tag:blogger.com,1999:blog-38600807.post3308454777512942011..comments2023-11-05T04:16:44.937-05:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): Roundup 2/6Unknownnoreply@blogger.comBlogger3125tag:blogger.com,1999:blog-38600807.post-29151035046333373832010-02-07T11:55:18.981-05:002010-02-07T11:55:18.981-05:00"I wouldn't expect regression to be that ..."I wouldn't expect regression to be that strong, at least among the top teams."<br /><br />I guess that depends on what you mean by "that strong". If you're thinking that top college teams don't become bottom-dwellers and vice versa, then yes, I agree with you. But if we accept that the top teams are essentially working from their own baseline, then I think it is pretty clear to see regression-to-the-mean taking place amongst them.<br /><br />The top teams probably have a mean of ~10 wins per season and, when they're "up" they get 12-13 wins and when they're "down" they get 7-8. After all, a 7-5 season is considered quite disastrous at a football powerhouse like USC or Florida. But the reinforcing feedback loop of college football's recruiting system ensures that these top teams rarely ever fall completely out of that top tier.<br /><br />I've also noticed that college football's regression tends to follow longer time cycles. A team like Alabama that won it all this year will almost certainly NOT regress back to, say, 8-4 next season. But it's not hard at all to imagine that they could be in such a situation in 3-4 years.<br /><br />A recent example that comes to mind is LSU. They had several several consecutive championship-caliber seasons. Now they are, by LSU standards, in the doldrums. In 3-4 seasons they could be right back at the championship level. But they never really get "bad". They never fall out of that top tier of college teams.<br /><br />This probably makes sense because most impact players in college are starting for 2-4 years and a program's surge could be defined by having 2-or-more consecutive superior recruiting classes. This means that top programs probably experience windows of 2-4 years when they have their best shot at winning a title.bytebodgerhttps://www.blogger.com/profile/15989876051555196561noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-81507220262233668412010-02-06T19:56:41.857-05:002010-02-06T19:56:41.857-05:00Speaking of interceptions, have you run a comparis...Speaking of interceptions, have you run a comparison of runs and passes using your new baseline for the value of an interception? I noticed you're still using 45 instead of 60.John Morganhttps://www.blogger.com/profile/06396900809037959125noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-73207671486105042892010-02-06T12:09:01.439-05:002010-02-06T12:09:01.439-05:00Interesting note from the PFR league totals. . . i...Interesting note from the PFR league totals. . . it looks like the number of interceptions per game has dropped substantially, while the number of pass attempts has increased. Could this help explain the current over-reliance on the running game? Perhaps a holdover from an era in which passing was rightfully considered to be a much riskier option?Brianhttps://www.blogger.com/profile/08130926015265781983noreply@blogger.com