Position Adjusted Big Ten Rebounding

[Bumped.  I will say, Austin Thornton comes up with a lot of long rebounds you think he has no shot at when he goes after them. -KJ]

[Edit from Mike: David Hess alerted me that I boneheadedly was using 70.2 rebound per game when I should've been using 35.1 instead. One of those facepalm moments on my behalf. Numbers have been fixed and I tired to fix some of the paragraphs but if I missed a few sentences referencing numbers that aren't in the table, I apologize. All of the tables are 100% correct, barring another blunder on my behalf. Thanks for the help, David!]

After my post yesterday, Spartan Dan mentioned grading against position averages and extending this out beyond big men. I, too, had that same thought and was wondering how to attack it. Do I make it just forwards and guards? Break it down by Centers, Power Forwards, Small Forwards, Shooting Guard, Point Guard? Do I make it just bigs, wings, and guard(s)?

What I decided to do was to just take the rosters from ESPN and match them up with the KenPom individual stats under each teams' overall stats. I chose to individually take the players listed on the team pages* because I wanted players with less than 40% minute percentage to be included as well. The players that appear below the team totals on KenPom are those used in at least 12% of possessions and the lowest player in terms of minute percentage I collected data for was Wquinton Smith of Wisconsin.

Now, I chose to break the positions down as forwards, wings** and guards because ESPN also listed some players as Guard-Forwards; I took this as a 'tweener' of sorts and assumed that wing could accurately describe this. I'd assume one could also do this by breaking it down by height which is something I thought about doing. Say, 6-foot-4 and under for guards, 6-foot-5 through 6-foot-7 for wings and 6-foot-8 and above for big men (these are just the arbitrary thoughts I had) but I'm decently pleased with the way it came out, save for the fact that I have only 12 players listed as a "wing" to go with 48 forwards and 50 guards, though it intuitively makes sense. Two forwards (C/PF), two guards (SG/PG) and the remaining spot (SF/a third guard) as a "wing."

I used the same methodology from my previous post, though, as the title suggests, these are adjusted for positions.

If we look at the game plan page for the Spartans, we see at the bottom rebounding distribution for MSU by position and for the entire Division 1 teams by positions as well. I used the 30% offensive rebounding percentage average for C/PF's David Hess estimated in this post. He didn't specify a number for defensive rebounding, but I estimated it at around 25%. To get this down to a per-player average, you multiply the team share of offensive rebounds by a C/PF combo (the 30%) by the average rebound percentage as a team in D1 (32.6%) and that yields 9.8% average for a C/PF player -- these are the forwards in my classifications.

I have these values for the three positional breakdowns I looked at:

Where TS OR/DR% is the Team Share OR/DR% and PL OR/DR% is the individual player average. So, using guards as examples, I estimated around 9.5% of all offensive rebounds for the average D1 team to be grabbed by either of the guards on the floor. Given teams rebound 32.6% of their own misses, we'd expect the average guard to pull down around 3% of his times misses on the offensive glass. Repeat this process for the offensive and defensive ends for each position and that's the above chart.

So what does all of this rambling mean? Rankings! Who doesn't love rankings?

The top 15 players in offensive rebounding, regardless of position are...

Who's on top of that list? Yes, it's a guard that gets less than 50% of his teams minutes, but his OR% sitting at 12.4% is about four times better than the average guard's OR% of 3.1%. If we were to have a forward rebounding offensively at four time his positional rate, that forward would have to have an OR% closing in on 40%. Joshua Smith of UCLA leads the nation at just over 21%, so that's not likely to happen. Still, Indian's Victor Oladipo is rebounding very well for a guard -- even if he is 6-foot-5 which could border of "wing" player.

MSU-wise, I'm not going to even pretend to try to make the argument of Austin Thornton as our best rebounder -- we know that's not likely the case. Still, given his limited use, he's been hitting the offensive glass pretty well. Is this sustainable? I'd venture to guess it's not. Is this indicative of his true talents? Again, probably not. Is he better than Durrell Summers as a rebounding guard? No. But it's nice to give him his due.

As for Spartans-only, here's the list:

The last column is their rank out of the 110 players I have for the Big Ten. We see Thornton and Summers, factoring in their playing time, both grabbing over a rebound more than we'd expect an average player to grab. Payne has also been very good. Delvon Roe's commitment to defense is wonderful and he has looked better in recent weeks, so his rating could/should rise.

Here's the top 15, regardless of position on the defensive end:

The list is dominated by big men, which you'd expect. Only Taylor Battle and E'Twuan Moore are guards in the top ten and Oladipo and Tim Frazier join them inside the top 15. This is where Jared Sullinger really shines. We'd expect the average forward to grab defensive boards 16.9% of the time and he's grabbing them almost 28% of the time. Combined with a high amount of minutes and he's a rebounding machine grabbing about .60 more defensive rebounds than the second best defensive rebounder and nearly six rebounds more than the average forward.

Draymond more than makes up for his average offensive rebounding with quality work on the defensive glass. Probably a testament to his sound positioning more than anything. The rest of the Spartans are here:

Durrell's pretty good on the defensive end as well, grabbing an extra half-defensive board more than average. Austin Thornton kicks in and Nix being the best non-Draymond Green big man on this list and being marginally better than average shows the lack of a quality post player we have defensively-- and Nix has been almost a rebound better than Garrick Sherman yet Sherman gets nearly double the minutes of Nix.

Finally, all players, all positions, top 15 in the conference:

Jared Sullinger's defensive rebounding really propels him. Not that he's a slouch offensively sitting in eighth place in offensive rebounding. No one else grabs one more rebounds above average in the conference other than these fifteen, so this is a nice cutoff.

MSU-only total rebounds are...

Draymond, Durrell, and Thornton really drive the teams rebounding when they're on the court. Unfortunately, the cynic in me must point out that, indeed, an undersized 'big' and two guards have been driving our rebounding. Still, it would be nice, as Garrick Sherman's the reserve big man getting the most minutes to pick up an extra board here or there to help out. Either that or Nix should get some more time -- though I know that we have more problems than just rebounding from the bigs. I don't have this data for years passed, but I wonder how far down (if any) Kalin's rebounding is.

* = KJ, Spartan Dan, LVS, anyone: is there a good place to parse data from (manually I mean. I'm not a programmer nor do I have the necessary computer skillz to rip data from the internet super highway with some code)? Kenpom's is a pain to copy-paste data from if I wanted all of the players he listed on the team pages like I wanted for this project. I guess I could calculate it but I'd rather not waste time trying to format a spreadsheet to calculate this data because that would take a long time given how naive/unfamiliar I am with tempo-free stats versus advanced baseball stats.

** = You can quibble with the positions if you'd like. I just went by ESPN. One man's forward may be another man's wing player. All up to interpretation. Truth be told, I'm actually decently pleased with how it turned out without adjusting the positions any (aside from re-labeling G-F's to Wings). My gut feeling is that there should be more wings than I have listed for sure, but I didn't want to spend the time trying to classify 110 players to what I think they should be.

I welcome any and all suggestions or rotten tomatoes you'd like to throw my way!