Home Unfiltered Articles Players Baseball Prospectus
Basketball Prospectus home
Click here to log in Click here for forgotten password Click here to subscribe

Click here for Important Basketball Prospectus Premium Information!

<< Previous Article
Surprise, Surprise, Su... (11/19)
Next Article >>
Around the Rim (11/23)

November 20, 2007
Measuring Coaching Performance
Bozeman's Effect

by Ken Pomeroy


One of the more colorful off-the-court stories from last season involved Morgan State head coach Todd Bozeman, who was charged with misdemeanor assault after a dispute at a Virginia restaurant over a sandwich order. (The charges were later dropped after Bozeman apologized to the restaurant owner.) The incident was the only newsworthy event of the season for Bozeman, who after enduring eight years of NCAA coaching sanctions finally returned to the D-I ranks, but at the division's lowest level.

While the situation at the sandwich shop was initially embarrassing to Morgan State, it probably benefits the Bears program in the long term. That’s because any other D-I athletic director that might be interested in Bozeman’s services will not only have to ask the Bears coach about the indiscretions that resulted in the NCAA’s actions–-namely paying the parents of a prized recruit when he was the head coach at Cal in the ‘90s–-but also his ability to represent the school positively off the court. This is relevant, because without the baggage, Bozeman would likely be in high demand following this season. Heck, even with the baggage, he may get some interest from bigger schools. Based on his first year in Baltimore, Bozeman knows what he’s doing on the court, even if he might have some troubles handling himself off of it.

In fact, Bozeman was so effective last season that the case could be made that he did a better job than any other first-year coach in the country. Measuring a coach’s performance is difficult, of course, but it is possible to get a fuzzy idea of how each coach did last season. We can start by comparing how each team performed against expectations. The difficult part is establishing what those expectations should be. The simplest way to do this is to rely on an idea I first saw introduced by Dean Oliver, one of the pioneers of basketball’s advanced statistics movement. That concept is something he calls “the pull of parity.” Bad teams tend to improve the next season, and likewise good teams tend to decline. The amount of expected change is proportional to how good or bad the team was. In general, teams get pulled towards the .500 mark over time.

This concept is more difficult to apply to the college game than to the NBA, but with a few adjustments it still seems to work. The main alteration is that teams don’t gravitate towards the middle of Division I-–they tend to gravitate towards the middle of their conference, and the quality of each conference remains pretty constant from year to year. Using my adjusted Pythagorean ratings, I recomputed each team’s rating to reflect how far it deviates from the conference average. Then we can compare every team’s relative position within its conference in 2006 and 2007 to determine how strong the pull of parity is from year to year. It turns out that the average team in ’07 was about 64% as good or bad as it was in ’06 relative to its conference’s mean. Conceptually, that change can be described like this...

Expected 2007 Rating = 0.64 * (2006 Rating – 2006 Conference Average Rating)

Here’s an example: In 2006, the team that exceeded its conference’s mean by the most was Memphis-–the Tigers' Pythagorean rating was 1.63 standard deviations better than the CUSA mean. Using that 64% figure, we would have expected Memphis to have been 1.01 standard deviations better than the CUSA mean in 2007. In reality, Memphis did move back towards the middle of the conference last year, but just slightly–-it was 1.55 standard deviations better than the CUSA average, exceeding its expectation by 0.54. This figure will represent our coach rating, since most people associate beating expectations with a good coaching job. Here are the new coaches whose teams, in 2007, exceeded expectations the most.

Anthony Grant (VCU) +0.80
Todd Bozeman (Morgan St.) +0.76
Mike Anderson (Missouri) +0.55
Tom Schuberth (UT Pan American) +0.53
Brad Brownell (Wright St.) +0.52

What this does is put every coach on a level playing field based on expectation. Washington State’s Tony Bennett won a coach of the year award from just about every organization that had one. In a way, Bennett earned the accolades by turning the longtime doormat of the Pac-10 into a three-seed in the NCAA Tournament in his first season at the helm. Realistically, Anthony Grant could not have done what Bennett did in terms of raising VCU into the AP top 10 and garnering a high seed in the NCAA Tournament, nor should Grant be held to that standard. (For the record, Tony Bennett’s effort under this system put him seventh on the list.) Based on what each coach had to work with, the jobs both Grant and Bozeman did deserved recognition. Grant got some recognition by leading VCU to a first-round upset of Duke, and nearly got a financial reward by taking over the Florida job from Billy Donovan. Bozeman, though, got neither for his efforts.

The case of Brad Brownell is an interesting one as well. Brownell left UNC Wilmington, seemingly in good standing, making what would generously be described as a lateral move to Wright State. Brownell turned the Raiders into an NCAA Tournament team after an 8-8 conference record in 2006. Meanwhile, UNC Wilmington suffered the second-worst drop in performance under a new coach.

This brings me to another point-–teams with new coaches tend to struggle. Perhaps it’s not surprising, since most coaching changes occur because the administration at a school is dissatisfied with the state of the basketball program, and the instability that a coaching change brings doesn’t help matters initially. Just 24 of the 60 new coaches last season saw their teams exceed expectations. Eight of those 60 coaches saw their teams rank among the top 25 underachievers in all of D-1. These numbers make the jobs done by Grant and Bozeman all the more impressive.

Most hoops fans are familiar with Anthony Grant’s work at VCU. Based on what Todd Bozeman did last season--and on what Morgan State did in its season opener this year, when the Bears took UConn to the wire before falling 69-65--it’s clear he can coach. Even with all of the baggage, he’ll get a shot at running a better program sometime soon, because he may be one of the best coaches in the game right now.

Ken Pomeroy is an author of Basketball Prospectus. You can contact Ken by clicking here or click here to see Ken's other articles.

0 comments have been left for this article.

<< Previous Article
Surprise, Surprise, Su... (11/19)
Next Article >>
Around the Rim (11/23)

State of Basketball Prospectus: A Brief Anno...
Tuesday Truths: March-at-Last Edition
Easy Bubble Solver: The Triumphant Return
Premium Article Bubbles of their Own Making: Villanova, Temp...
Tuesday Truths: Crunch Time Edition

2007-12-03 - Back and Forth: The Eyes of Texas Edition
2007-11-30 - Headlines You Won't Find Elsewhere: College ...
2007-11-27 - The Frances Pomeroy Naismith Award: Handicap...
2007-11-20 - Measuring Coaching Performance: Bozeman's Ef...
2007-11-19 - Temple of Doom: Fran Dunphy Relives a Coachi...
2007-11-12 - Back and Forth: Tip-Off Weekend
2007-11-09 - The Best of the Rest: Elsewhere in Division ...

Basketball Prospectus Home  |  Terms of Service  |  Privacy Policy  |  Contact Us
Copyright © 1996-2017 Prospectus Entertainment Ventures, LLC.