Yucata is a turn based board gaming site with 70 different board games. They have two different types of rating systems which seem pretty arbitrary. I remember when I first showed the site to Robb I commented on how I cared about raising my overall level but that the individual game skill number didn't matter. He responded by saying the individual game skill number was relevant and the metagame level didn't matter...
In a way I think we were both right. Neither of the ratings really do anything at all. They're numbers which go up and down and which may have some positive correlation to skill but they aren't used in any real way. Every now and then you'll run into someone who started a public game with a note asking for only people of a certain level/rating to join but it isn't enforced by the system. (Personally I think it would be rude to go against those requests but I don't care enough to make any myself.) On rare occasions people might even send a private game challenge to someone with a high rating in a game. (I helped test Torres and the test games seemed to have counted for rating so when that game launched I was #2 on the ranking and got a couple random challenges from people I'd never heard of before.)
The idea of a number that represents how good you are at something is a very appealing one but the existence of such a number means some people will behave differently in an attempt to make their number higher. Not by getting better at the games but by being very careful who they play against and what games they play. A particularly recent example is with A Few Acres of Snow where you could create a game and guarantee you'd play a specific side. With the game being unbalanced and one side being practically guaranteed to win it's an easy way to inflate your rating numbers. I imagine the people at the very top of that leaderboard had never even played each other since they were all playing as Britain and only playing against the suckers who joined their public invites as France. (Personally I only played games with random sides except against Andrew where we played one game as each side.)
At the end of the day the 'trueskill' number for each game doesn't so much show how good you are at a game so much as it shows how good you are at making that number bigger. The metagame level doesn't really show how good you are at games in general so much as it shows how dedicated you are to playing games you're good at... Or that are fast! Let's just say there's a reason why I've played 980 games total on the site and 370 of them were Roll Through The Ages!
No comments:
Post a Comment