Separated groups on average in pro games

I know a Go proverb or a condition, the more group of stones you have on the board, the more difficult maintain, that’s why its always recommended to connect groups if possible (especially the weak one), or there is always sacrifice or exchange.

Is there any statistic, how many group of stones on average that pro player usually manage in their game?

thanks in advance

3 Likes

I think the idea is not about how many you use to have but about how many MORE as your opponent.

Nothing special to notice if both players have 7 groups.

2 Likes

I just did a quick check for a recent 200 pro games, and only counted the games that went into scoring (for games ended in resigns for pro games, they often involved intense fighting, and hard to say which group in the fight might have lived or died, or even part of a group at all). The average result is 2.94 groups, so roughly about 3. The max is 5 groups, and the min is 1 group (obviously), and they are very rare (<5% for both 1 and 5, with the majority in 2 and 3, almost equally, slightly less for 4 groups compared to 2 and 3). I remember seeing pro games with 6 or even 7 groups before, but they are very very rare (IIRC, one is from SJS like years ago)

And another fun statistic for the pro games ended in scoring is roughly 15%, only about 1 in 6 to 1 in 7 pro games went into scoring (IIRC, it used to be higher, but with faster and faster time settings over the years, the ratio of resigns vs scoring began to increase as well).

6 Likes

This would be rare and worth noticing, as Counting_Zenist’s stats show. The proverb about group count is “5 can live but the 6th will die”. Winning with 6 groups will be rarer than having 6 groups and losing. Winning with 7 even rarer, I once managed that at OGS Honinbo 2008 (Main Class), Round 3 Group 1.

3 Likes

thank you! very valuable insights

Cho Chikun’s games will be worth watching. I remember that he always has a lot of groups.

2 Likes

I also checked 180 recent games (after 2017) from Cho Chikun, and the result is pretty interesting. First of all, his “group number distribution” has a “fatter” tail, that is, he has fewer games with 4 groups than the current average, but many more games with 5 groups (almost twice as many compared to the average). And the first game I saw that went into scoring with 6 groups (and won) against a player 20 years younger than him (and quite recently, there are games like Shin Jinseo vs Li Weiqing that SJS would have 6 groups, but ended in the opponent’s resignation before yose).

Another interesting thing about Cho Chikun’s games is that he went into scoring a lot often than the average, not just a little, it is almost 30% of his games when into scoring (twice the ratio compared to the average)

The 68th Oza preliminary - Cho Chikun (B) vs Suzuki Yoshimichi (W) 2019-07-04

And this is a very typical Cho Chikun games, with lots of groups live quite small or meandering and poking around like narrow snakes. When I search Cho Chikun’s game, it is quite recognizable which color he played by just look at the shape of his groups.

3 Likes

But what is noticable if both players have 7 groups?

Thank you, Counting_Zenist, once again you deliver a better, more precise, more informative answer than we could have imagined!

That would be even rarer, but just 1 player having 7 groups is already very rare.