First-move Advantage in Chess - Winning Percentages

Winning Percentages

In 1946, W.F. Streeter examined the results of 5,598 games played in 45 international chess tournaments between 1851 and 1932. Streeter found that overall White scored 53.4% (W: 38.12; D: 30.56; L: 31.31). White scored 52.55% in 1851–78 (W:45.52; D: 14.07; L: 40.41), 52.77% in 1881–1914 (W: 36.89; D: 31.76; L: 31.35), and 55.47% in 1919–32 (W: 36.98; D: 36.98; L: 26.04). Streeter concluded, "It thus appears that it is becoming increasingly difficult to win with Black, but somewhat easier to draw."

Two decades later, statistician Arthur M. Stevens concluded in The Blue Book of Charts to Winning Chess, based on a survey of 56,972 master games that he completed in 1967, that White scores 59.1%. However, Stevens assembled his games from those that had been published in chess magazines, rather than complete collections of all the games played in particular events.

More recent sources indicate that White scores approximately 54 to 56 percent. In 2005, GM Jonathan Rowson wrote that "the conventional wisdom is that White begins the game with a small advantage and, holding all other factors constant, scores approximately 56% to Black's 44%". International Master (IM) John Watson wrote in 1998 that White had scored 56% for most of the 20th century, but that this figure had recently slipped to 55%. The website ChessGames.com holds regularly updated statistics on its games database. On March 17, 2008 the database contained 460,703 games. White won 36.81%, 36.50% were drawn, and Black won 26.69%, resulting in a total White winning percentage of 55.06%.

New In Chess observed in its 2000 Yearbook that of the 731,740 games in its database, White scored 54.8% overall; with the two most popular opening moves, White scored 54.1% in 349,855 games beginning 1.e4, and 56.1% in 296,200 games beginning 1.d4. The main reason that 1.e4 was less effective than 1.d4 was the Sicilian Defence (1.e4 c5), which gave White only a 52.3% score in 145,996 games.

Statistician Jeff Sonas, in examining data from 266,000 games played between 1994 and 2001, concluded that White scored 54.1767% plus 0.001164 times White's Elo rating advantage, treating White's rating advantage as +390 if it is better than +390, or −460 if it is worse than −460. He found that White's advantage is equivalent to 35 rating points, i.e. if White has a rating 35 points below Black's, each player will have an expected score of 50%. Sonas also found that White's advantage is smaller (53%) in rapid games than in games at a slower ("classical") time control. In the 462 games played at the 2009 World Blitz Chess Championship, White scored only 52.16% (W38.96 D26.41 L 34.63).

Other writers conclude that there is a positive correlation between the players' ratings and White's score. According to GM Evgeny Sveshnikov, statistics show that White has no advantage over Black in games between beginners, but "if the players are stronger, White has the lead". An analysis of the results of games in ChessBase's Mega 2003 database between players with similar Elo ratings, commissioned by GM András Adorján, showed that as the players' ratings went up, the percentage of draws increased, the proportion of decisive games that White won increased, and White's overall winning percentage increased. For example, taking the highest and lowest of Adorján's rating categories of 1669 games played by the highest-rated players (Elo ratings 2700 and above), White scored 55.7% overall (W26.5 D58.4 L15.2), whereas of 34,924 games played by the lowest-rated players (Elo ratings below 2100), White scored 53.1% overall (W37.0 D32.1 L30.8). Adorján also analyzed the results of games played at the very highest level: World Championship matches. Of 755 games played in 34 matches between 1886 and 1990, White won 234 (31.0%), drew 397 (52.6%), and lost 124 (16.4%), for a total White winning percentage of 57.3%. In the last five matches in Adorjan's survey, all between Anatoly Karpov and Garry Kasparov, White won 31 (25.8%), drew 80 (66.7%), and lost 9 (7.5%), for a total White winning percentage of 59.2%.

Chess Engines Grand Tournament (CEGT) tests computer chess engines by playing them against each other, with time controls of forty moves in one hundred and twenty minutes per player (40/120), and also 40/20 and 40/4, and uses the results of those games to compile a rating list for each time control. At the slowest time control (40/120), White has scored 55.4% (W34.7 D41.3 L24.0) in games played among 38 of the strongest chess engines (as of May 27, 2009). At 40/20, White has scored 54.6% (W37.0 D35.2 L27.8) in games played among 284 engines (as of May 24, 2009). At the fastest time control (40/4), White has scored 54.8% (W39.6 D30.5 L30.0), in games played among 128 programs (as of May 28, 2009).