In this article, I'll be showing you the preliminary compu-picks 2011 top 25, comparing
it to the public rankings, 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 Dodds, SI Staples, and others.
| 1 || Oklahoma || 16 || TCU |
| 2 || Bama || 17 || Notre Dame |
| 3 || LSU || 18 || Miss St |
| 4 || Oregon || 19 || Auburn |
| 5 || Florida St || 20 || VA Tech |
| 6 || Ohio St || 21 || ASU |
| 7 || Boise St || 22 || Missouri |
| 8 || OK St || 23 || Florida |
| 9 || Stanford || 24 || West Virginia |
| 10 || Texas A&M || 25 || Texas |
| 11 || Arkansas || 26 || Georgia |
| 12 || South Carolina || 27 || Penn St |
| 13 || Nebraska || 28 || USC |
| 14 || Wisconsin || 29 || BYU |
| 15 || Michigan St || 30 || Utah |
And now below is the compu-picks top 25, as well as two notable omissions:
| Rank || Team || Rating || Public Rank |
| 1 || Oregon || 0.74 || 4 |
| 2 || Ohio State * || 0.68 || 6 |
| 3 || Alabama || 0.64 || 2 |
| 4 || Oklahoma || 0.57 || 1 |
| 5 || Boise State || 0.55 || 7 |
| 6 || Auburn || 0.54 || 19 |
| 7 || South Carolina || 0.54 || 12 |
| 8 || Arkansas || 0.52 || 11 |
| 9 || Stanford || 0.51 || 9 |
| 10 || Florida State || 0.49 || 5 |
| 11 || Texas Christian || 0.49 || 16 |
| 12 || Louisiana State || 0.47 || 3 |
| 13 || West Virginia || 0.47 || 24 |
| 14 || Notre Dame || 0.45 || 17 |
| 15 || Oklahoma State || 0.44 || 8 |
| 16 || Texas A&M || 0.41 || 10 |
| 17 || Arizona State || 0.41 || 21 |
| 18 || Southern California || 0.39 || 28 |
| 19 || Virginia Tech || 0.37 || 20 |
| 20 || Nebraska || 0.35 || 13 |
| 21 || Utah || 0.34 || 20 |
| 22 || Missouri || 0.33 || 22 |
| 23 || Texas || 0.32 || 25 |
| 24 || North Carolina State || 0.30 || NR |
| 25 || Mississippi State || 0.29 || 18 |
| 29 || Michigan State || 0.27 || 15 |
| 40 || Wisconsin || 0.15 || 14 |
1) Ohio St's surprisingly high rating is one of the most obvious things on the table. For now, I'm going to punt on this rating (as shown by the asterisk).
IF Tresselgate had never happened (and for some reason all those players had decided to return anyway), Ohio St would have been a clear top 5
(and probably top 2) team. Instead it's going to be much tougher to project how they'll do, especially with Tressel's ultimate punishment
still very much up in the air (and even more so with the late-breaking car dealership story).
For now I've made a minor adjustment due to this situation, but there's a good chance that by the time
I post numbers over the summer, that Ohio St's rating will decline, perhaps even materially. In other words, I am NOT going to stand
behind the projection (at least not right now).
2) The rating most personally surprising to me was Auburn's. Even though the "public" rating had them 19th, that was skewed by a couple very 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 so much 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).
They're also probably being temporarily helped by the fact that I don't have numbers for returning lettermen or starts lost to injury;
I would presume that they're going to get dinged by the lettermen
numbers, and that'll hurt a bit (though I'd be surprised to see it make a major impact).
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 fairly 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 skeptical, 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.
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 3 (much less #1, as some have) is a bit of a stretch.
4) Oregon at #1 isn't a huge difference compared to the public perception (which has them at #4), but it's definitely higher.
The first, and most obvious, thing in Oregon's favor is that basically lost nothing to the draft. Casey Matthews in the fourth round, and that was it.
Compared to other top-shelf teams, or even Oregon's standard draft losses, that's a major outlier on the low side.
This means that, while they do lose a relatively large number of starters, virtually all of their key players return.
Moreover, they've been recruiting well, and trending positively, which is another positive sign.
Of course, there are negative signs as well. 2010 was obviously an outlier for them; they're normally not nearly THAT good.
Their +13 turnover margin tied with 2005 as their most (by a substantial margin) over the last eight years. And they do lose
a number of starters, which means that they will have plenty of holes to fill. Nonetheless, it seems very likely that Oregon
will once again be an extremely good football team in 2011, and have a very realistic chance to return to the title game, and perhaps
pull out the win this time.
5) And on the flip side, 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).
6) West Virginia is a somewhat interesting case. They've got a very awkward off-field situation, where they've hired a guy to be a one-year
coach in waiting before stepping in to replace the current coach, who is effectively being fired as of the end of the season. Given how uncommon
this situation is, I haven't made any adjustment for it, since they won't actually replace the coach until next year. That said, it's quite
possible that this arrangement will crash and burn.
Nevertheless, there are a number of positive indicators for this team, who ought to be the best in the Big East. They had low draft losses,
they actually suffered their worst turnover margin (-5) in a number of years,
they improved as 2010 went along, their recruiting is on an upward trend, and 2010 still registers as a bit of a low outlier
(though given that it's the third straight "meh" season, maybe it shouldn't). So under normal circumstances, you'd expect substantial
improvement from this team. Whether it actually happens in 2011, though, will be an interesting case study in the effect of the one-year
forced apprenticeship system.
7) 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.
8) 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.
There are a few important notes and caveats I need to make about this model:
1) This 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.
2) 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.
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 South Carolina to be any better than Arkansas, much less a lot better.
3) The model is still missing some information. I'm waiting on numbers for returning lettermen and starts lost due to injury, and there are a few other pieces of
information that I may try to bring into the model if possible. Moreover, there are a few fields, such as returning starters, that may change over the offseason
due to injuries and/or players leaving their programs.
4) There are a few model changes that I may still implement; both this methodology and the overall compu-picks methodology should be pretty close to final,
but neither has not been finalized yet.
2010 Compu-Picks Blog
Questions, comments or suggestions? Email me at firstname.lastname@example.org