ken d
Hall of Famer
Posts: 17,490
Joined: Dec 2013
Reputation: 1226
I Root For: college sports
Location: Raleigh
|
RE: Is the SEC on the decline?
(12-10-2015 10:43 PM)JRsec Wrote: (12-10-2015 10:17 PM)ken d Wrote: (12-10-2015 07:38 PM)JRsec Wrote: (12-10-2015 06:55 PM)ken d Wrote: (12-10-2015 08:45 AM)quo vadis Wrote: The BCS system relied in part on computers, and computers are less biased than people (not necessarily more accurate, but less biased, two different things), so if anything, the SEC's titles won during the BCS era were at least as 'valid' as the poll-titles won before the BCS. Controversies about polls and rankings certainly didn't start in the BCS era.
You may not think Alabama deserved a rematch in 2011, but every ranking had them #2. The AP and Coaches polls both had them #2.
Bottom line is that the SEC did win 7 straight titles using the system all the conferences agreed to. Again, there is no precedent for that. The only other time a conference won as many as 3 titles in a row was the Big 10 from 1940 - 1942. It really had no precedent at all, nothing remotely close to it.
Heck, counting LSU in 2003, the SEC won 8 titles in 10 years. No other conference ever won more than 5 (SEC from late 50s to late 60s, Big 8 in the 1970s).
Computers, or to be more precise, the data they crunch, are not only less biased, they are more consistent from year to year. They don't care about "eye tests", which may be the most biased way of evaluating teams and conferences this side of preseason polls.
Performance on the field certainly matters the most. And during the recent SEC dominance, their on field performance has been statistically dominant. When that happens in a single season, it could just reflect the normal variation that is always present in all data. But when it happens repeatedly over time, it tends to reflect reality. The output of the "computers", or the algorithms those computers employ, tend to be pretty reliable.
Though their run started before 2009, I only have the data that far back. During that time, the SEC's performance, as measured by the data, was very consistent until 2014. That year, their performance spiked to its highest level of the 7 year period. But in 2015 it dipped sharply, to its lowest level. The average of those two seasons is right in line with the average of the previous 5 years.
What was remarkable to me was how consistent the SEC's performance was from year to year during their dominant stretch. All the other P5's besides the B1G had a much wider range between their strongest and weakest years. The PAC and B12 had the greatest variability, with the ACC in the middle.
In any given year, you can argue whether one specific team or another "deserved" to be picked for a bowl or a playoff. That's a large part of the fun of sports. But as for overall performance by conferences over a significant period of time, the data are pretty compelling. The SEC has been significantly better than everybody else for the past decade.
If we want to know if that will continue, we'll have to wait to see the data after the fact. Because there simply is no reliable data that let's us predict the future (no matter what recruiting services would like us to believe).
When the computers program themselves I will accept this. Right now they are programmed by people. There will always be bias. Furthermore why even use a computer, a committee, or a poll. Just structure it to be decided by the play on the field. End of story.
Are you suggesting that programmers want their ratings to specifically make the SEC look better than they should? Better than other conferences? And not just one programmer, but lots of them? That can't be a result of unintentional bias. That requires deliberate planning. I can't imagine a conspiracy of that magnitude being undetectable.
And how do you propose to decide how 128 teams should be ranked on the field, and over multiple seasons? The whole point of using computer algorithms is that you can't get everybody to play everybody else enough times to "decide it on the field". Now if you are suggesting that we not use computers, polls or committees to decide who should be in a playoff, I agree with you 100%.
But that's not the issue of this thread. The issue is whether the SEC, all 14 teams, is in a state of decline from its recent dominance. I don't see how you decide that on the field.
Ken the various polls all had conferences that they favored and they were different conferences. And its pretty obvious that 128 schools don't belong in the same division when some of them have resources that approach 200 million a year and others don't have 20 million to work with. 4 P conferences where play is confined to the conference would each determine on the field a champion. The four champions then play for the title. We need two divisions within the FBS. They both can crown a champ and do so with on field play only. There are breaks in the levels of revenue generated about the 60th position, again at the 65, and again between the 71st and 72nd. It doesn't matter what point you choose but the differences are stark.
It's time to acknowledge the differences and get on with it.
Now as to the tie in to the SEC, to imply it is in decline is to imply that other conferences are stronger. If we move to a P4 with a playoff that debate will cease. The best of each will play and the outcome will be evident every year. The other bowls if finally set up with some logic can match schools that finished in the same position within each conference and that too would then provide a degree of direct competition from which a better assumption at which might be stronger could arise.
The problem with the bowls right now is that somebody's #3 will play somebody else's #5 and so on and so forth based on tie ins, who played in that bowl last year, and what rivalries might be renewed, and most importantly which fan base travels better. Because of this bowls are specious at best.
I say if you want to have a great bowl season you will take the champs for the play off and then pair off each conferences runner up, each conferences #3, all the way down to the 5th or 6th position. Those results would be a better indicator of conference strength when taken in total than anything we do now and that may be the only way to make such a determination on field.
I think it would spur the interest in the bowls and provide much better match ups than the garbage we get now. And it also gives the participants a little more incentive than the second runner up of a major conference getting stuck with Small School U's opportunity to tag a big boy. While the latter is interesting the conclusions that can be taken away from it will seldom be reliable. But to play a school from a rival conference with a little conference pride on the line might be a lot better motivation to show up.
You say that various polls favored different conferences. That may or may not be true. But if you are saying that some of those were not really polls, but rather computer algorithms, then I'd need to see some pretty strong evidence of that.
Do you think that Sagarin, Massey, Colley, Billingley or Wolfe have a conference specific bias in their algorithms? And if not, would that not suggest that the AP and Coaches' polls, which produce rankings very similar to those algorithms, aren't biased either (at least not enough to matter)?
Again, I'm not suggesting that they be a primary determinant of who plays where in the postseason. I personally have a strong bias toward conference champions. But at some point, even some champions are going to have to be left out, and it's just not practical in football to determine which those are on the field without some type of seeding. And seeding requires applying some form of judgment, whether it's a committee, a poll or a computer algorithm.
As for avoiding that dilemma by subdividing the FBS to produce four 16-team "high resource" conferences, I believe that, aside from some fans, there is no stomach for forcing schools to either associate with others to create symmetry or be left out of the game. I also believe that if a majority of the P5 schools were to unite to try to force that to happen, they would invite successful legislation at both state and federal levels that would prevent it from happening in my lifetime.
I doubt you could find 64 schools among the current P5 members that would be willing to be part of such a division. More likely, you might get 32. And if the number you do get gets small enough, the NCAA now has enough leverage to expel them and exclude them from all other championships including basketball. That might not mean much to an Alabama or Auburn, but I bet it does to Kentucky.
The idea of a symmetrical super division sounds great in theory, but I just don't believe it is possible. And if it's not, then we all just need to learn to accept imperfection and asymmetry, because it's the best we can do.
(This post was last modified: 12-11-2015 11:16 AM by ken d.)
|
|