Comments/Ratings for a Single Item

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

First move advantage in Shogi[Subject Thread] [Add Response]

Jörg Knappen wrote on Tue, Sep 25, 2012 08:18 AM UTC:

After following the long thread on the first move advantage in chess, I am
curious about the first move advantage in Shogi.

The major difference between chess and shogi lies in the "decisiveness"
of the two games: Western chess is rather drawish, while almost all shogi
games come out as wins or losses.

Is there a first move advantage in shogi (I don't know statistics, but I
suspect that there is a first move advantage, although some Shogi pages
claim the opposite) and how large is it?

Matteo Perlini wrote on Tue, Sep 25, 2012 09:36 AM UTC:

I really curious too. In the past I made some internet researches but
didn't find anything satisfying. So I hope someone can find some reliable
statistics.

Anyway, I could find just these two pages (of the same author):

http://shogi.typepad.jp/eweblog/2009/06/winning-percent.html

https://groups.google.com/forum/?fromgroups=#!topic/shogi-l/GixqTlUmSpI

Jörg Knappen wrote on Tue, Oct 2, 2012 04:02 PM UTC:

Thanks, Matteo, for digging out the reference. It says

"Of the 2,323 public matches in fiscal 2008, white players won 1,167
and lost 1,156, a win rate of 50.2 percent, it was discovered on
Tuesday. The previous highest win rate was 49.5 percent in fiscal
1968, and the lowest 46.4 percent in fiscal 2004."

So, there was a constant black (who moves first in Shogi) advantage for
4 decades, but in 2008 the situation was reversed. Given the relative small
number of recorded Shogi games, the 2008 result may be just a statiscal
fluctuation. Are there more recent numbers published somewhere?

Matteo Perlini wrote on Wed, Oct 3, 2012 03:24 PM UTC:

Hi Jorg,

I don't think it was just a statistical fluctuation in 2008. There were "2,323 public matches in fiscal 2008", not few matches. Maybe it is because Shogi is really balanced?

I would like to hear some comments from someone with a strong statistical knowledge.

p.s. Small sample, anyway: http://www.iggamecenter.com/stats/game95.html
Number of wins by the 1st player (excluding quits) 285 (41.91%)
Number of wins by the 2nd player (excluding quits) 357 (52.50%)

Jörg Knappen wrote on Thu, Oct 4, 2012 05:49 AM UTC:

The square root of 2323 is 48, but the difference is just 11. Nothing of
statistical significance.

To get at some conclusions one has to sum up the results of many years or
to extend the base of recorded games.

5 comments displayed

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

Permalink to the exact comments currently displayed.