tag:blogger.com,1999:blog-166382609543481231.post7984631793944726622..comments2023-04-15T03:43:53.550-07:00Comments on College Football by the Numbers: Home Field Advantage - Stage 2Anonymoushttp://www.blogger.com/profile/09862656571382266427noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-166382609543481231.post-29896468850926988032008-09-03T11:32:00.000-07:002008-09-03T11:32:00.000-07:00Just a question about your analysis. In controllin...Just a question about your analysis. In controlling for a team's performance at home, you had only one option: the team's road games. So I assume that the control sample was the team's road games (ignoring games at neutral sites, since they are rare). My question is: couldn't we, then, interpret this list as badness on the road, as well as goodness at home? Hawaii, for example, based on the distance they must travel, is notoriously bad on the road.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-166382609543481231.post-15057581940070978682009-09-03T06:16:03.143-07:002009-09-03T06:16:03.143-07:00Johnny,Very good question, and looking back on it,...Johnny,<br>Very good question, and looking back on it, I realize it might be a little misleading in the post. Here's the actual calculation in the case of Arkansas State: They did 19.24 points better at home than on the road. Their home opponents were 12.4 points stronger, but that is not taking home field advantage into account for these teams. If we assume that these teams were 7 points better at home than on the road, the real SOS difference would be 12.4-7=5.4. So we subtract 19.24-5.4 and divide by 2 and we get approximately 6.85. The HFA(PD) values above are more precise, because I use the real HFA value for each team instead of this approximate 7 point value, but the principle is the same. Does that answer your question?<br><br>ScottScott Albrechthttp://www.blogger.com/profile/09862656571382266427noreply@blogger.com