As I was reading your article this morning about BCS crashers I wondered..."where does Okie state rank on defense?" From an S&P+ perspective they rank 9th. I saw that, thought it odd because they have never been particularly good on defense, but said, "wow if they are that good on defense, Bill is right they have a legit shot at going unbeaten." So then I looked at their "conventional" measure, and I found their defense is average as grits.....Yet by S&P measures they are 9! Part of this must be explained by the amount of yardage they surrender in garbage time and the number of turnovers they have gotten. Are their S&P numbers skewed positively by a smaller number of possessions when the game is close? Nonetheless, I have a hard time believing a defense is somehow that different when its winning big vs. when its close. We would not expect to see the same skew with other good defenses.
-- Michael C.
The Big 12 has long been the home of dissonance when it comes to the difference between raw defensive stats and those of the opponent-adjusted variety. (And for the last couple of years, there has been dissonance of the conference realignment variety too.) The best season for this was 2008, when a deep group of experienced offenses (Sam Bradford and his 60-point crew at Oklahoma; Colt McCoy, Jordan Shipley and Quan Cosby at Texas; Graham Harrell and Michael Crabtree at Texas Tech; Chase Daniel, Chase Coffman and Jeremy Maclin at Missouri; Zac Robinson, Dez Bryant and Kendall Hunter at Oklahoma State) completely torched otherwise solid offenses. After opponent adjustments, you had odd results like Missouri ranking 118th in passing yards allowed and 17th in Def. Passing S&P+ and Oklahoma ranking 100th in passing yards allowed and sixth in Def. Passing S&P+. The phenomenon has continued in recent years, though it has never been quite as stark as 2008.
As a whole, however, Oklahoma State's performance indeed has more to do with garbage time than the quality of the offenses they've faced. To be sure, quality has had a slight role to play -- Texas A&M ranks 12th in Off. S&P+, Arizona 26th and Tulsa 37th -- but Oklahoma has benefited most from taking care of business and putting the game out of reach in little time.
Let's walk through the numbers a bit. We'll stick with general per-play numbers (instead of my preferred S&P) because it's a bit more universal in terms of who does or doesn't understand it.
And as a reminder, a game is considered "close" as long as it meets one of these criteria: within 28 points in the first quarter, within 24 points in the second, within 21 points in the third and within 16 points in the fourth. Teams with leads of this size win almost 100 percent of the time. Anything that does not fit one of these criteria fall into "garbage time." Filtering out garbage time results in stronger, more predictive numbers.
|29 for 86 (3.0),
|47 for 189 (4.0),
|50 for 274 (5.5),
|22 for 165 (7.5),
|33 for 113 (3.4),
|52 for 369 (7.1),
|74 for 471 (6.4),
|20 for 106 (5.3),
|61 for 372 (6.1),
In non-garbage time, Oklahoma State has held every opponent under its per-play average. Against Tulsa, they held the Golden Hurricane to nearly half of their full-season rate. The type of opponent adjustments that go into the rankings at Football Outsiders are more complex, but this is the general sentiment.
As you see, Oklahoma State has completely taken their foot off of the accelerator when it comes to garbage time defense. Tulsa's per-play output nearly doubled, while those of UL-Lafayette and Arizona improved by at least one full yard. Does this signify that OSU's defense probably isn't really as good as it has shown in non-garbage time thus far? Perhaps. But two things bear mentioning (or, in one case, repeating):
1.) As a whole, filtering out garbage time results in more telling, descriptive and accurate numbers. While Oklahoma State could certainly end up an outlier in this sense, there is a very specific reason for all but writing off what happens when the game is put out of reach. Obviously the best approach would be to stop counting when the winning team takes its starters out, but pulling that information isn't possible with the tools we have at our disposal. And besides, teams rarely pull all of their starters at once.
2.) In the one full game they had to play this year, the Oklahoma State defense got much, much better as the game progressed. Sure, it was against Texas A&M, a team that has taken the third and fourth quarter off for the past three games, but still, OSU answered the bell for a full game the one time they were asked to do so.
Based on Nebraska's second half against [Ohio State], statistically, have they turned a corner?
No. Comeback wins can sometimes spur a team to great things in the future, but it is far from guaranteed. And as mentioned in this week's Numerical, the Huskers' win over Ohio State turned in such stark manner with the injury to Ohio State quarterback Braxton Miller that it is quite difficult to say the same thing would have happened with Miller behind center. A Miller fumble did indeed allow Nebraska to cut OSU's lead to 27-13, so perhaps momentum would have continued if Joe Bauserman had not had to come into the game. That said, on a yards-per-play basis, the contrast was significant:
Yards Per Play, Pre-Miller Injury: Ohio State 7.4, Nebraska 4.4 (Score: OSU 27, NU 13)
Yards Per Play, Post-Miller Injury: Nebraska 7.6, Ohio State 2.6 (Score: NU 21, OSU 0)
Teams, their momentum and their chemistry are such odd entities that you never know what will prompt good things or bad things in the future. In terms of known knowns, however, letting Ohio State go up 21 points in the first place told us as much about Nebraska as their comeback did. But with upcoming games versus Minnesota and Northwestern (with a tough visit from Michigan State in between), it isn't hard to see Nebraska developing a bit of a win streak here as October expires, and I'm sure that if or when that happens, the comeback will be given credit for the turnaround.
Statistically, Nebraska is still quite iffy. They are poor in terms of passing downs offense and questionable in terms of standard downs defense. That defensive tackle Jared Crick is lost for the season only complicates matters further. It isn't too late for the Huskers to pull it together and make a run to win their division, but a) their performance to date doesn't suggest that is likely, and b) if that happens, I'll have a hard time believing it was because they overcame a big deficit once the other team's starting quarterback got hurt. I've been wrong before, however.
How likely are we to get 3 or more unbeaten teams from AQ conferences?
My Outsiders colleague Brian Fremeau tracks something called "Mean Remaining Wins" with his FEI ratings. It asks, basically, the following question: if teams were to play their remaining schedule a hundred times, how many games would they win on average? For the undefeated teams from BCS conferences (sorry--I've just never been able to call them "AQ"), here are their games remaining and Mean Remaining Wins:
LSU: 5.0 Remaining Mean Wins in six games.
Alabama: 4.4 Remaining Mean Wins in five games. (Note: the Tide actually have six games remaining, but one comes versus Georgia Southern, and Fremeau's numbers ignore games versus FCS competition).
Oklahoma: 6.2 Remaining Mean Wins in seven games.
Oklahoma State: 5.7 Remaining Mean Wins in seven games.
Kansas State: 3.6 Remaining Mean Wins in seven games.
Wisconsin: 6.2 Remaining Mean Wins in seven games.
Michigan: 5.1 Remaining Mean Wins in six games.
Illinois: 3.3 Remaining Mean Wins in six games.
Georgia Tech: 5.0 Remaining Mean Wins in six games.
Clemson: 4.6 Remaining Mean Wins in six games.
Stanford: 6.2 Remaining Mean Wins in seven games.
Because They're Basically BCS At This Point
Boise State: 6.8 Remaining Mean Wins in seven games.
When all is said and done, I don't honestly know if we will have that many undefeated teams. Sure, it could go as we think it will right now (the LSU/Alabama winner rolls through the rest of the schedule, as do Oklahoma, Wisconsin, Stanford and Boise State, at the very least) but as I mentioned on Wednesday, we typically have a lot of undefeated teams at this time, and we typically start envisioning doomsday scenarios of crazy numbers of undefeated teams getting snubbed from a shot at the title. Crazy things happen -- in 2004 and 2009, we ended up with the dreaded "more than two undefeated BCS conference teams" scenario -- but odds are still pretty good that the ranks of the undefeated will shrink considerably, almost to zero, over the next seven weeks.
Will Dr Pepper ensure that I have a real good time, or merely just a good time?
Two answers for this one:
1) Does saturating a specific audience (in this case, college football fans) with a specific, repetitive commercial really have the positive impact intended? Because I've got to say ... after seeing the same commercial 14,324,690 times before the midway point of the season, I've developed a reflexive reaction, and it isn't a positive one. The wife and I went to Sonic last week, and when she couldn't decide between a Dr. Pepper and a cherry limeade, I told her she should get the cherry limeade because "screw Dr. Pepper and those damn commercials."
2) LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME. LET'S HAVE A REAL GOOD TIME, LET'S HAVE A REAL GOOD TIME
The Morning Tailgate runs weekdays. Catch up on the archives.