Granularity

This concept has proved difficult to work out in any detail, and is still under discussion. You can read and expand the discussion on the ''Discussion'' sub-page — please do not add detail to this page while the discussion continues.

Granularity has been proposed as a measure of one aspect of the quality of rule sets, in particular for territory scoring versus area scoring. The essential idea is that it is desirable for rules to measure the result of play as accurately as possible, in the sense that there are more possible final scores, so that final positions that are equivalent under ‘coarse-grained’ rules may be scored differently under ‘fine-grained’ rules. Territory scoring appears at first sight to be finer-grained than area scoring because area scoring usually leads to an odd score difference (flipping the color of an intersection is a 2 pts area score change), while the score difference under territory scoring can be any whole number; this conclusion is, however controversial.

(A different concept of granularity is distinguishing the quality of moves, in the sense that there are fewer optimal moves in any given position, making the game more challenging to play well; of course distinguishing more final scores tends to decrease the number of optimal moves.)

See also:

• TerritoryAndAreaScoring#granularity
• TerritoryAndAreaScoring#komiandresults
• FreeTeireIfEvenDame
• RulesAndAreaAndTerritoryScoring#areakomi
• Taiwan Rule
• Button Go