The Search For Balanced Coaching

There is a recurring statement made in the comments section of Barking Carnival game recaps that goes something like “Shaka Smart sucks at offense”, or “Shaka can’t coach offense”, or “Shaka sucks at coaching offense”. This happens nearly every time the team wins and absolutely every time the team loses; it’s not just at BC, I see it over at Burnt Orange Nation, on Reddit, Twitter, and a number of other forums. It’s the same complaint that was lodged against Rick Barnes many years, though with him they tended to say he was overly-reliant on having a superstar point guard to make his offense function. Texas fans have been saying some derivation of this about their head basketball coach since the advent of social media, and likely were saying it in forums before that. There was probably a Usenet group rec.sport.basketball.college.texas.penderssucks back in the day, instead it complained about the lack of defense. Texas fans are not historically a happy bunch.

Most Texas fans agree that Shaka can build a cohesive defense, a defense that gets better as the season progresses, and a defense that can rise to the level of truly elite when everything comes together (like this year). The Shaka detractors tend to focus on the offensive side of the ball; they desire an offense that is as good as the defense. But how many coaches out there can develop an offense that is on par with their defense, or a defense that’s on par with their offense? There are a lot of D-I coaches with a good reputation on one end of the floor, how many who are considered complete coaches? ENTAR TEH SPREADSHEETZ

I mulled over a way to quantify coaches who should be considered ‘complete’ on both the offensive and defensive side of the ball, and as usually happens I ended up digging through data on Ken Pomeroy’s site. He has a 17-year pool of data to draw from, so it gives me a reasonable sample size to tinker with. What I ended up doing is pulling down the teams ranked in the top 10% in both adjusted offensive efficiency (AdjO) and adjusted defensive efficiency (AdjD). Each of these stats measures how a team performs relative to the average D-I team in a season. For AdjO; the higher the number is the better; for AdjD, the lower the number is, the better. So if a team has an AdjO of 105, they’re 5% better than the national average, and if their AdjD is 105, their defense is 5% worse than the national average. I like these stats because the stats don’t reward teams that run just because they run or punish a slower team that executes well just because it scores fewer points; it’s effectively tempo-free data. Pomeroy then ranks all of the D-I teams in each statistical category, so I pulled down every team each year that ranked in the top 10% for that year. For example, this season there are 351 teams in D-I, so I snagged all of the teams that ranked in the top-35 of both categories. (In other seasons there were fewer teams in D-I, so the numbers changed a bit; for example, in 2003 there were 327 D-I teams so I rounded 32.7 to 33 and pulled in teams ranked in the top-33 of both categories.) The data I used for this season is as of 01/15/2018, so the rankings may change a bit based on games played after ~7 PM CT last night.

All told, there were 222 teams in the 17-year span from the 2001-2002 season to the 2017-2018 season that were ranked in the top 10% of both AdjO and AdjD. 67 coaches are responsible for those 222 teams, with 36 coaches appearing more than once. From there, I wanted to filter down a bit more to see if I could suss out the coaches who are more consistently able to field top-level offenses and defenses. If the idea here is to find consistently complete coaches, seeing a name show up once in 14+ years probably doesn’t meet the definition, even if it means removing names that might surprise us (hello, Lon Kruger & Bruce Pearl). I settled on picking coaches who had been head coaches for at least five years in the last 17 and showed up at least 20% of the time. This removes some of the issues with small sample sizes, but it also means a coach like Greg Gard (once in three years) doesn’t make it onto this list just yet. With that said, here is your list of coaches who qualify:

That’s the 25 most consistent coaches in the last 17 years, 16 of which (highlighted in green) are active in D-I right now. The names at the top probably aren’t much of a surprise, they’re the coaches most people associate as being the best in the business. Nearly every name in the top 10 will be in the Hall of Fame sooner than later, and many of the other 15 are going to have their names called eventually. These 25 coaches account for 169 of the 222 teams listed in the spreadsheet, which means 37.3% of the coaches who have fielded a top 10 percent team account for 76.1% of the teams in the list. In other words, this list is pretty top-heavy. There aren’t a lot of coaches out there who are consistently developing balanced teams. Well, consistently elite balanced teams; there are a number of coaches out there who are fielding teams that are consistently bad. :)

(Side note: 10 of the 25 coaches listed spent their entire tenure in the last 17 years coaching a single team, hence the ‘1 team’ remark to the right of their stats.)

What about Shaka Smart? Here are his rankings by year:

As you would expect, Smart cracks the top 10% on the defensive side of the ball more often than not. His offensive numbers aren’t as good, but they’re also not terrible most years. He had one year where he barely missed making the top 10% list, ironically it was the defensive end that let him down (he was in the top 10.26% that year). His average year over the last 9 years puts him in the top 22.22% in AdjO and top 10.54% in AdjD, where his average AdjO rating is 110.02 and his average AdjD rating is 94.41. Those numbers are well above-average, but the offensive side keeps Smart from being within arm’s reach of the elite. Quelle surprise.

There are limitations to the predictive nature of this data, I can see examples that don’t fit neatly into this rating system. For example, the year Rick Barnes got fir-err, resigned from Texas, he was 47th in AdjO and 14th in AdjD. Perhaps someone could use this as a launching point to further refine and/or alter this concept to make a more definitive view of which coaches are the most balanced in their approach. Still, it’s interesting to tinker with and can maybe help slot a coach more specifically than “Shaka sucks”; now you can say “Shaka sucks compared to Bill Self” and a few more people might nod along. Maybe this data can nudge all sides a little closer to consensus — or at least a temporary truce — on where Smart ranks among the D-I landscape.