In this week's Varsity Numbers, Bill Connelly takes a page out of baseball's playbook and attempts to isolate power from efficiency.
24 Sep 2010
by Bill Connelly
It has been 22 days since Wake Forest and Presbyterian kicked off the 2010 season (unfolding one of the greatest trick plays of all time in the process). It is still too early to worry much about drawing serious conclusions about much, at least not without the delete key nearby, but what do we think we know so far?
We know that Alabama has done nothing to lose the title of "Best Team in the Country" just yet. (And as you'll see in the ratings below, it is not particularly close.)
We know that the worst football in the country is probably being played just north and south of the Durango-Silverton railroad, at Colorado State, New Mexico, and New Mexico State.
We know that Denard Robinson, Marcus Lattimore, LaMichael James, Trent Richardson, Jeffrey Demps and the Rodgers Brothers are really fun to watch.
We know that coaches like Minnesota's Tim Brewster, New Mexico's Mike Locksley and possibly even Mississippi's Houston Nutt need to turn things around very quickly. We also know that nobody particularly likes Lane Kiffin.
We know that Bill McCartney, 70 years old and out of coaching since 1994, probably isn't the answer in Boulder.
And we know that it is still early enough in the season that some pretty creative teams have found their way into the S&P+ Top 25. Conclusions any more committal than this are likely to fall apart over the course of three hours on a Saturday afternoon. But admitting we don't yet know much isn't going to prevent us from diving into any number of topics, now is it?
College Gameday's presence in Auburn last weekend was rather prescient, as Auburn and Clemson put together perhaps the best and most dramatic of last weekend's marquee games.
In last week's Seventh Day Adventure podcast, I mentioned that I didn't need to see Clemson win at Auburn to name them the potential favorite in the ACC -- I just needed to see them looking good. Apparently I should have been a bit more demanding, as Clemson all but took home a major road win, falling short in gut-wrenching fashion. They certainly became an ACC favorite in the process (somewhat by default), but that probably didn't mean much to them late Saturday night.
|Field Position %||41.5%||29.5%|
|Close Success Rate||43.9%||42.6%|
|Close Success Rate||43.5%||42.2%|
|Close Success Rate||44.4%||43.8%|
|SD/PD Sack Rate||0.0% / 5.3%||14.3% / 11.1%|
|Turnover Pts Margin||+7.5||-7.5|
|1st Down S&P||0.778||0.456|
|2nd Down S&P||0.788||1.145|
|3rd Down S&P||0.620||1.042|
|Projected Pt. Margin||+8.4|
|Actual Pt. Margin||-3|
Clemson dominated the field-position battle and controlled the game for three of four quarters, but big plays and one large momentum swing allowed Auburn to even things and eventually win in overtime. In their four regulation scoring drives, all of which took place between the 1:14 mark of the second quarter and the 1:03 mark of the third, Auburn unleashed plays of 12, 12, 14, 15, 16, 18, 23, 34, 35 and 78 yards. These ten plays accounted for 257 yards, while Auburn's other 51 plays managed just 167 yards. For the most part, Auburn was far from consistent, but the Tigers made their big plays count and used their 15-minute surge to make up for 45 minutes of submission to their rivals from the east. Newton completed just seven passes all night, but they went for 203 yards.
If you believe in the theory that standard downs are the game-planning downs, then Auburn's game plan was not a particularly good one. The team's 41.9 percent standard downs success rate was much lower than Clemson's, and their 0.456 S&P on first downs was downright terrible. But big plays cure quite a few ills. That, and dropped passes, like the Kyle Parker-to-Jaron Brown pass that could have won the game for Clemson in overtime.
In the end, Auburn won this great battle, and while that's all that matters in the present tense, the game unfolded in a way that suggests good things of Clemson's overall talent and execution level. While teams like Virginia Tech and Florida State continue to try to figure out what they have, Clemson and, as of last night, Miami, have provided very nice showings against BCS teams in front of ESPN's cameras. That alone makes them possible favorites in the ACC's Atlantic and Coastal divisions, respectively.
For the last couple of Tuesdays, I have taken to Twitter to list out the projected scores from the week's notable games. It is an interesting way to see which games closely reflected their final scores and which were skewed by heaping doses of luck, turnovers, special teams craziness, or all of the above. Here is an example, and here are some of this week's more interesting results:
Stanford 68, Wake Forest 24
EqPts: Stanford 47.6, Wake Forest 17.2.
T/O Pts: Stanford +11.0.
Stanford > Wake Forest +41.4.
When you see below that S&P+ loves Jim Harbaugh's Fightin' Trees, this is why. Sometimes huge point totals and winning margins are skewed by big special teams plays or odd breaks. Other times, they are completely and totally earned. Stanford earned every point of its 44-point, bullying win over the visiting Demon Deacons, and the per-play rankings below are enamored of that. The rankings liked Stanford's destruction of UCLA as well.
Texas 24, Texas Tech 14
EqPts: Texas 15.0, Texas Tech 4.5.
T/O Pts: Texas +0.2.
Texas > Texas Tech +10.7.
If you told the average college football fan that Tommy Tuberville was taking over a team known for its offensive prowess (in a spread, no less) but lacking in consistent defense, the first two reactions would likely be: "He's going to improve their defense," and "He might screw up that offense." Both appear to be true. Again, it is quite early, but Tech, a team that consistently ranked in the top 15 in Offensive S&P+ under Mike Leach, currently ranks 73rd in raw Offensive S&P and 54th in Offensive S&P+. Meanwhile, they rank 22nd in Defensive S&P+.
Of course, Tech's offense might be just fine -- it might be the Texas defense that needs to be singled out here. Tech still has most of the personnel from last year's strong offense, and they were held to just 144 yards and 4.5 EqPts by this fast, attacking unit.
West Virginia 31, Maryland 17
EqPts: West Virginia 25.8, Maryland 13.1.
T/O Pts: Maryland +12.9.
Maryland > West Virginia +0.2.
In the fourth quarter against Marshall and the first quarter against Maryland, West Virginia looked like one of the best teams in the country. They were poised and ridiculously fast, with quarterback Geno Smith, running back Noel Devine, and receiver Tavon Austin all looking magnificent. They outscored their two opponents 29-0 in that span, and it could have been much worse. In the other six quarters, however, they were outscored 31-26 and looked listless, unfocused and horribly inconsistent. Turnovers could have very easily cost the Mountaineers this game.
Florida 31, Tennessee 17
EqPts: Florida 22.9, Tennessee 14.0.
T/O Pts: Florida +9.7
Florida > Tennessee +18.6.
A timely fake punt and generally solid play led Florida to victory here. They are getting incrementally better each week. Meanwhile, Tennessee is talented and competitive, but they might just be too damn young to do much damage this season.
Oregon 69, Portland State 0
EqPts: Oregon 48.1, Portland State -1.3.
T/O Pts: Oregon +8.8.
Oregon > Portland State +58.2.
Oregon earned almost every bit of this shellacking, and I almost feel embarrassed acknowledging that this game took place. The Ducks have looked incredible so far in 2010, though they face an interesting test in Arizona State tomorrow.
TCU 45, Baylor 10
EqPts: TCU 38.7, Baylor 13.6.
T/O Pts: none.
TCU > Baylor +25.1.
Without referencing games that took place in other seasons, give me one reason why TCU should be ranked below Boise State right now. They too have looked outstanding so far in 2010, first disposing of a strong Oregon State team, then humiliating a Baylor team that, while still Baylor, is faster and more well put together than recent seasons.
Boise State 51, Wyoming 6
EqPts: Boise State 38.6, Wyoming 7.6.
T/O Pts: Boise State +1.6.
Boise State > Wyoming +32.6.
If Boise State does have a claim to a higher ranking than the Horned Frogs from Fort Worth, it comes with the fact that they beat up Wyoming much worse than Texas did, and they did it on the road (Texas beat Wyoming 34-7 in Austin the week before, though they trailed late in the first half.). The Broncos took care of business, and they must do so again against Oregon State this weekend.
Oklahoma 27, Air Force 24
EqPts: Air Force 26.0, Oklahoma 23.5.
T/O Pts: Oklahoma +4.3.
Oklahoma > Air Force +1.8.
Like Ole Miss did with Jacksonville State, Oklahoma had this one well in hand, leading 27-10 in the fourth quarter when the Sooners took their eye off the ball and almost paid big. The Falcons and their tricky running game scored twice to throw a scare into the fans at Owen Field, but Oklahoma held on for the win. Turnovers did the Sooners plenty of favors.
We now have 2010 S&P+ rankings up and rolling, and they will be updated weekly. (Right now the 2009 pages are still the defaults with no year in the URL, but that should change Friday night.)
After I posed the "Which early season approach to ratings makes more sense?" question last week, I did a little research and found that, from a predictive point of view, incorporating more raw data and giving less heft to preseason projections seems to be the best approach. So while the rankings over the second half of the season will not change, we are adjusting the way we calculate these early figures. As you might expect, that means some teams' rankings have shifted drastically from last week.
San Diego State (55 spots, from 113th to 58th). Last week, the Aztecs were saddled with low projections and a weak early schedule. One near-upset of Missouri later, they find themselves ranked higher than they have been in a long time. We do not yet know how they will hold up to the rigors of the tougher portions of their schedule, but this is certainly a more intriguing version of the Aztecs than we have seen in a while.
UCLA (46 spots, from 64th to 18th). It's time for everybody's favorite game, "Fun with Small Sample Sizes!" Let's do the math: Houston gained 497 yards and scores 68 points against Texas State, then the Cougars put up 656 yards and 54 points against UTEP. They head to the City of Angels, and UCLA holds them to 360 yards and 13 points. UCLA's defense must be amazing! Well, maybe, but the numbers do not take into account things like "Houston loses both its first- and second-string quarterbacks to injury mid-game." Powered by some skewed defensive numbers, the Bruins surged in the rankings. Unless they stay close to Texas this weekend, they will likely begin to drift back to the pack.
Kentucky (42 spots, from 51st to 9th). If we are taking raw, schedule-unadjusted numbers into account now, Kentucky is bound to benefit drastically. They have done exactly what you are supposed to do when faced with a less-than-stellar level of opposition: dominate. Are they the ninth-best team in the country? Probably not. Though they will get a chance to prove themselves this weekend at Florida.
Nevada (38 spots, from 63rd to 25th). Though they have executed the Pistol Offense to perfection against overwhelmed non-BCS conference teams, Nevada had not yet broken through with a nice win over a BCS squad. That they were able to outmatch California in terms of both points and speed was damn impressive. If they catch fire, they might be a) hard for Boise State to beat, and b) just what Boise State needs to support what will be, by the time they meet on November 26, flagging strength of schedule numbers.
BYU (72 spots, from 13th to 85th). Positive projections and a decent-looking early win over Washington had BYU propped up pretty well in the rankings. Now, with less of a projections safety net and a slumping offense (quarterback Riley Nelson is lost for the season, leaving true freshman Jake Heaps in charge), the Cougs have tumbled. Though they boast a stellar defense -- one that could give Nevada fits in Provo this weekend -- their offense will hold them back until Heaps gets his footing. But at least they have wonderfully named running back J.J. Di Luigi (their leading rusher and receiver) around which to build.
New Mexico (41 spots, from 76th to 117th). New Mexico fell apart so quickly and completely that their five-year history doesn't paint a very accurate picture of their current program health. They won nine games as recently as 2007, and now they are 1-14 under already embattled head coach Mike Locksley. Their projections were artificially bolstered by their 4- and 5-year history numbers, and their strength of schedule actually got a nice bump from playing Oregon in the season opener. Now, with raw S&P data in the mix, they are where they belong -- near the bottom of the FBS pile.
Minnesota (40 spots, from 50th to 90th). Yes, they had a respectable showing against USC last weekend. That doesn't make up for the loss to South Dakota. (Then again, South Dakota's big win over the Rangers of NW Oklahoma State last weekend sure can't hurt the strength of schedule ...)
Kansas (37 spots, from 62nd to 99th). So ... Is it just an even weeks versus odd weeks thing with them?
Using the more combustible version of the S&P+ rankings yielded a much more creative and interesting F/+ Top 25 this week. (Here's your reminder that the F/+ rankings are the combination of my S&P+ rankings and Brian Fremeau's FEI. They are the closest thing we have to a college version of DVOA. Now with extra volatility!)
|F/+ Top 25 (After Three Weeks)|
|2||Ohio State (3-0)||+29.6%||281.7||3||.216||4|
|7||South Carolina (3-0)||+21.3%||255.5||7||.178||13|
|F/+ Top 25 (After Three Weeks)|
|19||Boise State (2-0)||+14.8%||224.2||36||.201||8|
|20||Virginia Tech (1-2)||+13.3%||220.7||38||.187||5|
|F/+ Top 25 (After Three Weeks)|
|22||Penn State (2-1)||+12.4%||227.3||32||.138||23|
|23||Boston College (2-0)||+12.1%||227.2||33||.131||19|
Really, this gives you an interesting mix of not only per-play and per-drive data, but also preseason projections, early strength-of-schedule adjustments, and raw, "Nevermind the opponents: Did they succeed, or didn't they?" numbers. Again, many of these teams are benefiting (or suffering) from small sample sizes, and the ratings will work themselves out down the line. But for now this paints a decent picture of who has looked the best after just two or three hours of football each. The first thing to note: Alabama is so far ahead of everybody else. The distance between the No. 1 Crimson Tide and No. 2 Ohio State is the same as the distance between Ohio State and No. 15 USC. That's incredible.
While dominance of weak foes explains how Kentucky slid into the No. 17 spot, the most surprising name on this Top 25 has to be Mike London's No. 21 Virginia Cavaliers. The Hoos are just 1-1, with a win over Richmond and a tight loss to USC. But while other teams have struggled against an "FCS Tier 1" team like Richmond, the Cavaliers pulled away for a 34-13 win. And while others have struggled against the USC offense -- the Trojans averaged 40.5 points per game against Hawaii and Minnesota -- Virginia held them to 17 points and 329 total yards. Plus, they have averaged 5.8 yards per play on offense. Running backs Keith Payne and Perry Jones have combined for 321 yards on just 51 carries, and the passing game has been semi-sufficient. It is somewhat easy to see how the opponent-adjusted statistics would like them this early in the season. Plus, their preseason F/+ rank of 48th did not exactly leave them starting from scratch.
(By the way, it is odd that London's first game in charge would come against the team he left behind last December. Then again, it has happened before; Notre Dame kicked off the Gerry Faust era in 1981 by taking on Archbishop Moeller High School of Cincinnati.)
So the question moving forward is, Does Virginia have any chance of maintaining this level of play for the rest of the season?
What They Have Going For Them:
What They Have Going Against Them:
To the Big 12, Nebraska and Colorado for getting it done. Both the Huskers and the Buffaloes agreed to exit terms to ensure that both programs will be playing in their new conferences in the 2011-12 academic year. Nebraska's exit was going to happen no matter what, but now that Colorado has finalized its exit, it prevents what would be the lamest of lame duck seasons, the rare second lame duck season, in fact.
Obviously the "Mike Leach to _____?" rumors were going to flare up sooner than later, but could we please wait until we have better proof than "Former New Mexico employee tells Facebook friends a rumor he heard that is based on a loss that hasn't happened yet"? Twitter briefly blew up about this yesterday despite the unlikeliness of the situation (a coach from outside the program is going to jump in mid-season?), and the fact that the rumor was based on a rumor of a rumor. Give it time, people. The Mike Leach rumors will generate themselves organically. We don't have to force it.
Three more words: more mascot fighting.
In honor of Boise State, the team who, even if they massacre Oregon State this weekend, will see its national (i.e. BCS) standing fall ever so slightly each week, just by playing their conference schedule.
"Ain't No Love," by David Gray
"I'm Happy But You Don't Like Me," by Asobi Seksu
"Little Man," by Tom Waits
"Nobody Ever Gives Me What I Want," by Blind Mississippi Morris
"'Nuff Respect," by Big Daddy Kane
"Respect," by Otis Redding
"Respectable," by the Rolling Stones
"The Underdog," by Spoon
"Whipping Post," by the Allman Brothers Band
"You Like Me Too Much," by the Beatles
And, of course, everything by The Mighty Underdogs.
Georgia over Mississippi State. Spread: Georgia +1 | F/+ Projection: Georgia by 5.5. OK, so picking a one-point underdog to win is not exactly much of an upset. Regardless, Mark Richt needs this one or else, justifiably or not, the buzzards will start trying to circle. It's ridiculous to think that Richt could be in trouble, but a 1-3 start would be enough to get the crazies to start yelling a little louder.
South Carolina over Auburn. Spread: S. Caro. +2.5 | F/+ Projection: S. Caro. by 0.6. Another less-than-risky pick. If Auburn's offenaw is as inconsistent as they were last week, South Carolina and Marcus Lattimore are more than capable of shortening the game and magnifying their offensive mistakes.
Boston College over Virginia Tech. Spread: B.C. + 4 | F/+ Projection: B.C. by 2.4. Virginia Tech is still getting plenty of respect from Vegas and bettors, and it is not entirely undeserved. But looking only at this year's results, the Hokies have not proven as much as Boston College, and the Eagles could be well positioned for a big, early conference win.
New Mexico over UNLV. Spread: UNLV -10.5 | F/+ Projection: UNLV by 0.8. This says much more about how poorly UNLV has been playing, as New Mexico (their third mention in today's VN!) has done nothing to earn anybody's trust.
One more thing we know: Next week we will introduce a slightly different format for Varsity Numbers -- two parts! While I very much enjoy reviewing last week's games, it doesn't necessarily make sense to do to that six days later on Friday afternoons. The goal will be to introduce a "The Week That Was" version of Varsity Numbers around Tuesday morning, and a "The Week That Will Be" version on Friday afternoon. Two smaller, more palatable and timely portions for your dining pleasure.
25 comments, Last at 26 Sep 2010, 4:46pm by Alexander