| |
BGonline.org Forums
Suggested Idea for Performance Grading
Posted By: Timothy Chow In Response To: Suggested Idea for Performance Grading (David Rennie)
Date: Sunday, 5 October 2014, at 6:14 p.m.
I happen to agree with most of your objections to the "cult of PR." However, I don't think that Rick's proposed system is such a bad thing. Yes, they are "bot titles" and not real titles. So what? If they incentivize people then they are probably good for the game.
Phil and Iancho and perhaps you seem convinced that this sort of thing won't draw new players into the game. I'm not so convinced. I suspect that it will draw in the same kind of people who are drawn into the chess world. Now, maybe we don't want those people—maybe there aren't enough of them to matter, or maybe what we really want are fish who are deluded about how good they are, or fish who know they're fish but get entertainment out of gambling.
But all this is a separate question from how close to perfect XG is. I think that XG is close enough to the equilibrium strategy that "bot titles" have some meaning. Snowie wasn't quite there and it was recognized that it wasn't quite there although it was close. XG still isn't quite there for some types of games but they don't come up that often in practice.
I think your estimate of 4-5 PR for "optimal" play is pretty far off unless you think that the best way to beat a weak player is to immediately go aggressively for an extreme backgame with 12 or more checkers back and count on accumulating enough of an edge over the course of a very long game to compensate for what XG thinks are super-whoppers. This doesn't seem to be how any of the most successful players play against weak players, though.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.