tag:blogger.com,1999:blog-38600807.post4666825553439220253..comments2023-11-05T04:16:44.937-05:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): Efficiency Rankings - Week 10Unknownnoreply@blogger.comBlogger10125tag:blogger.com,1999:blog-38600807.post-80203510862993896992010-11-17T15:31:53.596-05:002010-11-17T15:31:53.596-05:00Thanks Brian, sorting is working great.
Also- I t...Thanks Brian, sorting is working great.<br /><br />Also- I tried to make a spreadsheet that replicates your Game Probabilities calculations based on your explanation here: http://www.advancednflstats.com/2009/01/how-model-works-detailed-example.html<br /><br />Although most of my answers are close to your published numbers on the Fifth Down, none of them are exact and I was wondering if you've changed the coefficients in your formula or if there is something else going on. The most egregious example is NYJ vs CLE, I kept getting NYJ as a 61%/39% favorite while you had CLE as a 53% favorite.<br /><br />Here are my calcs:<br />CLE Logit = 0.46*6.2+0.25*4.2+-19.4*0.035+-19.4*0.018+-0.62*6.7+-0.25*3.9+-1.53*0.4 = -2.8672<br />NYJ Logit = 0.46*6+0.25*4.7+-19.4*0.02+-19.4*0.017+-0.62*5.8+-0.25*3.3+-1.53*0.56 = -2.0606<br /><br />CLE GWP = e^(-.36+.72-2.8672+2.0606)/(1+e^(-.36+.72-2.8672+2.0606)) = 39%<br /><br />NYJ GWP e^(-.36-2.0606+2.8672)/(1+e^(-.36-2.0606+2.8672)) = 61%Davenoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-75572992313052379742010-11-11T14:59:34.074-05:002010-11-11T14:59:34.074-05:00Dave-Sorting should be working now.Dave-Sorting should be working now.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-61498864516092984832010-11-11T01:38:30.448-05:002010-11-11T01:38:30.448-05:00Brian,
I can imagine that the problem with yards ...Brian,<br /><br />I can imagine that the problem with yards per attempt is that some teams may get a 3 and out repeatedly, others might march halfway up the field and then punt repeatedly, and both still have 0 points, but the one which got halfway up the field will have a much higher efficiency. <br /><br />Have u ever thought of utilizing other variables which may account for such discrepancies. For example, I would think the number of punts or 4th downs per possesion (minus possessions ending in turnovers and end of halves)could be helpful. I was just curious about whether or not you have ever looked into the statistical significance of these stats.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-87867384977072517402010-11-10T21:34:20.266-05:002010-11-10T21:34:20.266-05:0049ers at #25 is the best in the NFC West. The oth...49ers at #25 is the best in the NFC West. The other 3 West teams make up 3 of the bottom 4 teams in the overall efficiency chart. <br /><br />One of these teams will make the playoffs and likely host a game against a superior NFC East wild card team, such as PHI or NYG. LOL.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-27963617530694710952010-11-10T17:13:29.391-05:002010-11-10T17:13:29.391-05:00David-Yes, and it's not too hard. Do a search ...David-Yes, and it's not too hard. Do a search for 'how the model works' in the site's search box above. That post should tell explain.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-48097150295426219572010-11-10T17:01:40.922-05:002010-11-10T17:01:40.922-05:00Never mind. I looked into it and it's a bit m...Never mind. I looked into it and it's a bit more complicated.Davenoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-81396202276331384842010-11-10T15:46:10.585-05:002010-11-10T15:46:10.585-05:00Isn't there any easy way to get to your weekly...Isn't there any easy way to get to your weekly Win Probabilities from this chart? Like you take the GWP's for each team and make an adjustment for home field advantage. <br /><br />If team A were at home, would it be:<br /><br />Team A Win Chances = (GWPa*(.57/.50)) / (GWPa*(.57/.50)+GWPb*(.43/.50))Davidnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-70947289070326591112010-11-10T12:03:49.357-05:002010-11-10T12:03:49.357-05:00...Touché Mr Burke......Touché Mr Burke...Davenoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-42381281931850238402010-11-10T11:40:26.619-05:002010-11-10T11:40:26.619-05:00That happens to me too, but only on Firefox on my ...That happens to me too, but only on Firefox on my work computer. On all my other computers, it works normally.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-30379587575431356462010-11-10T11:26:47.492-05:002010-11-10T11:26:47.492-05:00Good stuff. For some reason the headers only sort...Good stuff. For some reason the headers only sort Descending when I click on them. I can't figure out how to make them Ascending.Davenoreply@blogger.com