Wednesday, November 07, 2007

Is LSU Really Lucky?

Before Mayhem dashed to commercial at 7:33 this morning after their opening to Tony Barnhart's segment (brought to you by BMW of South Atlanta, just north of your expectations!), Mike Bell and Barnhart were joking that they need to go to Las Vegas with Les Miles because Miles is really lucky right now. Since I'm already in the position of defending everything the guy does on the chance that he ends up as the next head man in Ann Arbor, I'm going to dispute the notion that Miles and LSU have been lucky. I am also doing so because I enjoy taking positions that can be supported by facts and logic.

On the evidence, there's not much of a case to be made that LSU has been lucky this year. The Tigers have been in four close games and won three of them, which isn't especially reflective of luck. They most certainly aren't Virginia (winners of five of six games decided by one score), Ohio State '02 (winners of all seven games they played decided by one score), or Georgia '02 (winners of five of six games they played decided by one score). Moreover, if you look at LSU's four close games, you can't point to any of them as being instances in which the lesser team won. Here are the yardage and first down totals from the four games:

LSU - 391 yards, 25 first downs
Florida - 314 yards, 19 first downs

LSU - 403 yards, 22 first downs
Kentucky - 375 yards, 26 first downs

LSU - 488 yards, 23 first downs
Auburn - 296 yards, 16 first downs

LSU - 475 yards, 21 first downs
Alabama - 254 yards, 20 first downs

If anything, LSU's opponents, especially Auburn and Alabama, were lucky to be in their games against the Tigers. LSU has won the yardage battle in all four games and has had more first downs in three of the four games (with the one exception being the game that LSU lost). The better argument is that LSU was unlucky to be in close games because the fluky factors (turnovers, special teams returns, their kicker having the worst night of his career, etc.). There is an argument to be made that special teams and turnovers are not necessarily random. I'll buy that special teams aren't random, but LSU's special teams aren't bad at all. The case is weaker that turnovers are not random. Phil Steele says that they are random and he's never wrong about anything. Yardage and first downs are better indica of a team's merit and LSU kicks ass in those departments.

Finally, take a gander at LSU's statistical rankings. They're in the top ten in every defensive category and in the top 25 in scoring offense and total offense. Bearing in mind that LSU has played a very difficult schedule, do they really look like a lucky team? I didn't think so.

Finally, let's compare LSU's yards per play gained and allowed against the other contenders for the national title:

LSU - 5.93 gained, 3.67 allowed, 2.26 margin
Oregon - 6.61 gained, 5.4 allowed, 1.21 margin
Oklahoma - 6.56 gained, 4.43 allowed, 2.13 margin
Missouri - 6.42 gained, 5.02 allowed, 1.4 margin
Kansas - 6.35 gained, 4.19 allowed, 2.16 margin
West Virginia - 6.7 gained, 4.17 allowed, 2.53 margin
Ohio State - 5.96 gained, 3.56 allowed, 2.4 margin

Since LSU has played a demonstrably tougher schedule than any of these teams, save Oregon, it can't be said that LSU is lucky to be where they are. Only West Virginia and Ohio State are ahead of LSU in yards per play margin and they haven't come close to LSU in terms of quality of opponents. The best conclusion from this comparison, by the way, is that Oregon and (to a slightly lesser extent) Missouri are poseurs as national title contenders because of their defenses.

(PS - I did this dance last year and reached the startling conclusion that Notre Dame was a fraud. Actually, in retrospect, the analysis was fairly useful, as it flagged Florida as being in the class of the top contenders and it spied USC's weakness.)

And one other thought: how does Tony Barnhart confidently state out of one corner of his mouth that "if [LSU] play[s] Ohio State, the Tigers will win" (call me crazy, but Barnhart is never this assertive except when he's playing to his audience and mocking the Big Ten) and out of the other corner of his mouth deride LSU as lucky? If LSU is lucky and therefore not as good as their record, then how can it be such a certainty that they would beat Ohio State?

4 comments:

Unknown said...

Michael,
Tell me if you disagree - but wouldn't your post and reasoning support the idea that LSU is not a well coached team?

If you dominate four games statistically on offense and defense - but win three extremely close games and lose to UK, doesn't there have to be a reason for that?

Or maybe they were just unlucky.

Side note: I challenge Phil Steele to watch 4 years of Reggie Ball game tape and tell me turnovers are always random.

Michael said...

Les is going to lead Michigan to four straight national titles, so yes, I disagree. :)

Yards per play gained and allowed are as much a function of coaching as anything else. If LSU were a talented, but poorly coached team, then it would not move the ball well on offense (like, say, Florida State) and/or it would not stop its opponents (like, say, Nebraska). A well-coached team doesn't simply avoid turnovers and penalties; it also does well at the basic functions of the game.

I agree with you that turnovers aren't totally random. Some quarterbacks make dumber throws than others. Some defenses can create pressure with four players and thus are more likely to force hurried throws. Some running backs never fumble. (See: Hart, Michael.) That said, there is a heavy random element. Fumbles are a 50/50 proposition when they are on the ground. Deflected interceptions are a luck proposition. Return yardage is a very random element. Empirically speaking, Steele has had a good betting career focusing on yardage and first downs instead of turnovers. The Vegas consensus agrees with him.

peacedog said...

Steele is looking at TOs from a defensive standpoint, isn't he? When you look at things that way, it makes more sense. And the stats bear it out.

Michael said...

No, I think Steele looks at aggregate turnover margin. His preview is full of references to team x being undervalued at a certain point in the season because they had outgained their opponents, but had a bad turnover margin. Look at the "turnovers = turnaround" page.