tag:blogger.com,1999:blog-38600807.post2470074960267034938..comments2009-05-05T11:40:31.905-04:00Brian Burkenoreply@blogger.comBlogger5125tag:blogger.com,1999:blog-38600807.post-70720339219777911522009-05-05T11:40:00.000-04:002009-05-05T11:40:00.000-04:00Regarding factoring in team strength. One big problem I've noticed looking at football (soccer) numbers, I think this is something you've touched on before Brian with home-team advantage, is that the further into a game you get, the less 'team strength' (as defined by previous games) affects the win probability.<br /><br />I'm not sure how easy the drop-off effect would be to analyse, or indeed whether it's worth looking at (most interest is in the end-game WP and those are pretty much 100% dictated by the match situation).Mark Mhttp://www.blogger.com/profile/00437341898224311426noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-2317790224822590702009-05-02T12:22:00.000-04:002009-05-02T12:22:00.000-04:00For now there's no home ice or team strength factored in. So yes, any tied game will be 50/50.Brian Burkehttp://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-68448801692709123062009-05-02T01:57:00.000-04:002009-05-02T01:57:00.000-04:00Wouldn't the Poisson distribution and goals scored per minute need to be based upon the average of each individual team, rather than the NHL average?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-87828049243148388012009-05-02T01:05:00.000-04:002009-05-02T01:05:00.000-04:00It looks as if New Jersey had a 97-98% chance of winning up one with 1:30 left...<br /><br />Brian, I assume the probability is simply 50% when tied?Zachnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-65420601513993118412009-05-01T14:08:00.000-04:002009-05-01T14:08:00.000-04:00what were the chances the devils would lose that game (and the series) a goal up with 1:30 to play?Anonymousnoreply@blogger.com