Drafting with Context: John Calipari

Dec 27, 2014; Louisville, KY, USA; Kentucky Wildcats head coach John Calipari reacts to the officials call during the second half against the Louisville Cardinals at KFC Yum! Center. Kentucky defeated Louisville 58-50. Mandatory Credit: Jamie Rhodes-USA TODAY Sports

My earliest work with draft projections focused on the simple correlation of different box-score statistics between leagues. This work has resulted in some pretty clear results. For example, NCAA players who block shots become NBA players who block shots with precious few exceptions. There is a similarly strong relationship between collegiate and professional rebounding, but in that case at least things like age and strength-of-schedule are important considerations. In other areas, particularly scoring efficiency, the relationship between leagues is not nearly as strong.

My aim here is not to reiterate these patterns, but rather to look at how different contexts might impact these relationships. In particular, I am investigating the correlation of different statistics between college and the NBA while isolating specific coaches. Different coaches run different offensive and defensive systems. Sometimes there are strong theoretical reasons to expect these systems to inflate/deflate a player’s ability in different areas relative to those under a different coach. For example, Briante Weber is collecting a ridiculous number of steals. I assume he is actually really good at that, but is it reasonable to assume he would have the same rate under Bill Self rather than Shaka Smart? Probably not.

This will be the first in a series of posts looking for these patterns. My ultimate goal is to help improve expectations for highly-touted draft prospects. Most of us already do this kind of thing in a loose subjective fashion, but looking at the data and testing assumptions should help fine-tune evaluations.

These analyses use my NCAA-to-NBA Correlations tool, which is available here at NylonCalculus. I will be putting together posts at my convenience, and will limit myself to whatever insights a few minutes poking around can reveal. If you are interested in the findings, I highly recommend you pull up the tool and do some research for yourself. You can look at any coach, any box-score statistic, and filter by position. If you find anything cool, let me know on Twitter @VJL_bball and maybe I will work the finding into future posts.

___________

John Calipari:

Teams: Kentucky, Memphis

Mentors: Larry Brown, Roy Chipman, Paul Evans

Mentees: Josh Pastner, Derek Kellogg, Bruiser Flint, Tony Barbee, Bill Bayno, Chuck Martin, Ed Schilling, Steve Roccaforte

Prospects of note:

Karl Towns, C; Willie Cauley-Stein, C; Devin Booker, SG; Tyler Ulis, PG; Trey Lyles, PF; Dakari Johnson, C; Alex Poythress, SF/PF; Marcus Lee, PF

___________

Patterns:

Rebounds and blocks —

As I noted above, rebounds and blocks are very stable. Those collected under Coach Cal are no exception. As you see in the two plots, Cal’s players’ rebound rates have historically translated into the NBA nearly perfectly. Obviously the rates themselves are expected to decline against tougher competition, but players’ rates relative to other college players hold steady. This is great news for Karl Towns. He is one of the best prospects in the draft at blocking shots and collecting rebounds, and we should expect this to carry over into his NBA career.

I should note that this year’s team may be a special situation. Cousins, Davis, and Noel all played under Cal, but they didn’t play at the same time. The overlap of talent with Cauley-Stein, Towns, Johnson, Lyles, and Lee is extreme. It may be wise to consider the possibility that these players are stealing rebounds and blocks from one another, thus depressing their rates relative to what we should expect in the future.

Personal fouls —

Every NBA big in my data who played under Calipari in college went on to foul less than expected based on his collegiate rate. I assume this stems from the fact that Cal teams play aggressive pressure defense that errs more on the side of accepting the occasional foul than most players will see in the NBA. Interestingly, none of the Kentucky bigs are struggling with fouls this season, even though both Dakari Johnson and Willie Cauley-Stein had issues in the past. I actually would have expected the opposite pattern since Kentucky has comical depth in the paint and thus minimal reason to worry about fouling out, so let’s see if things change as the season progresses.

Passing —

My reading of the Cal-specific assist data is that the dribble-drive offense shares the wealth more than most systems. Players without much passing skills may have their assist-rate inflated under Calipari, while passing savants will have their college rates depressed relative to their eventual NBA role. This is good news for Tyler Ulis whose current rate fits him right between Rose and Wall.

Conversely, Calipari’s offense also seems to hide mistakes. Tyreke Evans (who notably changed positions in the pros) is the only on-ball guard not to see his NBA turnover rate considerably outpace NCAA-based expectations.

More?

That is all I have for this installment. Please, take a look at Calipari in the linked correlation tool and let me know if you find anything else I should comment on. I will be back sometime soon with a post looking at one of the other top college coaches.