MGoBlog has moved. The new site can be found at MGoBlog.com

Tuesday, October 30, 2007

Guest post, this from FOB Nick Mahanic.

(I really need to give massive credit to Vijay from iBlog For Cookies, who pulled all of the data that I have here out of James Howell’s database. Any Hot Blog Groupies who get turned on by this should send gratitude his way.)

There’s a strain of conventional wisdom which suggests that Lloyd Carr is a very good coach against top teams but struggles against mediocre ones. I wanted to cut to the core of it: how good is Lloyd against top teams, and how bad is he against weaker ones?

To solve this problem, I decided to break out the statistics. I took a look at Lloyd Carr’s record against teams in different ranking groups based on the final AP poll. The record often cited by announcers is his record against teams in the top ten when Michigan plays them. There’s nothing wrong with this, but not all rankings are equal. Rankings early in the year reflect more speculation and less information based on actual performance; does beating #9 Colorado in 1997 when they wound up 5-6 (or 0-11 counting those pesky forfeits) really count as beating a top team? Maybe Carr's stats were propped up by running into an above average number of overrated teams.

The problem with these numbers is that there isn’t a well-known basis for comparison. Everyone knows that a .300 batting average, a 2.00 ERA, and rushing for 150 yards in a game are pretty good because they see these stats used all the time. But you rarely see a stat showing that Coach Sobchak is 4-7 against teams ranked 11-25 in final AP polls.

The best way to figure assess this information was to compare Carr’s records to those of his coaching peers. To simplify, I compared him to coaches who are either approximately as good as him (say, Mack Brown) or better, specifically coaches that we can all agree are excellent (Steve Spurrier, Lou Holtz, Tom Osborne). Then I threw in some random people like Mark Richt and John Cooper.

Here’s the data. A couple of things to note. First, the records are based on tenure for one team. This would be the most successful of a given coach’s career: Florida for Spurrier, Notre Dame for Holtz, Texas for Mack Brown. You get the idea. Also, the AP only ranked the Top 20 until 1989, when it began to rank 25 teams. This explains why Tom Osborne only has 3 games against teams in that range despite coaching for many years. This oddity also affects Bobby Bowden and Joe Paterno. Lastly, I’d really like to do this with some more coaches (Beamer, Fulmer) and with the Sagarin rankings.

Versus Teams Ranked 1-5:

CoachW L TGamesWin %
Bob Stoops 3 4
7 43%
Lou Holtz 6 9 1 16 41%
Steve Spurrier 7 14 1 22 34%
Lloyd Carr 4 8
12 33%
Mack Brown 3 6
9 33%
Pete Carroll 1 2
3 33%
Bobby Bowden 7 30
37 19%
Joe Paterno 6 27
33 18%
John Cooper 1 8 1 10 15%
Mark Richt 1 6
7 14%
Tom Osborne 5 30
35 14%

Lloyd Carr is in impressive company here. Paired with Steve Spurrier, Mack Brown, and Pete Carroll (albeit over a very small sample), and below only Lou Holtz and Bob Stoops. If you think about it, 4-8 is pretty good against Top 5 teams. The Top 5 in any given year is going to have unbeaten teams, 1 loss teams, and maybe 1 or 2 2-loss teams. To beat a team that finishes in the Top 5 is to likely give it its only loss. No one should be able to do this over half the time. Note Tom Osborne at the bottom.

Versus Teams Ranked 6-10:

CoachW L TGamesWin %
Mark Richt 2 0
2 100%
Bob Stoops 5 1
6 83%
Tom Osborne 11 3
14 79%
Pete Carroll 7 2
9 78%
Bobby Bowden 15 8 1 24 65%
Lloyd Carr 7 6
13 54%
Lou Holtz 9 8
17 53%
Steve Spurrier 6 6
12 50%
John Cooper 3 9 1 13 27%
Mack Brown 2 7
9 22%
Joe Paterno 6 22
28 21%

Mark Richt’s ranking is flimsy given the puny sample size. I’d be surprised if Stoops’ level is sustainable too but we’re never going to get a big sample here, and he is pretty close to Osborne, who coached for a long time. Carr is still faring very well here. He’s ahead of Lou Holtz and Steve Spurrier. Being behind the people he is behind is no loss; remember, these are some legendary coaches. Notice Paterno and Cooper again near the bottom. More on that later.

Versus Teams Ranked 11-15:

CoachW L TGamesWin %
Mack Brown 4 0
4 100%
Pete Carroll 4 0
4 100%
Tom Osborne 14 3
17 82%
Steve Spurrier 10 4
14 71%
Bobby Bowden 19 8
27 70%
Joe Paterno 24 11
35 69%
Bob Stoops 4 3
7 57%
John Cooper 6 5
11 55%
Lou Holtz 3 4 1 8 44%
Mark Richt 3 4
7 43%
Lloyd Carr 3 6
9 33%

Ugh. Now we get to it. Lloyd is woefully behind everyone here. At this point I wish I had thrown in a few coaches with obviously mediocre records (not Weis-like, but maybe Joe Tiller) to see who he sits with. But the point is clear: this is bad.

Versus Teams Ranked 16-20:

CoachW L TGamesWin %
Mark Richt 4 0
4 100%
Steve Spurrier 10 0
10 100%
Lou Holtz 8 1
9 89%
Mack Brown 5 1
6 83%
Tom Osborne 21 4 1 26 83%
Lloyd Carr 10 3
13 77%
Pete Carroll 5 2
7 71%
Bobby Bowden 16 8
24 67%
Joe Paterno 23 14 1 38 62%
Bob Stoops 3 2
5 60%
John Cooper 6 6
12 50%

Lloyd’s actually pretty good this time around, right in the middle of the pack. Better than Bowden and Carroll. The most significant thing here is that Steve Spurrier at Florida was damned good. If you doubted this somehow, it’s about to become even clearer.

Versus Teams Ranked 21-25:

CoachW L TGamesWin %
Lou Holtz 5 1
6 83%
Joe Paterno 14 3
17 82%
Bobby Bowden 8 2
10 80%
Steve Spurrier 6 2
8 75%
Tom Osborne 2 1
3 67%
John Cooper 5 3
8 63%
Mack Brown 1 1
2 50%
Mark Richt 3 3
6 50%
Pete Carroll 1 1
2 50%
Bob Stoops 2 2
4 50%
Lloyd Carr 3 4
7 43%

And back to the cellar for Lloyd. His record here (and in the 11-15) is actually worse than his record at 6-10 and 1-5 (technically 1-5/11-15 are equal). This is probably a sample size issue, since it is almost impossible that whatever strategic flaws he has, he’s actually more likely to win against (much) better teams. Nevertheless, the point is that while the man can do a great job against top opponents, he leaves something to be desired against weaker opponents.

Versus Unranked Teams:

CoachW L TGamesWin %
Steve Spurrier 83 1
84 99%
Tom Osborne 202 8 2 212 96%
Mark Richt 48 4
52 92%
Mack Brown 78 7
85 92%
Bob Stoops 69 7
76 91%
Lou Holtz 69 7
76 91%
Lloyd Carr 86 9
95 91%
Pete Carroll 47 5
52 90%
Bobby Bowden 227 26 3 256 89%
John Cooper 90 12 2 104 88%
Joe Paterno 290 44 2 336 87%

Yeah, Lloyd’s numbers are all well and good here (better than Carroll, actually), but be honest: you stopped at Spurrier. There was a man who didn’t fool around and took care of business. Kind of shocking that he only won one title at Florida (and that with a loss). Or was it? More on that soon.

First, though, another grouping:

1-10 W L T G %
Pete Carroll 8 4 0 12 67%
Bob Stoops 8 5 0 13 62%
Lou Holtz 15 17 1 33 47%
Lloyd Carr 11 14 0 25 44%
Steve Spurrier 13 20 1 34 40%
Bobby Bowden 22 38 1 61 37%
Mark Richt 3 6 0 9 33%
Tom Osborne 16 33 0 49 33%
Mack Brown 5 13 0 18 28%
John Cooper 4 17 2 23 22%
Joe Paterno 12 49 0 61 20%

11-25 W L T G %
Mack Brown 10 2 0 12 83%
Tom Osborne 37 8 1 46 82%
Steve Spurrier 26 6 0 32 81%
Pete Carroll 10 3 0 13 77%
Lou Holtz 16 6 1 23 72%
Bobby Bowden 43 18 0 61 70%
Joe Paterno 61 28 1 90 68%
Mark Richt 10 7 0 17 59%
Bob Stoops 9 7 0 16 56%
Lloyd Carr 16 13 0 29 55%
John Cooper 17 14 0 31 55%

So the conventional wisdom is right. Carr is terrific against elite teams but poor against mediocre ones, and fine against really bad ones. One fun fact: if Carr can beat OSU this year (and OSU finishes in the Top Ten) he will have as many wins against Top Ten teams as JoePa, except in five hundred million fewer years. Joe Paterno does not look particularly good relative to this group, which ordinarily would be no shame, but if you are in a race against death with Bobby Bowden for the most wins ever, you’d take this as a disappointment. Also, John Cooper sucks (at least relative to this group, which should be no surprise). Note, though, that Bowden, Osborne, and Paterno’s “Unranked” records should be upped a bit to compensate for the fact that at least some of the teams they faced might otherwise have fallen into the 21-25 group in a different era (all else being equal, their unranked group is tougher than everyone else’s, at least a little bit).

One more chart (last one, I swear).



Games

% of schedule

Total Games 1-10 11-25 UR 1-10 11-25 UR
John Cooper 158 23 31 104 15% 20% 66%
Lloyd Carr 149 25 29 95 17% 19% 64%
Mack Brown 115 18 12 85 16% 10% 74%
Mark Richt 78 9 17 52 12% 22% 67%
Pete Carroll 77 12 13 52 16% 17% 68%
Joe Paterno 487 61 90 336 13% 18% 69%
Bobby Bowden 378 61 61 256 16% 16% 68%
Tom Osborne 307 49 46 212 16% 15% 69%
Steve Spurrier 150 34 32 84 23% 21% 56%
Bob Stoops 105 13 16 76 12% 15% 72%
Lou Holtz 132 33 23 76 25% 17% 58%

This chart shows the breakdown of games played for each coach, both in raw numbers (the left) and by percentage (the right). While this doesn’t say much about anyone’s performance within a group, it does tell us how that is reflected in their overall record.

For example, remember when I mentioned how great Spurrier’s numbers were, but how he only won 1 national title? Looking at this, it’s no wonder he didn’t win more. 44% of his games were against ranked teams. 23% were against teams that finished in the Top Ten. That’s an average of almost 3 top ten opponents a season! A good part of this is because his tenure overlapped with FSU’s glory years, and the rest probably due to the SEC being strong (as well as Florida making a lot of SEC title games and good bowl games). In this respect, he’s very similar to Lou Holtz: excellent numbers but very daunting schedules.

On the other end of the spectrum, we have our good friends in the Big Twelve, Bob Stoops and Mack Brown. Stoops has played the smallest percentage of Top Ten teams; Brown barely plays more than 1 team ranked 11-25 in a season. This isn’t necessarily their fault: the Big Twelve hasn’t had much of a middle from the looks of things, and their foes in the North have been struggling to put together any elite team for a few years now. Still, it makes their accomplishments slightly less shiny when you consider the road they’ve traveled.

And finally, Lloyd Carr is on the higher end of all categories but not at the top of any. He’s certainly earned the achievements he’s won.

So what does this all mean? In a sense, this confirms one of the more common adjectives to describe a Lloyd Carr-coached team: maddening. Take an athlete with decent speed but not enough to be an elite runner. You understand why he can beat bad runners but can never beat the best. You understand it and can live with it. If Carr could handle weak teams but came up short against the top teams we could come to a conclusion that makes sense: Lloyd is decent but not up to par with the top coaches in the game.

Yet this evidence suggests he can coach his team well when he “needs” to. But he does something when playing weaker teams that really hurts Michigan’s record. A decent job against those teams means 4 or 5 more wins over his tenure. That’s a couple more Big Ten titles, a couple more BCS bowl berths, and, if they happened at the right time, another shot at a national title.

I cannot answer the “why”, but suffice it to say that Carr is what he is. Let’s hope he has at least one more big win in him.

0 Comments: