Author: Dann Corbit
Date: 18:03:49 11/20/98
Go up one level in this thread
On November 20, 1998 at 20:25:13, syed wrote:
>Assuming one did make a collection of all chess games ever played in the modern
>era (this century), would it be possible to figure out the best move ever made
>for a given position in the middle game, for a given opening? Could this
>possibly be achieved a series of SQL select queries?
A simple select will only give you a pretty good guess. Imagine how good the
estimate would be if a Pentium II 300 Mhz computer thought about that position
for twelve minutes. That is the initial estimate. However, once the data has
been completely analyzed, then you can refine the data in the database with more
queries. For instance, suppose that position x has five leaf nodes that either
analyze well or were played to success by various GM's or whatever. I can now
examine the leaf nodes and see what they look like. I can update node x with
the new information. I can continue the process as long as I like.
Quite frankly, I have no idea how useful the end result might be. One thing
that is guaranteed is that all the positions will have extremely thorough
'blunder' checks. We would be able to find the move that evaluates the highest,
or the move that has the highest win percent, etc. But best is very difficult
to quantify. However, once the best move has been determined, there is a 'bm'
field in the database to indicate that fact. Unless a move clearly and
unmistakeable leads to mate faster than any conceivable other possibility, we
won't really know if it is the best move or not. Perhaps a clever pin that
takes a queen in 7 prevents us from a clever move that mates in 9.
Consider the following query:
SELECT epd, acd, acn, ce, pv
FROM dbo_Epd
WHERE abs(ce) >= 32000
AND abs(ce) < 32750
ORDER BY abs(ce);
It produces the following result set with checkmates as far away as 41 plies:
epd acd acn ce pv
2k5/R4N2/4p3/8/2P1rP2/8/3K4/8 w - - 13 137286402 32726 Nd6+ Kb8 Rb7+ Ka8 Nxe4
Kxb7 f5 exf5 Nd6+ Ka6 Nxf5
2k5/R7/3Np3/8/2P1rP2/8/3K4/8 b - - 14 137125125 -32727 Kb8 Rb7+ Ka8 Nxe4 Kxb7 f5
exf5 Nd6+ Ka6 Nxf5
2r5/1K6/6p1/7p/3k3P/8/8/8 w - - 27 167880968 -32737 Kxc8 Ke5 Kd7 Kf5 Ke8 g5 hxg5
Kxg5 Kf8 h4 Kf7 h3 Ke8 h2 Ke7 h1=Q
1Q1Q4/8/8/3K2pk/r7/8/8/8 w - - 2 629 32738 Qxg5+ Kxg5
2K5/4k3/4P3/5P2/4B3/6p1/8/8 w - - 13 142641043 32742 Bg2 Kf6 Kd7 Kxf5 e7 Ke5
e8=Q+ Kd4 Qe4+ Kc3 Qe3+ Kc2 Be4+ Kb2 Qd2+ Ka3 Bd5 Ka4
2K5/4k3/4P3/5P2/8/6p1/6B1/8 b - - 13 154303007 -32743 Kf6 Kd7 Kxf5 e7 Ke5 e8=Q+
Kd4 Qe4+ Kc3 Qe3+ Kc2 Be4+ Kb2 Qd2+ Kb3 Bd5+ Ka4 Kc6
1R1Q4/5k2/p1p1pPb1/P4qp1/B3p3/4P3/5K1P/8 w - - 10 28021196 32744 Ke1 Bh7 Rb7+
Kg6 Qe8+ Kh6 f7 Kh5 f8=Q+ Kg4 Bd1+ Kh3 Qh5+ Kg2 Qe2+ Kg1 Qf1+ Qxf1+ Qxf1+ Kxh2
Rxh7+ Kg3 Rh3#
1k6/1P6/P7/P2p1pp1/3P1P1p/4K1PP/8/8 w - - 11 28674854 32746 fxg5 hxg3 g6 f4+ Ke2
g2 Kf2 Ka7 g7 g1=Q+ Kxg1 f3 g8=Q Kxa6 b8=Q f2+ Kxf2 Kxa5 Qxd5+ Ka4 Qbb3#
1R3r2/8/k7/3K4/p7/8/8/8 w - - 17 115517867 32746 Rxf8 Kb6 Ra8 a3 Rxa3 Kb5 Rb3+
Ka4 Kc4 Ka5 Kc5 Ka6 Rb6+ Ka7 Kc6 Ka8 Ra6+ Kb8 Ra5 Kc8 Ra8#
1r4k1/2R5/5PK1/8/8/8/8/8 w - - 13 39245897 32746 Rg7+ Kf8 Rh7 Ke8 Rh8+ Kd7 Rxb8
Kc7 Rb4 Kd6 f7 Kc5 f8=Q+ Kd5 Qf6 Kc5 Qb6+ Kd5 Rd4+ Ke5 Qf6#
1k1rr3/1p1qbppp/8/1R6/2Q5/4PBP1/5P1P/2R3K1 b - - 9 117393995 -32747 Qd1+ Rxd1
Rxd1+ Kg2 Rb1 Qd5 Rxb5 Qxb5 Kc7 Qxb7+ Kd6 Qd5+ Kc7 Qc6+ Kd8 Qb6+ Kd7 Bg4+ f5
Bxf5#
1k6/1P6/P7/P2p1pP1/3P3p/4K1PP/8/8 b - - 12 91180383 -32747 hxg3 g6 f4+ Kf3 g2
Kxg2 Ka7 g7 Kxa6 b8=Q Kxa5 g8=Q f3+ Kxf3 Ka6 Qgc8+ Ka5 Qd8+ Ka6 Qbb6#
1n4K1/1P6/5k1P/p7/P7/8/2B5/8 b - - 12 141176274 -32747 Ke5 h7 Nd7 h8=Q+ Kd5 Qh5+
Kc6 Be4+ Kd6 b8=Q+ Nxb8 Kf7 Nd7 Qd5+ Kc7 Ke7 Kb6 Kxd7 Ka6 Qb7#
1r4k1/6R1/5PK1/8/8/8/8/8 b - - 15 161223356 -32747 Kf8 Rh7 Ke8 Rh8+ Kd7 Rxb8 Kc7
Rb4 Kd6 f7 Kc5 f8=Q+ Kd5 Qd8+ Kc6 Kf5 Kc5 Qb6+ Kd5 Rd4#
1r4k1/7p/p1b3p1/4Q3/2qBN3/2P2P2/1nB4P/1R5K b - - 11 125857624 -32747 Qxd4 Qxb8+
Kg7 cxd4 Nd3 Bxd3 a5 Qe5+ Kh6 Qg5+ Kg7 Qe7+ Kh6 Qf8+ Kh5 Nf6+ Kg5 Rg1+ Kf4 Nd5#
2R5/5Qpk/2P4p/3N2q1/6P1/4P2K/1b3P1P/2r5 b - - 10 134451009 -32747 h5 gxh5 Qg1
Qf5+ g6 Rc7+ Bg7 Rxg7+ Kxg7 Qf6+ Kh7 Qf7+ Kh6 Qf8+ Kxh5 Qh8+ Kg5 f4+ Kf5 Qe5#
1k6/1P6/P7/P2p1pP1/3P4/4K1pP/8/8 w - - 11 17136767 32748 g6 f4+ Kf3 g2 Kf2 Ka7
g7 Kxa6 b8=R f3 g8=Q g1=Q+ Kxg1 f2+ Kf1 Kxa5 Qxd5+ Ka4 Qa8#
1r5k/6R1/5PK1/8/8/8/8/8 w - - 9 2785741 32748 Rh7+ Kg8 f7+ Kf8 Rh8+ Ke7 Rxb8 Kd6
f8=Q+ Kd5 Rb4 Kc6 Qc8+ Kd6 Rd4+ Ke5 Qc5+ Ke6 Qd6#
2b5/1p1n2k1/2pq3p/8/1P1PQN2/1B2P3/1K3PP1/8 w - - 9 9670213 32748 Qe8 Qf8 Ne6+
Kf6 Nxf8 Nxf8 Qxf8+ Kg6 Qf7+ Kg5 Qg7+ Kf5 e4+ Kf4 Qg3+ Kxe4 Qe5+
2R5/5Qpk/2P5/3N2qp/6P1/4P2K/1b3P1P/2r5 w - - 9 12159282 32748 gxh5 Qg1 Qf5+ g6
Rc7+ Bg7 Rxg7+ Kxg7 h6+ Kg8 Qc8+ Kh7 Nf6+ Kxh6 Qf8+ Kg5 Ne4+ Kh5 Qh8#
2R5/5Qpk/2P5/3N3P/8/4P2K/1b3P1P/2r3q1 w - - 7 2193692 32748 Qf5+ g6 Rc7+ Bg7
Rxg7+ Kxg7 h6+ Kg8 Qe6+ Kh7 Qf7+ Kxh6 Qf8+ Kh5 Qh8+ Kg5 f4+
2r5/5r1k/4q1p1/QP4B1/P1ppp1Pp/R1n2P1P/6K1/5R2 b - - 10 21845161 32748 exf3+ Kg1
f2+ Kg2 Qe4+ Kh2 Qe5+ Kh1 Qe2 Kh2 Qxf1
1nr5/2q1P1Q1/1p1k2P1/pPp1RP2/P2p4/3P2K1/8/4R3 b - - 11 116183922 -32749 Nd7 Kf2
Qb7 e8=Q Rxe8 Rxe8 Kc7 Qxd7+ Kxd7 R1e7+ Kd6 Rxb7 c4 dxc4
1R1Q4/5k2/p1p1pPb1/P4qp1/B3p3/4P3/7P/4K3 b - - 11 141104134 -32749 Bh7 Rb7+ Kg6
Qe8+ Kh6 f7 Kh5 f8=Q+ Kg4 Qxf5+ Bxf5 Bd1+ Kh3 Rb2 Bg4 Qh8+ Bh5 Qxh5#
1r4k1/R5Rp/3q1p2/6p1/3p4/1rnP1BPP/5P1K/4Q3 b - - 11 152924579 -32749 Kh8 Bd5
Nxd5 Qe4 Qxg3+ Kxg3 Rxd3+ Qxd3 f5 Rxh7+ Kg8 Rh8+ Kxh8 Qxd4+ Nf6 Qxf6+ Kg8 Qg7#
1R6/6k1/2PK1p2/6p1/4p3/4P3/4bP2/R7 b - - 11 157790309 -32749 Bg4 Ra7+ Kh6 Rh8+
Kg6 Rg8+ Kh6 Ke7 Kh7 Kf8+ Bd7 cxd7 Kh6
1rq2n1k/pN2R2p/6p1/P2B1p2/3P4/1Q6/1B4PP/5RK1 b - - 10 136138121 -32749 Qc7 Bf7
Nd7 d5+ Ne5 Rxe5 Qc3 Re8+ Kg7 Qxc3+ Kh6 Qh3+ Kg5 Bc1+ Kf6 Qc3+ Kxf7 Nd6#
2R5/5Qpk/2P5/3N2qP/8/4P2K/1b3P1P/2r5 b - - 11 126788147 -32749 Qg1 Qf5+ g6 Rc7+
Bg7 Rxg7+ Kxg7 h6+ Kg8 Qc8+ Kh7 Nf6+ Kxh6 Qf8+ Kg5 Ne4+ Kh5 Qh8#
I suspect that in *most* of the cases, the actual players did not find these
mates.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.