Nd5? (The danger of the passed pawn is over the computer's horizon) 35.Bxa6 (Houdini is in mild time trouble---18 minutes until the time control; I have over an hour on the clock. Clearly Houdini's evaluation algorithm has gone haywire. Giving me an unstoppable passed pawn is a bad idea.) Rb8 36.Qe2 (Just connecting my pieces and anticipating Nb4/Nd3) Qg5 (With the obvious hit on h3) 37.Kh2 h5? (At long last the win should be a matter of technique) 38.Bc4 Nce7 39.Bxd5? (Missing the simple a4/a5/a6---Passed pawns must be pushed!) Nxd5 40.Qe1 (Just getting ready to push the a-pawn. I have good piece coordination and two extra pawns Re8 41.a4 Bc8 42.Nc3 Nxc3 43.Qxc3 (I was threatening to walk off with the e-pawn) h4 44.Qe1 (When I push the a-pawn I can get Black to give up pressure on the h3 square. I can combine threats to Queen the pawn and occupy the e5 square with my Knight) Ba6 45.Nd2 Bd3 46.a5 Re6 47.g4 (Necessity is the mother of invention; I need to defend the g2 square) g6 48.Ra2 Kg7 49.Rb2 Re7 50.Ra2? (White was exhausted) Re6 51.Ra3 Ra6 52.Ra1 Qd8 53.Kg2 c6 54.Nf1 Ra7 55.Nh2 (Draw offer declined) g5 56.Nf1 Bxf1 57.Kxf1 Rf7ch 58.Kg2 Rf3 59.a6 Rg3ch 60.Kh2?? (Kh1 =) Qc7 0-1

Houdini 3-Moody Game 10 (What is surprising is that Houdini suffers from he horizon effect) 4.Ng5 Bc5 5.d4 d5 6.Bxd5 Nxd4 7.Nxf7 Qe7 8.Nxh8 Bg4 9.Qd3 (White should play Bf7ch) Nxd5 10.c3 Nb4 11.cxb4 Bxb4ch 12.Nc3 Qd7 13.Kf1 Bxc3 14.f3 (Obviously if this is best White has serious problems) Bxf3 15.gxf3 Qh3ch 16.Kf2 Bb4 17.a4 Rd8 18.Bg5 Bc5 19.Qc3 Nb3ch 20.Qxc5 Nxc5 21.Bxd8 Kxd8 (Houdini claims a big advantage Black here) 22.Nf7ch Ke7 23.Ne5 Qe6 24.b4 Nxe4ch 25.Kg2 Qxe5 26.Rad1 Qd6 27.Rxe5ch Kf8 28.b5 c5 29.bxc6 Qxc6 30.Rae1 Qd7 31.a5 g6 32.R1e2 Kf7 33.Rb2 Qc7 34.R4b4 b6 35.axb6 axb6 36.Rxb6 Qf4 37.Rb7ch Kf6 38.Rxh7 Qg5ch 39.Kh3 Qf5ch 40.Kg2 Qg5ch 41.Kf2 Qc5ch 1/2-1/2

This is a test of the Berliner Gambit in a variation that BCO 2 says is equal.

Moody-Houdini 3 Game 12 There is a pretty zugawang position here.

4.Ng5 d5 5.exd5 Nd4 6.c3 b5 7.Bf1 Nxd5 8.Ne4 Qh4 9.Ng3 Bg4 10.f3 e4 11.cxd4 Bd6 12.Qe2 O-O (Be6? 13.Nc3 +-, 13.Qf2 +-, 13.Qxb5ch +-) 13.fxg4 Bxg3ch 14.Kd1 c6 15.Nc3 Rfe8? (Black must keep pieces on the board at all costs; simplification just loses without allowing counterplay) 16.Nxd5 cxd5 17.b3  (You can't give White three obvious moves in a row) Bf2 18.Bb2 b4 19.Rc1 e3 20.d3 Rac8 21.g3 Qf6 22.Bg2 Qd6 23.Rc2 h6 24.Kc1 Red8 25.Bf3 Rxc2ch 26.Kxc2 (I can hit d5 three times) Qf6 27.Kb1 g6 28.Rc1 a6 29.Rc5 Qe6 30.Bc1 Kg7 31.Kc2 Bg1 32.Kd1 Rd7 33.Qg2 Bf2 34.Bb2 a5 35.Ke2 Re7 36.h3 Qd6 37.Bxd5 Bxg3 38.Bc4 Kh7 39.Rxa5 Qc7 40.Re5 1-0

Here is the second game in this variation, Moody-Houdini 3 Game 8. Houdini richly deserves its name; here it squirms out of a slightly inferior position and swindles a win:

5.Bxf7ch Ke7 6.Bc4 Na5 7.Bd3 d6 8.Nc3 h6 9.Nf3 Be6 10.Be2 Nc6 11.d3 a6 12.Na4 Ba7 13.O-O Kf7 14.h3 Rf8 15.b3 Kg8 16.c3 Qd7 17.Be3 Bxe3 18.fxe3 b5 19.Nb2 Bxh3 20.gxh3 Qxh3 21.Nh2 Qxe3ch 22.Kh1 (Black doesn't have quite enough for the piece, but I get outplayed from here) Ne7 23.Rf3 Qc5 24.Qd2 Ng6 25.R1f1 Nf4 26.Nd1 d5 27.exd5 Qxd5 28.Ne3 Qd6 29.Rd1 Rae8 30.Qc2 h5 31.Bf1 c5 32.Ng2 Ng4 33.Nxf4 exf4 34.Qd2 Ne3 35.Re1 g5 0-1 able to hold this position, but not me so I resigned)

1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 6.Nc3 Bxa6 7.Bg5! (For some reason this obvious move is not part of Houdini's book. It is obviously a huge oversight in theory because it is not mentioned by MCO 15. Now White is minutely better; as this game shows I had no difficulty in coming up with a simple practical plan against Houdini 3) Bg7 8.e4 Qa5 9.Qd2 d6 10.Bxa6 Qxa6 11.Nge2 Nbd7 12.O-O O-O 13.f3 Rab8 14.Rab1 Ne5 15.b3 (White completes his anti-Bishop strategy of taking away targets on the diagonal; a major purpose is to keep the Knight out of c4, Black's only source of counterplay. The fundamental problem with the Benko Gambit is that once White finds Bg5, he can complete his development without having to distort his position.) Rb7 16.Rfd1 +/= (White has spectacular piece coordination and a solid passed pawn) Re8 17.h3 c4 18.Nd4Nfd7 19.Be3 Nc5 20.Qc2 Ra8 21.b4 Ncd3 22.b5 Qa3 23.f4 Nd7 24.Nde2 Nb4 25.Qd2 Nd3 (Houdini tacitly offers a draw. Now White must make slow but steady progress. White must improve the position of his only undeveloped piece, his King) 26.Rf1 Nb6 27.Bxb6 Rxb6 28.g4! Bf6 29.Kh2 Rb7 30.Kg3 h6 31.h4! Qc5 32.a4 h5 33.g5 Bg7 34.Rf3 Qc8 35.f5 Be5ch 36.Kg2 Qc5 37.Rbf1 Rbb8

There is no way for either side to make progress without risk  so  the position is equal.

Post members maintain that the Berliner Gambit is "obviously" bad, but much to my surprise Deep Fritz 14 plays the Berliner Gambit exclusively. In the critical position, it plays one of two moves. Here it plays its preferred choice:

Reality is a wonderful place, sloughterchess.  You can see it when you open your eyes.

Deep Fritz 14 played book in the main line of the Two Knights' Defense; only White has winning chances but I could see no way to make progress an forced a draw by repetition. An endgame specialist might have winning chances here.

While the Fried Liver gives White a sizeable initiative, Black seems to be able to hold. Here in the Lolli, Black can rely on the Heisman Variation to sacrifice a piece for two pawns and although Phren claims that Black does not have full compensation for the piece, I have been unable to secure a lasting advantage with White.

Here in one of the key variations, White picks up the exchange for a pawn, but was unable to make progress. Maybe someone with better endgame skills could prevail.