Brian Fremeau looks back at some of his favorite mistakes from 2013, and unveils the final preseason FEI ratings.
18 Oct 2012
by Bill Connelly
We are seven weeks into the 2012 college football season, and those of you who have read FO for a while know that, after seven weeks, the preseason projections are all phased out of the numbers. What you see now from our F/+ rankings and all of the other stat pages are pure, unadulterated, opponent-adjusted 2012 data.
And as is customary after seven weeks, there are some numbers and rankings that you might either call crazy or intriguing, based on your rooting allegiances. Let's have a look at some of what I will politely call the "quirks" of the Week 8 FO rankings.
Want to know how perfect Oklahoma's performance was against Texas in Dallas last weekend? The Sooners zoomed from a plus-23.5 percent F/+ last week to plus-43.2 percent this week. (That they have played just five games, instead of six or seven, helps -- this perfect performance accounts for a full 20 percent of the Sooners' output thus far.) Oklahoma allowed Texas to run just 21 percent of its plays on the Sooners' side of the 50, and Texas only had a chance to run 12 plays before the game was officially not "close" anymore. For those 12 plays, Texas had a 0.0 percent success rate. Zero. Point. Zero. Meanwhile, the Sooners were ripping off 95-yard touchdown runs. Close-game S&P? Oklahoma 1.202, Texas 0.024. This was quite possibly the best performance of the season by any team.
Will it last? Maybe, but probably not. By all means, a total destruction of the F/+ No. 20 team, plus an easy road win over the No. 5 team (Texas Tech) will get you ranked rather high. But I struggle with the thought that Oklahoma will continue to look as good as it has the last two weeks. That said, the receiving corps has improved dramatically over the last month, with the emergence of freshman Sterling Shepard and the sudden eligibility of Fresno State transfer Jalen Saunders. Running back Damien Williams showed against Texas that he has quite the fifth gear. The Sooners have proven that they are a worthy Top 5-7 team over the past two weeks, but I struggle with the thought of them being better than, at the very least, Oregon and Florida. Plus, you know, they did already lose to Kansas State.
This is mostly an FEI-driven ranking -- the Hawkeyes are 21st in FEI and 40th in S&P+. The Hawkeyes grade out quite well in DFEI (No. 13), and the offense has at the very least stopped being a complete liability. Thank "Hebrew Hammer" Mark Weisman for that.
So how is Iowa ranked higher in the combination F/+ measure than it is in either S&P+ or FEI? First, the addition of the Special Teams F/+ adjustment (we split out Brian Fremeau's Special Teams Efficiency measure to account for about 14 percent of the total F/+ equation) to this week's rankings helped them a bit; Iowa ranks 24th in Special Teams F/+, which gives Kirk Ferentz's squad a slight edge over USC (No. 33 in Special Teams F/+), Michigan (No. 57), Wisconsin (No. 68), Michigan State (No. 74), Utah State (No. 76), Arizona (No. 87), and West Virginia (No. 110). Plus, there is an enormous batch of teams with almost exactly Iowa's F/+ rating.
Will it last? Probably not. Let's put it this way: at plus-17.9 percent, Iowa is 3.5 percent below No. 15 Stanford and only 4.3 percent ahead of No. 29 TCU. With one average performance, they could drop pretty far. Plus, I'm pretty sure you couldn't even find many Iowa fans who believe their team is among the 16 best in the country right now.
The state of Iowa has donated a large sum of money to the FO coffers, apparently. Both of the state's FBS teams produce eyebrow-raising ratings. But when it comes to Iowa State, look at it this way: if you believe that Kansas State is the No. 4 team in the country, that Texas Tech is No. 5, that Iowa is No. 16, and that TCU is No. 29, then Iowa State's ranking makes some sense, doesn't it? The Cyclones lost to KSU and Texas Tech by a combined 17 points, beat Iowa in Iowa City, and whipped TCU in Fort Worth. Therefore you would expect them to be ranked below the former two and above the latter two. If you think that the statistical perceptions of those other teams are not quite right, then you almost certainly won't believe Iowa State, either.
Will it last? It's a broken record, but probably not. The Cyclones have a damn strong defense (13th in Def. F/+) and decent enough special teams, but since I can't really buy into Iowa being in the top 20, Texas Tech being in the top 10 and Kansas State being in the top 5, I have to figure ISU doesn't end up top 15 either. Top 30, though? Quite possibly. And that alone would be a serious upgrade for Paul Rhoads and the Cyclones.
Thanks to S&P+'s extreme crush on Alabama (and, therefore, any team Alabama has played), Michigan still grades out pretty well in overall F/+: they come in at 18th, which is about right considering their losses have been to No. 1 'Bama and No. 6 Notre Dame. But the Wolverines rank just 47th in FEI and 63rd in OFEI; typically when there is a big disparity between a team's Off. S&P+ and OFEI, it suggests that a team has a lot going for it on a play-for-play basis but struggles to avoid leaving points on the board. That appears to be Michigan's problem right now.
Will it last? Probably not. The Wolverines rank seventh in Off. S&P+, and the odds are pretty good that theses two rankings will begin to meet in the middle after a while. This is not an elite offense this year, but it's still pretty good.
So far this year, the Aggies have allowed 4.1 yards per play to Utah, 3.8 to Wisconsin, 4.4 to Colorado State, 4.1 to UNLV, and 4.7 to BYU; against San Jose State, the Aggies sacked quarterback David Fales 13 times and held the Spartans to a 0.578 S&P when the game was considered "close." (USU went up 28-3 early in the second quarter, allowed a couple of scores late in the second, then hit the gas again in the third.) The defense, therefore, ranks 14th in Def. S&P+, a figure that is probably too high … but not egregiously so. That USU has lost to two teams ranked below them (BYU and Wisconsin) suggests they are probably ranked a bit too high, but that those two losses came by a combined five points on the road suggests this is, at worst, a legitimate top 40 team.
Will it last? I'll give you one guess. Probably not. But again, Gary Andersen's squad is still a lot better than you probably think. Among other things, the Aggies will have to absolutely wreck four remaining opponents -- No. 81 Texas State, No. 113 Idaho, No. 118 UTSA and No. 121 New Mexico State -- to maintain opponent-adjusted figures that high, and they probably need to beat Louisiana Tech on the road as well. That probably won't happen, but I would imagine USU still finishes about 9-3 with a top 40 ranking.
How does an undefeated team from a BCS conference end up ranked below, among other teams, 2-4 Marshall (No. 92 in S&P+), No. 91 UNLV (1-6) and No. 90 Buffalo (1-5)? By playing a couple of truly awful FCS teams (Fordham and Delaware State), and by dilly-dallying. Thanks to turnovers, the Bearcats never did put away a Delaware State team that has also lost to Florida A&M and Delaware and ranks 198th in Jeff Sagarin's current rankings. (DSU was 240th last year, so … improvement!) Then, against Fordham (Sagarin ranking: No. 187) this past weekend, the Bearcats led only 14-6 at halftime before hitting the accelerator. If there's one thing we know about S&P+, it is that it is merciless to teams that fail to put away bad opponents. You can rank really high with a weak schedule if you take care of business (like Boise State in recent years); but if you don't, you are going to sink like a stone.
Will it last? Certainly not. In games versus FBS competition, Cincy has beaten Pittsburgh by 24, Virginia Tech by three, and Miami (Ohio) by 38. This might not be better than a top 30 team right now (Virginia Tech isn't good enough this year to be the cornerstone of a great resume), but the Bearcats are certainly better than 94th. But they have played just five games, and two were terrible efforts against low-tier FCS teams; right now, they are paying dearly for that.
See, Kansas State fans? Maybe I didn't design a system of numbers with the sole purpose of dumping on K-State whenever possible! (Honestly, this ranking is pretty well-deserved and not incredibly surprising. I just wasn't going to pass up an opportunity to antagonize.) Thanks to improvement by running back John Hubert and some explosive passing options (No. 2 target Tramaine Thompson is averaging 12.3 yards per target, No. 3 Tyler Lockett 9.9, tight end Travis Tannahill 13.6), Kansas State is proving more capable of scoring via big play this year, not just working enough controlled, "cloud of dust" drives to win tight games. Many of the big gainers come late in the game, too, after three quarters or so of pounding away with quarterback Collin Klein and Hubert. The Wildcats scored 35 points in the fourth quarter against Missouri State, 21 against Miami, and 14 against Oklahoma. And they scored 28 against Kansas in the third quarter. Eventually they get you.
Honestly, I am a little surprised to see the Wildcats grading out this high, simply because of their occasional ability to dilly-dally as well. They led Missouri State by just seven points until the fourth-quarter explosion, they beat North Texas by just 14, and they had to hold on for dear life to take out Iowa State last weekend. But thanks to the overall explosiveness, they are still winning games with more margin for error this time around.
Will it last? I think so. If they don't finish in the top 5 of Off. S&P+, I could definitely see top 10. This is a controlled attack capable of greater diversity than last year. The defense still bends a little bit too much for my tastes, and the team still relies a bit too much on not only turnovers but timely turnovers. (KSU recovered a fumble for a touchdown and recovered another fumble at their own 6 in a five-point win over Oklahoma -- do the math on that one.) This is a much sturdier team in 2012, and if KSU wins in Morgantown on Saturday, this sturdy team could win the Big 12.
1 comment, Last at 19 Oct 2012, 6:45pm by Kal