In Ira's report of the ACC meetings, football coaches including Jimbo were asking the CFP committee some tough questions about why FSU dropped despite being defending champions, having a pretty tough schedule, and remaining undefeated.
Like most on here, I saw the dropping of FSU out of the top spot and then all the way down to #4 as a real slap in the face. FSU was getting everyone's best game and still coming out unscathed.
What I heard a month or two ago softened my stance on the committee a bit. While I would still be willing to bet that the ACC's weak representation on the CFP committee did FSU no favors, I'm no longer in the camp that it was mostly on the ACC's weak reputation.
Because of my research, I had the occasion to talk pretty frankly with someone with first hand knowledge of the committee's ranking process. To summarize, they are using some pretty advanced analytics to determine strength of teams. FSU looked pretty ugly at times, particular in first halves of games last year, and performed only slightly better than some pedestrian opponents (regardless of how motivated said opponents might have been to beat FSU). As a result, FSU's wins "looked" like a series of very lucky events to the advanced, detailed analytics programs used by the CFP committee. In fact, there were times when FSU was ranked below 5th place in the second half of the season.
In the past, before the CFP, computer rankings were only concerned with outcomes - who wins, who loses, and the final score. Margin of victory was taken out of the computer portion of the BCS formula altogether after the uproar from coaches RUTS against teams. (Still recall that glorious first Bowden Bowl!). The momentum prior to the CFP was decidedly away from analytics...only wins and losses and SOS (based on opponent's and opponent-opponent's wins and losses) could be figured in.
This year, the CFP playoff committee was stuck between competing forces with FSU. The human polls had FSU in the top 2. The analytics they were using had FSU pretty consistently in the 5-10 range during the second half of the season. Does the CFP committee go with the traditional way of ranking teams (like the AP poll) based almost solely on whether they win or lose? Or, do they go with what the "state of the art" analytics are telling them FSU should actually be ranked? Based on what I heard, the analytics most revered by the committee had FSU ranked outside the top 4 for the final poll.
Losing by 39 points probably validated the analytics on FSU to many on the committee.
When the committee spoke of "game control" and other such criteria, much of it was based on the analytics provided. Like them or not, analytics are changing the face of sports...from Moneyball and Billy Beane to Daryl Morey, and now to CFB.
FSU has to play better for the entirety of games to get the benefit of the doubt in the future. It will also help for ACC teams to not crap the bed against mediocre-to-bad OOC opponents early in the season. One major reason that the analytics loved the SEC West is because it went 28-0 (and looked mostly dominating) in all its OOC games last year. Until the ACC can start coming close to this type of dominance OOC, and get another voice or two on the committee to protect its teams, FSU will have to have 2013 type dominating seasons to ensure a top-2 seed.
TIFWIW.
Like most on here, I saw the dropping of FSU out of the top spot and then all the way down to #4 as a real slap in the face. FSU was getting everyone's best game and still coming out unscathed.
What I heard a month or two ago softened my stance on the committee a bit. While I would still be willing to bet that the ACC's weak representation on the CFP committee did FSU no favors, I'm no longer in the camp that it was mostly on the ACC's weak reputation.
Because of my research, I had the occasion to talk pretty frankly with someone with first hand knowledge of the committee's ranking process. To summarize, they are using some pretty advanced analytics to determine strength of teams. FSU looked pretty ugly at times, particular in first halves of games last year, and performed only slightly better than some pedestrian opponents (regardless of how motivated said opponents might have been to beat FSU). As a result, FSU's wins "looked" like a series of very lucky events to the advanced, detailed analytics programs used by the CFP committee. In fact, there were times when FSU was ranked below 5th place in the second half of the season.
In the past, before the CFP, computer rankings were only concerned with outcomes - who wins, who loses, and the final score. Margin of victory was taken out of the computer portion of the BCS formula altogether after the uproar from coaches RUTS against teams. (Still recall that glorious first Bowden Bowl!). The momentum prior to the CFP was decidedly away from analytics...only wins and losses and SOS (based on opponent's and opponent-opponent's wins and losses) could be figured in.
This year, the CFP playoff committee was stuck between competing forces with FSU. The human polls had FSU in the top 2. The analytics they were using had FSU pretty consistently in the 5-10 range during the second half of the season. Does the CFP committee go with the traditional way of ranking teams (like the AP poll) based almost solely on whether they win or lose? Or, do they go with what the "state of the art" analytics are telling them FSU should actually be ranked? Based on what I heard, the analytics most revered by the committee had FSU ranked outside the top 4 for the final poll.
Losing by 39 points probably validated the analytics on FSU to many on the committee.
When the committee spoke of "game control" and other such criteria, much of it was based on the analytics provided. Like them or not, analytics are changing the face of sports...from Moneyball and Billy Beane to Daryl Morey, and now to CFB.
FSU has to play better for the entirety of games to get the benefit of the doubt in the future. It will also help for ACC teams to not crap the bed against mediocre-to-bad OOC opponents early in the season. One major reason that the analytics loved the SEC West is because it went 28-0 (and looked mostly dominating) in all its OOC games last year. Until the ACC can start coming close to this type of dominance OOC, and get another voice or two on the committee to protect its teams, FSU will have to have 2013 type dominating seasons to ensure a top-2 seed.
TIFWIW.
