Opening Frequency

 I did a little survey of chess opening frequencies from several chess databases. After looking over a million games or so and keeping tabs of the first few moves, these are the results.

1.e4 was the most common opening, representing 70% of the sample. This was followed by 1.d4, representing 20%. This frequency of 1.d4 goes up as more master games are included and amateur games excluded.  More amateurs still prefer 1.e4 over 1.d4.  More grandmasters prefer 1.d4 over 1.e4.

The next most common was 1.c4 (the English opening), representing 5% of the sample. Again, this number goes up with more master games and less amateur games. But our sample represents chess at all levels.

The next most common was 1.Nf3 (the Reti or King's Indian Attack), representing 4% of the sample. Thus, 1.e4, 1.d4, 1.c4, and 1.Nf3 represent 99% of all openings in most databases.

The next most common was 1.b3 (Larsen's Opening), closely followed by 1.b4 (Polish or Orangutan), 1.f4 (Bird's Opening) and 1.g3. These openings represented .9% of the sample.

Next came 1.Nc3 (Dunst), 1.g4 (Grob), 1.e3, 1.Na3, and 1.Nh3

The rarest moves were 1.a3, 1.c3, 1.f3, 1.h4, 1.d3, 1.a4, and finally 1.h3.

The most common response to 1.e4 was 1...e5. The next most common was 1...c5 (the Sicilian Defense).

Next came 1...e6 (the French Defense), followed by 1...c6 (Caro Kann).

Next in frequency was 1...Nf6 (Alekhine's Defense), 1...d5 (Center Counter), then 1...d6 (Pirc).

Next came 1...g6 (Pirc), 1...Nc6 (Nimzovich Defense), followed by 1...b6 (Owen's Defense).

Then came 1...a6, then 1...g5.

Finally, the rarest moves were 1...f6, 1...a5, 1...Nh6, 1...Na6, 1...f5, 1...h5, and 1...h6. I found no 1...b5 move.

The most common response to 1.d4 was 1...Nf6, followed closely with 1...d5.

Next came 1...f5 (Dutch). These 3 moves represent over 90% of the responses to 1.d4.

Next in frequency was 1...e6, followed by 1...c5.

Still rarer were 1...b6, 1...b5, 1...d6, 1...Nc6, 1...g6, and 1...e5

Finally, the rarest moves were 1...c6, 1...h6, 1...Nh6, and 1...f6.

I found no 1...a5, 1...a6, 1...Na6, 1...g5, or 1...h5.

» posted in Opening Theory

| 1194 reads | 8 comments |

1 vote:

Comments:

by antne003 - 3 years ago
RIO GRANDE, NEWJERSEY United States

Member Since: Jul 2007
Member Points: 280

 bill   ,  GREAT  RESEARCH, I ENJOY WANDERING AROUND THE FORUMS FOR ARTICLES  BY  YOU  AND MY  OPENINGS THAT I LIKE

                                 THANKS  BILL    TONY   antne003

by figrock - 3 years ago
United States

Member Since: Jul 2008
Member Points: 1438

Billwall, thanx for the statistics. I always enjoy your "words of wisdom".

by srn347 - 3 years ago
California United States

Member Since: Jul 2008
Member Points: 268

I can't believe the most common response to e4 is the unforgivable mistake e5. And that scandinavian is below french(no cultural war jokes or anything intended). Anyway, at least it supports(note I don't say proves) my claim that d4 is better than e4 is GMs and masters play it more.

by billwall - 4 years ago
Palm Bay, FL United States

Member Since: May 2007
Member Points: 3486

Not sure if there is a way, unless there is a big database of just master level and above games.  Perhaps if someone looks at all the Chess Informant games, then a count of White wins and Black wins for each opening variation.

by KingLeopold - 4 years ago
Scottsdale, AZ. United States

Member Since: Oct 2007
Member Points: 402

Bill,

Is there a way to see exactly which opening has the most wins at Master level and above, also within that opening, which variation.

I think that would be interesting for both White wins and which Black opening gives the best results against White. 

by Stu-pot - 4 years ago
Grimsby, UK United Kingdom

Member Since: Aug 2007
Member Points: 4

Really interesting statistics.  I don't necessarily think that White's opening moves, be it e4, d4, c4 or Nf3 necessarily have that much between them - it's more a matter of personal taste and the resultant middlegame positions which these moves tend to lead to.  Each of these four moves achieves a lot for an opening move, whereas the lesser seen moves don't accomplish as much (or at least not as quickly).  I think the same pretty much applies to Black's reply in that some moves clearly accomplish more than others, and it seems to be these moves which are seen most, statistically speaking.

by billwall - 4 years ago
Palm Bay, FL United States

Member Since: May 2007
Member Points: 3486

I just looked over the opening frequency at the 2007 US Championship.  There were 162 games.  1.e4 was played 82 times.  1.d4 was played 61 times,  1.Nf3 was played 12 times.  1.c4 was played only 5 times.  There was one game each with 1.b3 (a draw) and 1.f4 (a win for Black).  After 1.e4, 1...c5 was played 29 times.  1...e5 was played 21 times. 1...e6 was played 16 times.  1.e4 d5 was played 6 times.  1e4 g6 or 1.e4 d6 (the Pirc) was played 4 times.  1.e4 c6 was played 3 times.  1.e4 Nf6 was played twice (a draw and a win for Black).  Against 1.d4, 1...d5 was played 21 times.  1.d4 Nf6 was played 38 times.  1.d4 f5 was played once.  White won 58 games.  Black won 45 games.  59 games were drawn.  The shortest victory was 18 moves (1.e4 g6, won for Black).  There was also a 20 move victory (the Slav, 1.d4 d5 2.c4 c6, won by White).