pentejr: Your opponent probably had more completed games than you. Until you accumulate a certain number of completed games, your ratings adjustment after each game is modified (increased) by a factor that declines as you gain experience. Since you have completed only 42 games of hypergammon, your ratings adjustment is greater than it would be if you had already achieved the level of experience at which the extra factor ceases to apply. Your opponent was probably closer to that level than you, and therefore had a smaller adjustment.

alanback: I figured it was something like that. But this is the key to the ratings inflation you find in the more frequently-played games on this site, is it not?

pentejr: I don't think so. It is rather considered smart than new people have a higher coefficient than people who have played a lot of games, and it could even compensate an inflation caused by people entering the rating list with a too high score. When there really is inflation it is likely to be caused by a too generous policy regarding first ratings, when the new overrated players inflate the rating pool.

nabla: I would expect the average ratings displayed on the boards to rise steadily, because a larger proportion of successful players than unsuccessful ones tend to stay active.

alanback: True, that is also a possible factor.

alanback: That would explain why the active ratings tend to slowly rise, but I believe (but cannot prove) that the average rating rises slowly over time.