| |
BGonline.org Forums
Cube error rates
Posted By: Fabrice Liardet In Response To: Cube error rates (Gregg Cattanach)
Date: Thursday, 23 July 2009, at 1:58 p.m.
Right or wrong, greater than 100% is not some 'calculation error', it is just the results of that formula.
That is exactly my point, and what that means is that the formula does not capture the information it says it captures.
I mean that "player B was a xx% favourite" sounds like a well-defined claim about the player's equity, where xx could in principle be computed accurately by the bot - under the assumption that the bot plays perfectly. But the naive Snowie calculation does not do the job. What would be the right way to compute it ? Would it require additional info such as the average length of a match ?
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.