Getting winning is not the same as staying winning. Across 300,758 analysed games in which a player reached an advantage of at least three pawns, 77,739 of them, 25.8%, finished with the evaluation back at level or worse.

Getting winning is not the same as staying winning.

Across 300,758 analysed games in which a player reached an advantage of at least three pawns, 77,739 of them, 25.8%, finished with the evaluation back at level or worse.

25.8%

Only positions from move five onward count as the peak, so opening noise cannot qualify as winning, and the endpoint is the final evaluation rather than the result, which also turns on the clock. One game in four that was won on the board was not won at the board.

Across 300,758 analysed games from 228 players.

The Club Ledger · No. 6

The most common single blunder on Eloist is castling short.

Across 335,411 analysed games containing a move that cost at least 300 centipawns, O-O is the culprit 3,712 times, 1.11% of them, ahead of every other move in the corpus.

1.11%

No move accounts for a large share of blunders; the point is the ranking. O-O tops a list of thousands of distinct moves, which is not where a player would look for the worst move of their game.

Across 335,411 analysed games from 246 players.

The Club Ledger runs again tomorrow.