| |
BGonline.org Forums
Suggested Idea for Performance Grading
Posted By: mamabear In Response To: Suggested Idea for Performance Grading (Timothy Chow)
Date: Sunday, 5 October 2014, at 8:34 p.m.
If this is supposed to be an argument against VRR and in favor of PR...
Egad, no! It doesn't mean that I'm carrying water for PR, just because I think there are issues with VRR, and that's not what I said in the post above yours. Five demerits for false dichotomy!
Digressing a bit, I don't think these PR, VRR or any other rankings will draw new people to the game; I think they will either have no effect on new people being attracted to BG, or if mishandled, will have the opposite effect. All these arcane whatzi are strictly for the amusement of the cognoscenti.
My argument is only that we can't consider VRR to somehow solve our problems with PR, or anything else. If anyone feels like measuring it, fine, but I have no interest in rankings based on it, and I sure don't hear much clamor for them from anywhere else in BG-land, either.
To take an extreme example, a bot that has luck completely backwards, regarding lucky rolls as unlucky and vice versa, will still report a VRR that averages to the correct value in the long run...
I realize this is just an example you are using, but to me the significant issue is that one player who tends to play more games of types the bot misevaluates at the lower ply, where it now evaluates the luck, will be affected for good or ill more than one who doesn't. That will never converge until the bot is made perfect--and then VRR will work fine. But then, in that case so will PR. Then in both cases, they would both measure very precisely whatever they are designed to measure. We'd still have to figure out exactly what that signified.
An example regarding the last sentence of the above paragraph: Player A is very steady, and will post similar PRs when ahead, behind, or playing even in a match. Player B gets depressed when losing, and plays very poorly, missing whatever opportunities come his way to turn the match around, but plays fine when ahead. Player C gets overconfident when ahead, and plays like a happy monkey, acting as though the win has already been scored, but pulls himself together and does the best he can when losing. If all we have to evaluate are 10 winning matches, or 10 losing matches, for all three of these players, we will probably get a good handle on Player A's strength, but we won't get a good picture of the overall strength of Players B and C whether PR or VRR is used.
That's not an argument against VRR or PR, really. I think it's more an argument in favor of including both some wins and some losses in everybody's sample of 10 for this ranking system, however the camel ends up designed by the committee.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.