Much has been made of Tristan Thompson's improvement as a rim protector from one of the league's worst (Seth Partnow's rim protection metric had Tristan in the bottom three last year if I recall correctly) to an 'average' contributor in this extremely important part of big man defense.
As most that frequent this board know, I have been 'low' on Tristan Thompson's prospects for a long time - in large part because I don't believe in his rim protection or his ability to develop as a rim protector.
Fear The Sword editor David Zavac likes to remind us that Tristan is an average rim protector rather than the poor one that many fans describe him as:
@RickyH23 The Cavs biggest issue is defense. Tristan is an average rim protector who can guard four positions. Hes been very smart this year
— David Zavac (@DavidZavac) May 22, 2015
The definition of 'average' here can be tough. The problem is that Partnow's rim protection metric involves how often a player protects the rim, as well as he does when opponents try to score on them there. On this end, a player who is stationed next to the rim is rated highly - case in point Chris Kaman and his +1.09 points saved per 36 minutes, courtesy of his high contest rate and in spite of his horrendous 52.9 FG% allowed at the rim, a number that is even worse than noted at-rim sieve Kevin Love.
This boost is a combination of Terry Stott's hyper-conservative pick and roll scheme on defense and simply just playing more time at center. Of course, our very own Tristan Thompson has very recently started logging a lot more minutes at center and has started contesting more shots at the rim as a consequence.
It is for this reason that we need to dig deeper when defining the term 'average'. Can Thompson - who played 78.8% of his minutes at center last year - really be compared to guys like Thad Young, Derrick Williams and Danilo Gallinari?
When we look back through Partnow's rim protection metric and look at only players that logged at least 50% of their minutes at center (an arbitrary number I know, but I did pluck it out of the blue rather than cherry pick) then we get this list (with points saved per game in parentheses):
Jared Sullinger: (-0.53), Nikola Vucevic (-0.26), Al Jefferson (-0.20), Robert Sacre (-0.04), Tristan Thompson (-0.01), Mason Plumlee (0.03), Spencer Hawes (0.03), Pero Antic (0.06), Donatas Motiejunas (0.07), Lavoy Allen (0.11), Zaza Pachulia (0.11), Steven Adams (0.22), Al Horford (0.28), Tiago Splitter (0.28), Kendrick Perkins (0.35), Gorgui Dieng (0.35), Marc Gasol (0.54), Chris Kaman (0.57), Aron Baynes (0.61), Bismack Biyombo (0.62), Meyers Leonard (0.64), Chris Andersen (0.79), Jusif Nurkic (0.83), Cole Aldrich (0.88), Omer Asik (0.90), Kyle O'Quinn (0.92), Tyler Zeller (0.95), Brook Lopez (0.96), DeAndre Jordan (1.07), Marcin Gortat (1.11), Derrick Favors (1.17), Alex Len (1.28), Ian Mahinmi (1.30), DeMarcus Cousins (1.35), John Henson (1.36), Andre Drummond (1.39), Tyson Chandler (1.40), Jonas Valanciunas (1.46), Robin Lopez (1.47), Timofey Mozgov (1.50), Tim Duncan (1.56), Pau Gasol (1.63), Serge Ibaka (1.76), Nerlens Noel (1.78), Hassan Whiteside (2.16), Roy Hibbert (2.57), Andrew Bogut (3.13), Rudy Gobert (3.23)
That is a list of 48 players, and Tristan ranks 44th. It appears that when we remove the Gallinari's, Mirotic's and Papinikolau's of the world that Tristan is not average - rather he is far from it.
This isn't to rag on Thompson who is a great pick and roll player offensively and guards his man at an absolutely elite level provided they're not a post-up brute. This also isn't to rag on David - at -0.01 points saved relative to the average player Tristan IS the definition of average in this respect. But when we parse out the positional differences and control for players that just play the same role as Tristan did this year, it appears that Tristan has a lot more growing to do to become average.
"Minutes at center' was taken from NBAwowy.com by filtering for players that the players above shared the floor with. Some definitions are blurry - is Ed Davis or Jordan Hill the center in a lineup that involves both? - but I think for the most part it is a decent proxy.