Can Houdini 3 be beaten?

when I first readed this thread I thought it was about Can Houdini 3 be beaten (by another chess program), Rybka development isn't further allowed?

I just looked it up because it seems hard to believe, but you're probably right! That's just insane

Phren is right---I still play endgames at a 1500 level, but the following game illustrates that I am really playing these games not Rybka.

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 g6 5. c4 Nf6 6. Nc3 d6 7. f3 Nxd4 8.Qxd4 Bg7 9. Be2 O-O 10. Qf2 Be6 11. Bd2 Nd7 12. Rc1 Qb6 (At this point Houdini recommends Qxb6; I don't even consider Qxb6 as a candidate move because it loses against a program like Houdini due to either axb6 or Nxb6) 13. b3! Bd4 14. Qf1 Nf6 (I prefer f5 here; according to how I define development this just loses a tempo)  15. Nd1! (This gains a tempo and prepares to contest the strong Bishop)  Rfc8 16. Bc3 Bc5 17. Nb2! Bb4 18. Na4 (Gaining a tempo) Qa5 19. Kf2 (Gaining a tempo) Bxc3 20. Nxc3 (White has made progress; he has eliminated the strong Bishop)  Qc5+ 21.Ke1 d5 22. exd5 Qe3! 23. Rc2 Bxd5 24. Qf2 Qe5 25. f4 Qf5 26. Rd2 Bc6 27. Qe3 Rd8 28. Qe5 Rxd2 29. Kxd2 Rd8+ 30. Kc1 Qxe5 31. fxe5 Bxg2 32. Rg1 Ne4 33. Rxg2 Nxc3 34. Bf3 b6 35. a3 * And this is why I lose to Houdini; it will win this endgame 100 out of 100 tries.

I equalized easily with Black against Houdini. Unless I am mistaken this may have significant theoretical value.

1. d4 d5 2. c4 c5 3. Nf3 Nf6 4. cxd5 cxd4 5. Qxd4 Qxd5 6. Nc3 Qxd4 7. Nxd4 a6 8. g3? e5! = 9. Nc2 Nc6 10. Bg5 Be6 11. Bg2 Be7 12. Rd1 Rd8 13. O-O O-O 14. Ne3 Rxd1 15. Rxd1 Rd8 16. Rxd8+ Bxd8 17. Bxf6 Bxf6 18. Bd5 Bxd5 19. Nexd5 Bd8 20.Kg2 f6 * =

You were following Levenfish-Lasker 1936 until move 10. 

I knew that without having any assistance. 

I'm very concerned about fracking.

2...c5? cxd5 and white has a significant advantage.

Your computer must be weak if you can equalize against it in that line.

Good job on equalizing with black against Houdini. Find an advantage for black and report back :-)

haha yea good luck

How does black equalize after 3. cxd5? I'm sure it's not with Qxd5 which just loses tempo to e3 or Nf3 followed by Nc3 when white has a clear advantage, or with 3...Nf6 e4! Nxe4 dxc5 Nxc5, when once again, white has a clear advantage.

So what move do you suggest on move 3 for black, knowing that the two most logical tries fail? e6 perhaps?

After 3...e6 dxe6 Bxe6 I don't believe black has enough for the pawn, although he probably has some sort of measely compensation. White should once again be better.

Maybe perhaps 3...cxd4 Qxd4 Nf6 e4 e6 Nc3 Nc6 Qd1, when white is up a pawn for nothing and perfectly happy?

The opening is refuted, I don't care if Lasker played it, or even Kasparov or Fischer. It is no good.

White has the first move advantage only and has no more likelihood of converting that advantage to the full point that any other major opening. Here is a sample drawing line where White is pawn up with minimal chances of winning between GM's.

1. d4 d5 2. c4 c5 3. cxd5 Nf6 4. e4 Nxe4 5. dxc5 Nxc5 6. Nc3 e6 7. Nf3 Be7 8.Bb5+ Bd7 9. Bc4 Bf6! 10. O-O O-O 11. Bf4 Bxc3 12. bxc3 exd5 13. Qxd5 Ne6 14. Qd6 Nxf4 15. Qxf4 Nc6 16. Rad1 h6 17. Nh4 Qg5! 18. Qxg5 hxg5 19. Rxd7 gxh4 20. Rxb7 Na5 21. Rc7 Nxc4 22. Rxc4 Rac8 23. Rxh4 Rxc3 * Double Rook ending are notoriously equal sometimes even two pawns up. This is no exception.

Wow, I already am not a fan of Bb5+, or some other moves white played in the game, and white still had a solid advantage throughout the entire game. Black didn't even have a chance to equalize.

I would prefer natural development through Be3 in that position, white has a solid lead in development and should look for a chance to quickly exploit it.

The same superficial hope chess again...

It's well known that black is in trouble after 6...e6 7.b4!, and the same move is also very strong against 6...e5.

I agree though that your chances to get at some decent position using that mentality are, errr, minimal.

The Four Knights is drawish and I was able to equalize with Black fairly easily.

1. e4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. Bb5 Bc5 5. O-O O-O 6. d3 Re8 7. Bxc6 dxc6 8.h3 a5 9. Qe2 h6 10. Be3 Qe7 11. Bxc5 Qxc5 12. Qe3 Qxe3 13. fxe3 c5 14. a4 c6 15. Rf2 b6 16. g4 Nh7 17. Rd2 f6 18. d4 cxd4 19. exd4 exd4 20. Rxd4 h5 21. gxh5 Bxh3 22. Kf2 Rad8 23. Rxd8 Rxd8 24. Rg1 Kf7 * =

One of the fundamental principles I've learned with Houdini is that if you develop rapidly and have good piece coordination there are no tactics. The more disjointed your army is the deeper the tactics. Here at a critical juncture post members may wonder whether I calculated 13...Qxc3 forking Bishop and Rook. The answer is that I don't have to calculate 13...Qxc3 because I have made nothing but developing moves and useful moves; I can't be hanging a piece so there has to be a tactical refutation which is trivial to find i.e. 14.Qb3. The point is that after 14...Qxa1 15.Ba3 Qg6 16.Bxf8 & Black can't play Kxf8 because of Qxb7. White has more than adequate compensation for the exchange.

I discontinued the game here because obviously the double minor exchange is worth a pawn.   

1. e4 e6 2. d4 d5 3. exd5 exd5 4. c4 Nf6 5. Nc3 Bb4 6. Nf3 dxc4 7. Be2 O-O 8.O-O Be6 9. Ng5 Qd7 10. Nxe6 Qxe6 11. d5 Qe5 12. Bxc4 Bxc3 13. bxc3 Qxc3 14. Qb3 Qxb3 15. axb3 Nbd7 16. Bf4 Rfc8 17. Rfd1 a5 18. Bd2 a4 19. Rxa4 Rxa4 20. bxa4 Nb6 21. Bb3 Nbxd5 22. a5 Rd8 23. f3 Ne7 24. Kf1 Ned5 25. Ke2 c6 26. Kf1 Ne7 27.Ke2 Nf5 28. Bc3 Rxd1 29. Bxd1 Nd5 30. Bd2 Kf8 31. Kd3 Ke7 32. g4 Nd6 * =

Tracking off. Houdini would beat you in its sleep.

tracking off too, IM Pfren has very sarcastic comments (also he mistakes sometimes-- he is human)

White just chopped wood and entered a BOC ending with a symmetrical pawn structure.

No one cares.  Please stop now.

I'm going to issue a challenge to the manufacturer of Houdini 3 that I can win or draw against Houdini 3 in at least 5 out of 20 games. I will consult with a local expert in the match. The stakes will be $1000. The games will be posted on this thread.