tag:blogger.com,1999:blog-2007702193183659560.post233704167223561865..comments2014-04-06T02:43:07.306-04:00Comments on Miami Heat Index: Raptors vs. Heat: Miami Favored To End Losing Streak With Extinction Level Eventreservoirgodhttp://www.blogger.com/profile/16673328205522511885noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-2007702193183659560.post-551190941369350662010-11-15T14:20:56.989-05:002010-11-15T14:20:56.989-05:00Yeah, on the one hand that sucks - but on the othe...Yeah, on the one hand that sucks - but on the other hand, I really want them to dominate. But it may be better to set proper expectations. Thanks for the comment.reservoirgodhttp://www.blogger.com/profile/16673328205522511885noreply@blogger.comtag:blogger.com,1999:blog-2007702193183659560.post-14646097724845233412010-11-15T14:03:57.697-05:002010-11-15T14:03:57.697-05:00Just a quick comment on your win probabilities. T...Just a quick comment on your win probabilities. The formula you're using is based on a linear regression that's not really designed to handle predictions when one team is a lot better than the other. For example, using that formula, if the home team were expected (by WP) to win 82 games, and its opponent were expected to win 0, you would get a (nonsensical) win probability of 160.6%. Given that the Heat are a lot better (in WP anyway) than just about every team they face (think of all the Wizards-Heat matchups where that's basically the situation), that means you're giving probabilities almost exclusively in the part of the curve where this model doesn't work very well. <br /><br />If you were posting these numbers for an average team, they would probably be pretty accurate. For the Heat, you're going to see a lot more losses than these probabilities suggest (even if it's not a lot in the grand scheme of things).Kennethhttp://www.blogger.com/profile/16020039547788588924noreply@blogger.com