mike860 wrote:genorb wrote:This is indeed a good way: using ply depth levels.
(...) I've also tried this approach with my chess software and I'm still getting ripped apart at one or two plies. This is probably because I have a Q6600 @ 3.30GHz and I have not limited my "deep" engines to one core instead of four.
This cannot be the reason, because a depth of i.e. 2 plies is always the same, disregarding the number of cpu cores used to achieve it (which will not make a noticable difference up to ~10 plies). It only means that higher depths will be reached quicker.
The only difference is that some moves may be not reproducable under the same conditions, but I think that should only, or at leat mostly affect positions where alternatives are within a narrow margin.
I think the strength difference to depth limited chess comps mainly comes from the much more effective extensions of the modern programs (but also from better evaluation). It's difficult. I just tried it a couple of times against Spike 1.2 at depth 2 and I always failed.
A major problem is playing too fast because the program always responds immediatly. I am probably not patient enough to play as slow as I would need to.
But actually, if I'd be in the mood to play slower and more serious, I'd probably rather play against an umlimited engine although that is hopeless.
To create an example, here is a game where Rybka "ultra optimistic" plays unlimited with approx. 3 seconds per move (D945 3.4 GHz) against Spike 1.2/2 plies. Unlike me, Rybka found many little tricks and all seemed to work.
[Event "experimental"]
[Site "?"]
[Date "2008.02.29"]
[Round "?"]
[White "Spike 1.2 Turin-2 plies"]
[Black "Rybka ~3 seconds"]
[Result "0-1"]
[PlyCount "83"]
{128MB, History107.ctg, COMPUTER-11} 1. f4 {B/0 0} d5 {B/0 0} 2. Nf3 {B/0 0}
Nf6 {B/0 0} 3. e3 {B/0 0} e6 {B/0 0} 4. Nc3 {-0.14/2 0} Nc6 {5} 5. Bd3 {
-0.04/2 0} Bd6 {3} 6. O-O {-0.06/2 0} O-O {3} 7. Ng5 {0.17/2 0} h6 {3} 8. Nf3 {
-0.28/2 0} e5 {(a6) 3} 9. fxe5 {0.06/2 0} Nxe5 {4} 10. Nxe5 {0.06/2 0} Bxe5 {3}
11. Qf3 {-0.32/2 0} Bg4 {3} 12. Qf2 {-1.08/2 0} Be6 {(c6) 3} 13. Qh4 {-0.65/2 0
} Ng4 {(d4) 2} 14. Qxd8 {-1.80/2 0} Bxh2+ {3} 15. Kh1 {-1.80/2 0} Raxd8 {3} 16.
Be2 {-2.00/2 0} Bg3 {3} 17. Nb5 {-2.57/2 0} Nh2 {(Nf2+) 3} 18. Rg1 {-2.60/2 0}
Rfe8 {(a6) 3} 19. Nxa7 {-2.12/2 0} c6 {(c5) 2} 20. Bd3 {-2.14/2 0} Ng4 {(h5) 3}
21. Rf1 {-2.99/2 0} Nf2+ {2} 22. Kg1 {-3.37/2 0} Nxd3 {3} 23. cxd3 {-3.04/2 0}
d4 {(h5) 2} 24. Rf3 {-3.32/2 0} Bb8 {(Bh4) 4} 25. b3 {-6.10/2 0} dxe3 {(Bxa7) 3
} 26. Nxc6 {-5.95/2 0} bxc6 {2} 27. Rxe3 {-5.64/2 0} Ba7 {(h5) 3} 28. d4 {
-6.94/2 0} Bxd4 {4} 29. Rb1 {-7.71/2 0} Bf5 {(Bxe3+) 2} 30. Rb2 {-10.63/2 0}
Bxe3+ {2} 31. dxe3 {-10.63/2 0} Rd1+ {3} 32. Kf2 {-10.63/2 0} Rxc1 {2} 33. Rd2
{-10.04/2 0} Rc2 {(Be4) 2} 34. Rxc2 {-9.85/2 0} Bxc2 {3} 35. Kf3 {-10.13/2 0}
Ra8 {(Bb1) 2} 36. a4 {-11.11/2 0} Kf8 {(Bxb3) 2} 37. e4 {-11.12/2 0} Ke7 {
(Bxb3) 2} 38. Ke3 {-11.21/2 0} Bxb3 {3} 39. Kf4 {-12.34/2 0} Rxa4 {3} 40. g3 {
-12.61/2 0} Bd5 {(c5) 4} 41. Kf3 {-13.55/2 0} Bxe4+ {2} 42. Ke3 {-13.66/2 0}
0-1
The triumph of humanity: