I recently played a game where I had almost as many captured pieces as territory:

http://online-go.com/game/625666

The game was closer than I expected, I think because capturing a lot of pieces lulls me into thinking I must be winning. But now I'm wondering: how rare is it to have such a low ratio of territory to prisoners? Are there games where the ratio is even less than one (say as a result of ko fights)?

One way in which this can happen, particularly in beginner games, is when hopeless invasions are made into the other player's territory, and the other player responds, or is forced to respond. So gradually the territory gets exchanged for captures. Or (as in this case) where one player tries to go for an overwhelming win, in a low risk/high reward frame of mind. I would be interested in seeing similar games, or receiving analysis of this game.

I think that in most games the number of captured stones would be fairly small compared to territory. I read somewhere a rough analysis to the effect that a typical go game lasts around 230 moves, leaving 130 total possible territory points, so a typical amount of territory is around 65 points. The upshot of that was that if you are playing and have about 60 points of certain territory you don't need to take risks to get more.

Anyhow, if the number of captured stones is large and the number of points is not a lot larger it suggests that there wasn't much territory except for what was created by actually capturing stones. It doesn't suggest a sophisticated level of play.

Sad but true.

Were you B or W ?
Also, it's easier if you post the SGF directly here.

I notice that pretty uniformly across games I win and lose that my opponent has typically made 10-15 more captures than I have.

_________________
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.

Or it suggest a very high level of play with many trades

Presumably you would have at least twice as many points as you had captures, unless for some reason you were forced to fill your territory afterwards. I supposed a low scoring game with a large number of ko fights is possible. Ko could generate a large number of captures without the respective territory, but even then you would still probably expect there to be at least 30-50 additional points per side beyond the captures.

If we are just looking at territory (and not total points), then a few ko fights and a big trade could get you over that 1:1 ratio.

I think this will depend on how you count 'points'. At the end of the game under territory scoring it is certainly possible to have very few open points left in the counting after using prisoners to fill in the opponent's territory. It is only by convention that we count prisoners as our points during a game. In the final counting they are negative points for the opponent. In games that go more than 360 moves we need some agreement to exchange prisoners or some counting of left over stones in order to determine the final result, no?

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

This got me wondering how long the longest pro games have gone. It took a bit of searching but this site seems to have some top contenders at 400+: http://web.archive.org/web/200810141808 ... ngest.html

Yamabe Toshiro - Hoshino Toshi, Oteai, 12/20/1950 (411 moves)


Fujisawa Kazunari - Kamimura Haruo, 53rd Honinbo preliminaries, 12/12/1996 (406 moves)


Yasui Sanchi - Ito Showa, 4/3/1839 (405 moves)


Found from http://senseis.xmp.net/?FamousGamesInvolvingKo

I'd say you can get lots of captures if you have a game with a long ko fight, even more so if many of the threats are plays within opposition territory.

_________________
Poka King of the south east.