Chinese Opening success rate

[RedStick]

Anyone who follows professional games knows that the low Chinese has been very popular recently. My question is simple: is it working?

Can anybody do a database search over the past year to see what percentage of the time B wins when starting with low Chinese?

[TMark]

I used our GoGoD database for the period 2009-2010 and then for 2010 alone. The figures I have are that for the first period we have 444 games with Black winning 48.2% and White 51.4%. For 2010 on its own, we have 291 games, Black 48.1%, White 51.9%.

Best wishes.

[RedStick]

Thanks! I'm surprised to see B winning less than 50%.

I wonder what opening will gain popularity next.

[Mef]

Thanks! I'm surprised to see B winning less than 50%.

I wonder what opening will gain popularity next.

Just speculating....it does look like it could be getting played more and more (153 in 2009, 291 in 2010?), perhaps it's that there are more game records available for later stages of tournaments, and those who are particularly good at playing against the Chinese opening have had success. Perhaps it ends up that there are a disproportionate number of games records published with a player holding white that is particularly good against the Chinese. Or of course, perhaps it's just that 1.8% is well within the statistical noise of the sampleset (if my fading memory of statistic serves correct, standard deviation for 444 games at 50% probability is something like 10 games, or approx 2.5%?).

[Joaz Banbeck]

Do we have figures for a control group? In other words, what are the black/white win proportions for other fusekis? Without that, the above figures are not meaningful information about the Chinese fuseki - even if there is something coming through the noise.

[TMark]

We have 2244 games from 2010, with a success rate for Black overall of 50.2%. Taking two quick examples, if both sides play nirensei, Black wins 41.4%, White 58.6% (29 games), but if Black plays the
"Kobayashi" fuseki, the figures are Black 54.9%, White 45.1% (144 games).

Best wishes.

[tj86430]

And then there is the question of causality vs. correlation