...the joy and despair of being a Wolverines fan.


November 28, 2007


Great coaches win at about a 75% clip, particularly when viewed over a long career:

                  Bryant - 78.2%

                  Paterno - 74.8%

                  Bowden - 76.2%

Over shorter stretches, great coaches can exceed an 80% winning percentage:

        Stoops - 81.9%

        Carroll - 84.4%

        Switzer - 83.7%

Back when Lloyd Carr was on the hot seat, defenders were quick to point out that the retiring Michigan coach has a 75.2% career winning percentage, seventh among active coaches and right up there with the all time greats. Is it reasonable, then, to ever fault a coach who is winning at an 84.2% pace, albeit over a three-year stretch? Most definitely, argue Les Miles's detractors, many who are still bloodspattered from defending Lloyd’s 75.2% winning percentage like the pass at Thermopylae.

To resolve this paradox, those who dislike Les Miles’s coaching accomplishments at LSU make the obvious point: will you look at all the talent he has at LSU! They are quite comfortable, then, establishing a new standard: a great coach may win at a 75% pace, but if a coach has a lot of talent, 75% is failure, not success. A higher standard, say 85%, is required if you are loaded with blue chippers (no one has defined this, but those making the argument have already declared Les's 84.2% at LSU inadequate, so it must be at least this).  Call this the “Carroll Standard,” after Pete Carroll’s sick 94.6% winning percentage and two national championships from 2003 to 2005, because clearly USC had ridiculous talent those years, and Pete was simply doing what should have been done.

Let's set aside the thorny problem of how we determine when great coaches have benefited from a glut of talent (say Lloyd in 97) and when they brilliantly overcame marginal talent, and grant the premise: Les Miles's winning 84.2% of his games these last three years is a substandard coaching effort because many coaches with that amount of incredible talent would have done better. In layman's terms, all Les did was mess up a sure thing.

Well, these people sure have high standards. Clearly, no other coach they put forward to replace Carr (Brian Kelly is a favorite, maybe Jeff Tedford) has to live up to that 85% winning percentage, but this is because they don’t have obscene talent at their disposal. If Les had crappy players and went 32 and 6, well then! Send the chartered jet!

Their argument breaks down if LSU’s talent does not trigger the Carroll Standard. So let’s take a look:

(data here)

Oh my. Where is the insane talent down in Baton Rouge? In case those reading this rejection of conventional wisdom need more detail, looking at the starting line-ups for Michigan and LSU:

The Rivals rating proves to be a very appropriate measure for our purposes here. Detractors of Les Miles argue that he is winning with Nick Saban’s talent. He has proven he can recruit (another qualification for the Michigan job), so examining the incoming rating of  current starters (many who were Saban's) will reveal a trend: if these numbers had been high, then the two losses each year prove that Miles is not meeting the Carroll Standard. If the numbers are lower than other SEC competitors, then the current players are outperforming their initial rating, suggesting that Les Miles is getting more out of his players than expected…a curious additional qualification for the Michigan job (Michigan being a place, as demonstrated by the data here, where players come in more highly rated and underachieve).

So why do so many think that LSU is loaded with extraordinary talent and not Michigan? Development?

Should Lloyd be held to the Carroll Standard?

The Miles-LSU Talent Glut argument is debunked. In fact, the analysis suggests the opposite: if one wants to claim that the best talent in the SEC is down in Baton Rouge, then Miles and his staff made them that way.

Posted by Meeechigan Dan | Permalink | |