The rating system and me

[gennan]

@oddlyenuf

If a player only plays one or two weaker players and always wins will that player reach dan level or cap out just above the weaker players?

As the rating gap grows, the winner will gain fewer an fewer points from each win. In practice this means their rating will cap out, similar to this graph

@su2

[oddlyenuf]

Thanks. Not quite like a guy buying a black belt and claiming expert status but close enough.

[gennan]

Slightly more accurately, I’d like to add that if that dominating player literally only plays against those same weaker players and never loses, the winner’s rating will grow indefinitely, but it will be more like a logarithm:

I deliberately left out numbers in that diagram, because those would depend on various (scaling) parameters of the rating system.

But this kind of situation is purely hypothetical. I don’t think it makes much sense to have a rating system with only 2 players of which 1 always wins.

And if we have such a pair of players of which 1 always wins, we can never find out how big the rating gap should be from even games only. To measure the skill gap in that situation, we would need to determine the handicap that allows the weaker player to win about 50% of the games.