clock menu more-arrow no yes

Filed under:

Regressing To The Mean

You have a string of numbers, say 1-20. You chose them, one after another, totally at random. If you choose an 18, the odds are very high that the next number will be lower. The exceptional happening was random, inexplicable, and often times enough to win an award.

Because of the layered aspect of even the most commonplace happenings, fans are all too willing to create an explanation for what is otherwise the perfectly random roll of an 18 ("What a run by Reggie Bush! It is because he is the best player in the world" or "Texas beat OSU because they had more guts" or "We scored 50 points this game, therefore, Greg Davis is good at his job").

Football is half understanding what happens and half knowing what should have happened. A decent player can take an opportunity that is given to him and execute it. A great player can create opportunity where none existed before.

The difference is not always easy to see, which is why you get polls like this, while true difference makers like Vince Young and Roy Williams (pick one) have to sit and watch other people give Heisman speeches. It's also why you listen to commentators and message boarders praise Colt and Quan for "guts," when the fact is that they are just doing what they should have done all game had their coaching staff unleashed them. That the biggest play came with 3 seconds left doesn't really matter. The timing was the 19, not the play.

(note that Colt and Quan always did have "guts." My point isn't to say they didn't. They had them all season. They had them when they lost to Tech. Quan was born with more cajones than most men accumulate in a lifetime. They did not get more gutty in the 4th quarter.)

This is just a pet peeve, but it has practical implications, prognostically speaking. Last summer I used it in very broad terms, i.e. "we won 10 games with an average offense and shitty defense, so we should be one win better with a defense that knows what it's doing." I predicted 11-1 based on little more than that sentence.

It's not prefect, obviously, since I used the exact same logic the year before. The defense has to be better than last year, right!?

This year I want to take a closer look at it.

Colt McCoy. There are two ways to score in a Greg Davis offense. One is to play a bad defense. The other is to be so good that you can overcome your disadvantages. In 2005, the offense was Davis-proofed. We had enough variety because each of our 3 run plays had many options within them, and the passing game had enough variety because nobody could afford the man power to scheme it. The guys at ESPN weren't lying when they said Vince ran a simple offense here. He did. Just didn't matter.

Last year, Colt McCoy essentially broke every rule we've ever been taught about offense. The trade-off used to be that the defense would take away deep routes and make you complete ten short ones in a row to score. This worked because it's virtually impossible to do more than once or twice a game. Davis is so repetitive and his running game so weak that we pretty much relied exclusively on big plays to win. The cat & mouse sensibility was completely lacking.

So we get some plucky youngster who thinks he can make it in the big city, and lo and behold he simply shatters the NCAA record for completion percentage. Of all the QBs in all the seasons, nobody ever did it even close to as well as Colt McCoy did. He was the anomaly that was finally good enough to win in a fairly inefficient way. And we finished 11-1.

He's already a 19. Improvement isn't coming from here.

Malcolm Williams. Quan Cosby was an underrated aspect of Colt's record. He was good, sure, but he never dropped any passes. He was never in the wrong spot. He was reliable in the way that a WR in our scheme last year had to be. Williams is not reliable. His effects on the game will be sudden, unpredictable, and way more important.

At some point, there is a trade-off here as well. Sure, Quan might keep a drive alive by catching 4 passes, 2 on 3rd and 6's. Williams can end a drive in one play, from anywhere on the field. Like humans, who at some indeterminate point developed enough brain power to overcome the overwhelming advantage in brawn against . . . pretty much everything, big plays kind of rule the roost in college football. Players are too undisciplined to win the same way every time. You need guys who can do things outside of the chalkboard.

Our offense last season was a perfect metaphor for this process. If Shipley catches that pass against Tech, we probably win. If we don't get that safety, if Gideon doesn't drop that pass, if Kindle doesn't recover that fumble, of the OSU safety lines up 8 inches closer, if Kirkendoll doesn't get those 4 inches, etc, then our season totally changes. These little things happen and you can't stop them.

The question really isn't whether or not he'll improve. It would be hard not to. The question is if his production will be enough to fill the gap left by Quan's reliable excellence. You simply cannot with any amount of reasoning assume that it will.

Brandon Collins and James Kirkendoll. This is a better comparison to Quan in terms of role in the offense, even if Williams will probably fill a more important role. Unless Collins breaks his leg skiiing the powdered slopes or something, he is going to have a breakout season. He is as naturally talented as Jordan Shipley, and although not as consistent as Quan, offers more versatility. Kirkendoll played the Brian Carter role (make two catches a game, at least one of which is huge, ballsy, and utterly game-changing) and looked like he'll excel in it. The biggest jump in a player's ability happens from year 1 to year 2, so expect bigger things from both guys. Big enough, in fact, to combine with Shipley to form a better overall unit at WR than we had last year. But better doesn't always mean much.

Greg Davis. For 6 glorious, nationally televised quarters last year, Greg Davis called the best football of his overvalued life. From the Shipley kickoff return to the final whistle of the Missouri game, we was calling plays like Rainman plays blackjack. He didn't just roll a 20, he rolled a 20 seconds before a meteor full of cash fell through his roof and killed the dinosaur that survived for millions of years behind his fridge. To say that we can rely on that again would be like relying on the money you won in the lottery to heal your lightning burns. Good luck with that.

The "good" news is that the good times didn't last. OkSU started a trend of overrelying on Colt, and the Kansas game started the overreaction of not relying on him enough. It wasn't really until the 4th quarter in the Fiesta Bowl that we found that groove again. Again, it was necessity (Blaine Irby's knee and OSU's manhandling of us have one thing in common, at least), but at least it happened. We lost a game during the stretch, sure, but we still went 5-1 and beat 3 very good teams. So we know we can win with an offense not firing on all cylinders.

Running game. Speaking of not firing on all cylinders. Gotta be better, right? Fozzy's health alone should insure that.

Overall, you're looking at an offense that will likely lose production at it's two top producing positions, but gain some everywhere else. This should break even, more or less, meaning we have an 11-1 offense again. Good thing it's a team sport.

Secondary. The difference was noticeable right away. Last year's spring game looked like a child's soccer game. There's the ball, go . . be around it. This year's secondary looked like they knew what they were doing. They communicated in that way that doesn't even need words. I know what your job is, you know what mine is, let's cooperate!

I've been adamant about my one year rule. First year is learning, second year is growth. There won't be anymore Crabtree-at-the-last-second moments this year. Chykie Brown and Aaron Williams are legitimate football players now, providing us with the extremely rare luxury of covering the third guy as well as the first.
Curtis Brown still has that Tre Allen/Greg Oden disease of being a phenomenal athlete who looks surprised at every single thing that happens. He may grow out of that, he may not. At worst we have a vicious gunner on the punt team. None of those guys is the story, though.

Earl Thomas. Eric Berry and Taylor Mays will get the press, and deservedly so. However, to quote a puppet, "there is another." Texas has been blessed at the safety position up until recently, and Earl may end up better than any of them. Michael Huff had the best single season but it took until he was a senior to put it all together. Earl is a sophomore and has already grown into a playmaker. Combine everyone's comfort with the scheme and The Earl's emergence as supreme leader, and you get a secondary that can win games on it's own.

Pass Rush. Going from Roy Miller to Ben Alexander isn't just losing the rush at one spot, it's losing the rush at two, since Lamarr Houston handles double teams like I handle breakups. This could not be a bigger downgrade. However, it's too early to panic. We have, for maybe the first time since . . . . before I can remember, we have a defensive coordinator who changes his scheme based on the situation! Ben Alexander likely won't play on passing downs, since we go to to a split front built for speed. We could also free up Houston from double teams with alignment and stunts.

Kindle and Acho may not be nose guards, but they have the size to play a 5 technique on a passing down, and I would bet serious of sums of money (like, $40, which is a lot to me), that we see a lot of ends on the field, in a lot of different places, and a lot of 5 man blitzes. Why? Because those would help us keep our 1-on-1 advantages and Will Muschamp isn't a slobbering philistine.

Still, you don't lead the nation in sacks, lose your two best players, then repeat your performance. The pass rush will be worse, believe me. But the combination of the rush and coverage will even out, if not improve, and the pass defense will be better this coming season.

Rush defense. Who gives a shit. Nobody runs in this clown-ass conference anyway.

Schedule. Huge improvement. Tech is breaking in a new QB and we drew them in September, which is usually before Leach's detox takes effect. A&M is a few years away even if something finally goes right for them. Baylor is starting to catch darkhorse buzz, but it's still a one man show in Waco. Missouri sucks until they show me something and Kansas is in Austin. I think our OOC schedule has a high school on it. We have three tough games:

OkSU - Really, I'm only counting this as tough because they schemed the shit out of us last year and the game is in Stillwater. Usually games like that are the exceptional events that normalize the next year. OkSU played a great game and Colt McCoy didn't. In which direction do you think both those factors will migrate? Without Pettigrew, OkSU loses the ability to exploit the middle of the field while Dez "PhD" Bryant draws all the attention outward. They killed us with blitzes on defense, but have a new DC that a) we've seen before and b) isn't a huge blitzer. I think I just talked myself into a bit of confidence, here.

Whoever plays well that given Saturday - Tech didn't beat us beat Tech is better than us. Tech was one of four teams that played well enough to beat us and got enough breaks at the end to win. There will be another one of those games this year. It's unpredictable and you don't know it's happening until it's in the 4th quarter and you never broke away.

Missouri, Kansas, Baylor (if Griffin goes nuts), OkSU, and whatever team we get in the Big 12, if we go, fit this mold. If Tech can beat us again without Crabtree I will take back every bad word I ever typed about them.

OU. Their new punter can't act. We won legitimately, and will improve. They got worse, although not by much. They can certainly win this game, but it's in our court.

Cliff's Notes. Offense will average out to about the same level of effectiveness and the defense will improve. Since we lost one game by two points, and Earl Thomas' growth alone is worth 3, we are a 12-0 football team. Whether we win 10, 11, 12, or 14 depends on the breaks. But I like the odds.