Originally posted by KingOnPoint
With the following partial game shown at http://www.365chess.com
Efim Geller vs. Istvan Csom
Budapest 1973 · Sicilian, Richter-Rauzer, Rauzer attack, 7...a6 (B66) · 1-0
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 Nc6 6. Bg5 e6 7. Qd2 a6 8. O-O-O h6 9. Be3 Bd7 10. f4 Qc7 11. Bd3 Nxd4 12. Bxd4 Bc6 13. Rhe1 O-O-O 14. Qf2 Nd7
Acco ...[text shortened]... " a bad move outright rather than just statistically? And is 14. ...Nd7 better than 14. ...Ng4?
Simple stats such as this tell you almost nothing, and can even give you a false impression.
For instance, did you know that out of 100 games with 11. Kb1 white won the first 60, but an innovation by black in the variation turned the tables, and black has won the last 40?
With 11. Bd3 it's the opposite case. Out of 100 games white had an ill-conceived plan in the first 75, but a subsequent white innovation at move 12 means that white has won the last 25 in a row.
Also, in the case of 11. Kb1, the white players average ratings were 200 points higher than the black players. With 11. Bd3 the black players were just over 300 points higher than the white players.
It is a special case with 11. h3. Computer analysis reveals that the move is an almost winning advantage with very accurate play, but the ending is very hard, and in practice (between two unassisted human players) black draws the position about 90% of the time.
Of course, I just made all this up. Don't we just love statistics!