In this article, I'll be showing you the final compu-picks 2011 top 25, comparing
it to the public rankings, displaying the actual projected ratings,
and highlighting a few key results and caveats.
First, though, is a table showing what the actual public top 30 is.
I've created this table as a composite of a number of publicly available
rankings, including ESPN Schlabach (multiple editions), CFN (multiple), CBS Dodd (multiple), SI Staples (multiple), preseason magazines,
sportsbook national title odds, and others.
| 1 || Oklahoma || 16 || Wisconsin |
| 2 || Bama || 17 || VA Tech |
| 3 || Oregon || 18 || Michigan St |
| 4 || LSU || 19 || Auburn |
| 5 || Florida St || 20 || West Virginia |
| 6 || Boise St || 21 || Missouri |
| 7 || Stanford || 22 || Miss St |
| 8 || Ohio St || 23 || Florida |
| 9 || OK St || 24 || ASU |
| 10 || Texas A&M || 25 || Georgia |
| 11 || Arkansas || 26 || Texas |
| 12 || South Carolina || 27 || USC |
| 13 || Nebraska || 28 || Penn St |
| 14 || TCU || 29 || Utah |
| 15 || Notre Dame || 30 || Miami |
And now below is the compu-picks top 25, as well as two notable omissions:
| Rank || Team || Rating || Public Rank |
| 1 || Alabama || 0.64 || 2 |
| 2 || Oregon || 0.61 || 3 |
| 3 || South Carolina || 0.58 || 12 |
| 4 || Oklahoma || 0.56 || 1 |
| 5 || Boise State || 0.56 || 6 |
| 6 || Florida State || 0.55 || 5 |
| 7 || Arkansas || 0.53 || 11 |
| 8 || Ohio State || 0.52 || 8 |
| 9 || Stanford || 0.50 || 7 |
| 10 || Notre Dame || 0.49 || 15 |
| 11 || Louisiana State || 0.49 || 4 |
| 12 || Texas Christian || 0.47 || 14 |
| 13 || Texas A&M || 0.42 || 10 |
| 14 || Brigham Young || 0.40 || NR |
| 15 || Auburn || 0.38 || 19 |
| 16 || Oklahoma State || 0.38 || 9 |
| 17 || Nebraska || 0.36 || 13 |
| 18 || Tennessee || 0.36 || NR |
| 19 || Missouri || 0.35 || 21 |
| 20 || Virginia Tech || 0.34 || 17 |
| 21 || Arizona State || 0.34 || 24 |
| 22 || West Virginia || 0.34 || 20 |
| 23 || Southern California || 0.33 || 27 |
| 24 || Utah || 0.33 || 29 |
| 25 || UCLA || 0.30 || NR |
| 29 || Michigan State || 0.27 || 18 |
| 31 || Wisconsin || 0.25 || 16 |
1) One interesing note this year is that there isn't anything close to an overwhelming preseason favorite, like Florida in 2009 or Alabama in 2010 (both teams
failed to win the title, but both were VERY good teams nonetheless). And, while Oklahoma has been a popular media pick for #1, there's been enough variation in
votes, including a number of sources that had them 3rd or 4th, that I think this perspective is reflected in popular opinion. Not only is there no clear favorite,
there are no clear examples of teams that you KNOW are going to be really good. If I told you that Oklahoma would go 8-4, would that stun you? How about Alabama?
Oregon? Really, other than Boise (mainly due to a soft schedule), I can't think of a single team that's close to guaranteed to hit 10+ wins. That should make this
an interesting, potentially upset-filled season. I can't wait.
2) It's not a huge surprise that there are two SEC teams in the Compu-Picks top three, but it IS a surprise which those two are. Bama is expected,
of course, but South Carolina is most certainly not. But, as noted in their league preview page,
they have a number of quite positive signs. They've been recruiting at a very high level (and getting much better lately), they have a slightly
high number of starters returning, they have virtually no meaningful draft losses, and while their 2010 season might have seemed like a lucky
outlier to some, they did it while suffering a very high 34 starts lost to injury and a net zero turnover margin (quite low for a top team, and since
turnovers are fairly random, this suggests a better number in 2011).
So while "they're just not usually this good" points towards decline, almost every other sign points to continued improvement.
Compu-Picks may well be going out on a limb here (and if things turn sour with QB Garcia then almost certainly), but there are serious
indicators that the general public is seriously underrating this team.
3) LSU is another very noticeable result from the model. Virtually everyone seems to be assuming that LSU is going to be a top five team,
and that may be a little bit of a stretch. There's plenty of evidence to think that they'll be very good: they've been very good the past few years;
they return a decent number of starters, including their QB; their recruiting has been good and on an upward trend; and their turnover margin was
pretty average compared to most other teams projected to be every good.
On the other hand, there are a few negatives, though none of them are huge. Their draft losses were fairly substantial. Their recruiting momentum,
while positive, was materially less than a number of other top-rated teams. And while most other top-rated teams registered material improvement over
the course of 2010, LSU really didn't; they won their bowl game, but lost to Arkansas the game before, and struggled against Ole Miss the game before that.
Of course, LSU is still projected to be a very good team, and they're certainly within striking distance of the top five, but the system thinks that putting them
in the top 5 (much less #1, as The Sporting News and Dennis Dodd [in January] did) is a bit of a stretch.
4) Another low projection is Oklahoma. Compu-Picks has Oklahoma #4 compared to the pundits' pick of #1. There are various positive and negative
indicators for the Sooners, but it mainly boils down to the fact that in 2010, they were simply a few steps below the other top-shelf programs.
They are projected to improve (thanks to recruiting, low draft losses, a positive performance trend in 2010, and a fairly high number of returning starters),
but are not projected to improve by enough to leapfrog the three ahead of them (partially due the sheer size of the gap, partially due to their
high 2010 turnover margin and various neutral indicators).
5) BYU is one of Compu-Picks highest picks compared to the consensus. What's actually surprising to me is that pretty much nobody (except Compu-Picks and Jon Wilner's January poll)
had them in the top 25. While last year was a major downer, this has been a pretty consistently strong program that recruits well (and getting better),
lost absolutely nothing to the NFL draft, had a brutal season of injuries (43 starts lost is very high), and in almost every way seems strongly likely
to rebound in a major way. Even if Compu-Picks is a bit too high on them (which is what I think), they really should be projected as a top 25 team.
6) Tennessee is one of Compu-Picks' most interesting optimistic projections. Unlike BYU, they haven't been at the same high level consistently (though they also weren't as bad in 2010).
But like BYU, they recruit very well (and getting better), they suffered very low draft losses, they had a nasty injury year in 2010, and they return a solid amount of experience.
Also, unique for Tennessee, they'll have coaching staff stability for the first time in years. That ought to be worth something too. While I personally think 18th is a bit of a reach,
I do think that the Vols at least have the potential to be a serious sleeper in the East (and since Compu-Picks is, as far as I can tell, the ONLY source to have the Vols ranked,
I think sleeper is clearly appropriate here).
7) UCLA, another team Compu-Picks is high on, basically boils down to recruiting, returning experience, and brutal 2010 luck in both turnovers and injuries. If that luck can rebound, UCLA
could very well be a serious contender in the Pac-12 South.
8) Auburn is an interesting case. Even though the "public" rating had them 18th, that was skewed by a couple very early high ratings
of the Tigers by people who thought that Newton and/or Fairley might return. So the real public perception of them is even lower.
So why does compu-picks like them in 2011, even after the draft losses and REALLY low number of returning starters (only 7; 3 off, 4 def)?
First of all, their draft losses, while material, weren't anywhere near the very top of the list. They lost two very highly regarded players,
and then only two other guys, each in the 7th round. That's much less than Bama (FOUR first rounders) or UNC (one first, three second, and a few more
scattered through the draft), and fairly comparable to a number of major programs, including LSU, Miami, Georgia, etc. The draft losses
were high enough to be a negative, but not high enough to be a huge negative.
On the other hand, they've been recruiting extremely well AND have been improving their recruiting. They've got a LOT of talent with which to fill
in the holes left behind by their departing starters, including star running back Michael Dyer. Their turnover margin, while a bit on the fortunate
side, wasn't a huge outlier and was actually less than what most other top teams put up. Moreover, they improved substantially as 2010 went along,
which is a good sign for 2011, even though a number of starters are gone (among other things, that usually suggests that the younger players
were successful in working their way into the system and becoming better able to fill in the holes created by departing players).
Overall, while Auburn's situation (very low returning starters without especially heavy draft losses) is fairly unusual,
it's far from unheard of. Over the last three years, there have been a number of somewhat similar examples among
top 50 programs: 2010 Arizona St (9 starters, no QB); 2009 Miss St (10 starters, no QB), TCU (10 starters, with QB);
2008 Georgia Tech (9 starters, no QB), Miami (9 starters, no QB), Oregon St (10 starters, with QB), Texas A&M (10 starters, with QB).
None of them had as few returning starters as Auburn, but overall the scale was fairly similar.
And by and large, they ended up
all right. 2008 A&M collapsed, but generally the teams ended up reasonably comparable to how they'd been the previous year, with a couple
(most notably Georgia Tech) actually improving. And it should be pointed out that in 2010, there was
a previously high-flying team that had only eight returning starters, that suffered their biggest draft losses in the last 10 years, and that everyone
simply assumed was in for a major rebuilding season. That team was 2010 Oklahoma St, and they ended up with a quite respectable
season, including a dominating win in the Alamo Bowl.
Does all that mean that Auburn is going to seriously compete for an SEC or national title?
I'm certainly a sketpic of going that far, but I also think it's safe to say that anyone who's projecting Auburn to go something like
3-5 or worse in the SEC is very likely in for a seriously rude awakening.
9) Wisconsin's exceptionally low projection is a function of a number of negative indicators.
They had substantial draft losses, among the highest in the country (less than Bama and UNC but comparable or higher than all other programs),
and the most they've suffered over the last decade.
They also benefited in 2010 from a +14 turnover margin, their best over the last eight years (2005's +13 came close, but otherwise they've
generally been far worse), and far higher than normal for even good programs;
it is highly likely that their 2011 margin will return to around zero, which is a major negative sign for the upcoming season.
Moreover, 2010 was a fairly material outlier; teams that make big jumps forward have a tendency to slide back to some extent.
Overall, the data available strongly suggests that Wisconsin is likely going to be in for a fairly material decline.
10) Michigan St is a much different case. For the Spartans, most of the available indicators suggest holding about steady, not decline.
12 starters is a bit low, their +5 turnover margin was on the fortunate side, and 2010 was better than usual
(not a major outlier, but the numbers also account for a bad 2009 campaign).
On the other hand, they return their quarterback, they had pretty low draft losses, and their recruiting has been trending positive.
So why does compu-picks not particularly like them compared to the public perception? Mainly it's because it didn't think they were all that
good last year. They got completely destroyed twice, and they were pretty fortunate in close games; 4-0 in games decided by single digits,
and 3-0 in games decided by less than a touchdown (including a trick play overtime win), is unlikely to repeat itself. Overall, while the Spartans
shouldn't be a worse team, the record is a whole other story.
There are a few important notes and caveats I need to make about this model:
1) Compu-Picks does not endorse implicitly or explicitly any form of illegal gambling.
Compu-Picks is intended to be used for entertainment purposes only.
2) No guarantee or warranty is offered or implied by Compu-Picks for any information provided and/or predictions made.
3) This preseason model is primarily based on the main compu-picks model. Essentially, it attempts to predict how well a team will rate given its rating history,
as well as a number of other data points, such as returning starters, draft talent lost, turnovers, recruiting, etc. This means, among other things, that the rankings
are power rankings based on how good a team projects to be, as opposed to a more cynical (though accurate) model that attempts to project how the BCS will rank a team
by making adjustments to favor those with easy schedules and punish those with tough schedules.
4) The top 25 here ties exactly to the list released a couple weeks ago. This article goes inot the numbers in more detail; it does NOT revise them based on subsequent
news, such as the Butch Davis firing, injury updates, and the Miami fiasco.
5) There is a substantial amount of noise in these projections, which is to be expected given the large number of unknowns (who will have good and bad luck with injuries,
which young players will improve and which won't, how specific matchups will come into play, etc.). Right now the standard error is a bit over 0.2 on a scale of about -1 to +1.
It's important to look at the projections with this in mind to get a sense of how material the projected differences are. Given a standard error around 0.2,
it is safe to project Alabama to be a much better team than Mississippi St, but it is not safe to project Arkansas to be any better than LSU, much less a lot better.
6) At this point, there are a number of model features that need to be investigated further. Chief among these is the distribution of extreme events.
It appears that the model may be overstating the probabilities of extreme events, such as 12-0 or 0-12 records, or major underdogs winning their division/league.
Please keep this in mind when looking at the distribution of win probabilities.
2011 Compu-Picks Blog
Questions, comments or suggestions? Email me at email@example.com