For all your fancy-pants statistical needs.

Praise for The Basketball Distribution:

"...confusing." - CBS
"...quite the pun master." - ESPN

Players of the Month: December (so far)

Welcome, all! Here I will be grading players according to their estimated offensive and defensive impacts (using my per-100-possession stat, SimplePlayerRating) via Month-of-December-State. I'm rolling out my Defensive SPR here, finally. Formula at the bottom.

EDIT: Fixed the per-game numbers.

Surprises of the month go to: Andray Blatche (#6 #7), Paul George (#9 #10), Kemba Walker (#10 #11) and JJ Hickson (#17 #21!!!).

Without further ado, here are your top and bottom 26.

The Top 26

PlayerSeasonOSPRDSPRTotal SPRSPR per Game
1Carmelo Anthony2012-1310.6-
2LeBron James2012-
3Blake Griffin2012-
4Kevin Durant2012-
5Chris Paul2012-
6Kobe Bryant2012-136.4-
7Andray Blatche2012-
8Ryan Anderson2012-137.9-
9Tony Parker2012-137.3-
10Paul George2012-
11Kemba Walker2012-136.4-
12Chris Copeland2012-1312.
13Russell Westbrook2012-
15Stephen Curry2012-135.0-
16Paul Millsap2012-
17James Harden2012-
18Tyson Chandler2012-
19Andrei Kirilenko2012-
21J.J. Hickson2012-
20Matt Barnes2012-
22Ed Davis2012-
23Dwyane Wade2012-135.3-
24David Lee2012-
27Kyrie Irving2012-135.9-
26Eric Bledsoe2012-

The Bottom 26

PlayerSeasonOSPRDSPRTotal SPRSPR per Game
387Daniel Gibson2012-13-5.5-1.7-7.2-3.9
386Kyle Singler2012-13-4.1-1.8-5.9-3.8
388Doron Lamb2012-13-8.9-1.6-10.5-3.8
382Mickael Pietrus2012-13-4.6-1.1-5.7-3.5
383Andre Iguodala2012-13-5.30.7-4.6-3.4
379Chris Singleton2012-13-5.90.1-5.8-3.3
381J.R. Smith2012-13-3.6-1.5-5.1-3.3
378Andrea Bargnani2012-13-3.2-2.2-5.4-3.2
377Victor Claver2012-13-11.32.0-9.3-3.2
376Jeff Taylor2012-13-3.1-2.3-5.4-3.1
374Jerry Stackhouse2012-13-4.0-2.6-6.6-3.0
375Bismack Biyombo2012-13-5.00.5-4.5-3.0
372Alonzo Gee2012-13-3.3-0.8-4.1-3.0
373Gerald Green2012-13-4.7-1.9-6.7-2.9
370Willie Green2012-13-5.6-2.2-7.8-2.8
369Festus Ezeli2012-13-7.6-0.9-8.5-2.6
368Dahntay Jones2012-13-4.1-2.4-6.5-2.6
367Austin Rivers2012-13-2.6-1.7-4.3-2.6
365Keith Bogans2012-13-9.1-2.6-11.7-2.6
364Sebastian Telfair2012-13-3.2-2.4-5.6-2.4
363Aaron Brooks2012-13-2.7-1.6-4.3-2.4
361Martell Webster2012-13-2.8-1.0-3.8-2.4
362John Salmons2012-13-1.8-1.8-3.6-2.4
359Jason Maxiell2012-13-5.41.6-3.8-2.3
360Tony Allen2012-13-6.11.5-4.6-2.3
358Nolan Smith2012-13-6.5-2.8-9.3-2.3

-The formula for DSPR is
DSPR = (1.3xSteals - 0.1xMissedFG + 0.2xDRB + 0.5xBLK)x100/Possessions Played - 3

-OSPR can be found here.

An Apology for my Quietude: EZ Score

SORRY FOR THE DELAY. I have been hired by the Losangephoenix Spurockets to do basketballysis!

Just kidding.
I've actually been pretty busy doing other church-music related things. Not my bball-twitter-peeps kind of material, I know.

I've had ten or so blog posts in the works, none of which ever reached fruition.
Displeased with my blog production level ( < 20%), I decided to post what I think could probably have made me the most money (I don't know, a couple dollars?) had I decided to streamline and sell it.

Yes, my compassion and eagerness outweighs my entrepreneurial sense. And yes, I did have to spell-check "entrepreneurial."

With sincere apologies to Evan Zamir who owns a 51% market share on the term "EZ" in the basketball-stats world, I present "EZ Score." EZ Score is a game charting system that takes into account every possession (so it requires a bit of rewinding your recorded video), and every player on your team. It is pretty simple to describe, but a little bit open to interpretation. If anyone cares, I can post the Excel-specific nitty gritty on how to accomplish it, but here are the main basic details:


This is a system that imitates Dean Oliver's offensive and defensive rating system, although it is a little bit more intensive in that every possession (both offensive and defensive) must be charted. Credit is only given to whomever directly contributes to the possession result. This is pretty wide-open to interpretation, but generally I follow it like so:
Everything past #1 for each data point could be optional if you want, but it will give you less-refined results.

Each possession must be entered manually, simply by entering the player's jersey number like so:

A dream-team including MJ, Hansbrough, and Penny.

P1=Most responsible / directly responsible for the possession result
P2=Less responsible than player #1
P3=Less responsible than player #1

If only one player truly deserves credit, only enter one player. My excel sheet distributes credit accordingly to the between 1 and 3 players(weights are noted at the end of this section).


Good possession (2+ points):
P1) Whoever scores
P2) Pass or screen or offensive rebound leading to score
P3) Pass or screen or offensive rebound leading to #2

Normative plays (1 point):
P1) Whoever scores. Optionally, the assister/etc can receive credit as P2 and/or P3, but this depends on your philosophy (is it the passer's "fault" that the player misses a free throw?, etc)

Bad plays (0 points):
P1) Turnover, missed field goal
P2) Not boxing out/missing easily available rebound

Good possessions (0 points):
P1) Forced field-goal miss or defensive rebound (if more causal than #2), fouls, forced turnovers
P2) Forced field-goal miss or defensive rebound (if less causal than #1), fouls, forced turnovers, help defense
P3) Same as #2

Normative possessions (1 point):
P1) fouler gets 100% credit

Bad possessions (2+ points):
P1) Your man or your zone scores / fail to switch / etc
P2) If a man is wide open due to #1, whomever helps, etc receives #2.
P3) Same as #2

Now, the weighting. In excel, I weight every possession depending on the # of contributing players
If there is 1 player, they receive 100% of the score & possessions
p1=(100% * points, 1 possession)

If there are 2 players, the first receives 66.67% of the score and possession, the second receives 33.33%.
p1=(66% * points, 0.66 possessions), p2=(33% * points, 0.33 possessions)

If there are 3 players, the first receives 50%, and the second two receive 25% apiece.
p3=(50%*points, 0.5 poss), (25%*points, 0.25 poss)

So, there are many obvious small changes one could make. Many of these would increase the work of the charter and might not necessarily be necessary; it's a balancing act. The most obvious to me is the ability to choose in the two or three-player scenarios between ranked and equal weights for player 2 & 3 (i.e. the ability to say that a scorer-screener-assister are weighted something like 50-30-20 rather than 50-25-25) But I would love to hear your suggestions. are my results for USA v. France in the Olympics this year. This took maybe 30 minutes more than it would have, had it been a regular game-watching experience.

I'll try to do this for a few games this year as my free time permits. Use responsibly!

Introducing SPR (Offense)

If you're like me, and you run regressions of statistics against multi-year regularized-adjusted plus-minus in your spare time, you'll note that there are numerous ways to come up with good estimates. Wanting to create a system that requires as little math as possible, I have found combinations that can estimate offensive value well, while not requiring people to utilize long strings of decimals.

So I've decided to finally release my Simple Player Rating (for offense) to the public. Here it go!
"SPR: Offense" represents an estimate of how much a player boosted their team's offensive rating above average.

SPR: Offense =
(Points - FGmade - Turnovers + 0.5 x (Oreb + Assists - FGmissed - Free Throw Attempts))*100/Possessions Played - 6.5

a)To come up with possessions played, you can find this on any box score:
Possessions Played = Pace x (Minutes Played) / (Game Minutes, Usually 48)

b) Also, if you want to convert this to a per-game statistic, simply multiply by:
(Minutes Played) / (Game Minutes)

Here is how this stat correlates with 8-year-RAPM:

In case you were wondering, with the raw values that my regression gave me the R^2 is only 0.01 higher.

And here are the per-game and per-100 SPR Offense from Miami vs Boston Game 6, aka "BRON-BRON GOES NUTS."

Defense is coming soon. Enjoy responsibly!

New Stat: SLEASY%

SLEASY% (Super-Lazy-Estimate-of-Assisted-Shots,Yo%)

Based on Dean Oliver's realization that a player's % of shots assisted on is extremely close to:
=1.14*(Team Assists - Player Assists)/(Team Field Goals Made)

I plugged this into a large NBA player dataset, where:
Team Assists per 100 = 5*Average(Assists per 100)
Team Field Goals Made per 100= 4*Average(Field-Goals-Made-Per-100)+playerFGMper100

The logic is a little hazy, especially considering that the "team" assists are made up of 5 average players, but we are subtracting a non-average player. So it's lazy.

OK, from here, we can get a linear regression. The correlation is EXTREMELY high, but definitely misses the mark on a few players with very high field-goals made or assists dished per 100:

SLEASY% = 0.73 - 0.03xAssistsPer100 - 0.01xFieldGoalsMadePer100

In my 8-year dataset for the NBA through 2011, we find the following top and bottom 10:

Pretty straightforward...low usage players up-top, frequent shooting PGs at the bottom. Can't assist yourself! Or can you ??

2012 Bracket Cheat Sheet

Why yes, it is the most wonderful time of the year.

I've simulated the NCAA tournament 10,000 times with the following specifications

-Brigham Young out. Iona is out. I mistyped.
-Miss. Valley State out.
-Injured/disqualified players out.
-Minutes adjusted (better players getting more and vice versa).
-Home-court advantage adjusted for.
-Lucky efficiency accounted for (somewhat)..i.e. TO% and Defensive Rebound% are worth more than "usual," for example.

Using all of this, I created a bracket cheat sheet. Guaranteed to beat Pomeroy, Sagarin, and the LRMC!*

Enjoy, and use responsibly. Don't gamble with this stuff.

*-OK, not really guaranteed.

The Orange Are Okay.

Before we begin: credit to the numbers here go out to Daniel Myers @DSMok1 and his NCAA statistical plus-minus sheet. I have my own version (which I made about a day before his :), but it didnt' respond to Sports-Reference's reformatting very well.

If we assume that Melo's minutes get replaced by Christmas, and also slightly replaced by James Southerland, we see the following:

Efficiency Margin per 100 (with Melo having high Minutes%):  31.7
Efficiency Margin per 100 (with Melo's minutes replaced by Christmas, then Southerland): 30.4

This is a difference of -1.3 points per 100 possessions, or just -0.82 points per game, in total point margin. So Syracuse moves from a (rough statistical estimate:) #6 team to a #9 team.

Why does this happen?
I will show you.

I am just making some guesses here, but:

-Melo was playing 58% of Cuse's minutes beforehand...he missed a few games. So I bump this up to 74% to compensate/guesstimate.
-Also to compensate, I bump underrated James Southerland's minutes down by 0.1%, and Christmas' minutes down by 16%.

So we then get the following:

Orange are okay! Just probably underrated by impressionable bracketeers.

All points are not created equally...

...therefore not all points predict the future equally.

The True Value of the Four Factors

Dean Oliver's four-factors are well-understood in how they impact the game.
To figure this out, people usually run a regression of game or team four-factors versus efficiency.

We find the following, roughly:
raw coefficients
However, to say that these are the proper weights for each of these is to assume that each of these is just as controlled by the offense as the defense. Ken Pomeroy has been blogging on the subject and it got me thinking: there is no way that these can be the true (relative) PREDICTIVE values for four-factors.

I ran a LOOCV test for each 2010-11 NCAA team, for each game they played (for both teams).
In layman's terms: I looked at each game, then averaged all the four factors for the other games a team played.

The results look may only slightly different. The difference is, in fact, huge.

predictive coef

The differences can be outlined here (x = predictive / raw) :


This is now the basis for my ratings. The strength of schedule part is a bit of a guess, but the results are so different no matter what I do, I can't really tell if it's working :)

All this to say: Murray State is actually more overrated than we thought (#227 in offensive TO%, somehow).

Conference Bias in the Polls: No Mountain-West Love

"The Nation's Best Point Guard" apparently.

I did a very quick study on how conferences/teams are underrated or overrated in the polls.

First, I combined the ESPN and AP Polls from Monday.
Here's a simple formula based on regression to add the vote totals from both polls for NCAA basketball:

Total Adjusted Votes= AP votes + (2*ESPN Votes+26)

The top-27 in this method is shown left.

All teams worse than #27 were given a simple "#40" as a placeholder . I then compared each of these team's rankings to my own ranking system, which is very similar to Ken Pomeroy's...I can just fiddle with it as I please :). From this, it's pretty easy to see that teams that have won a lot of close games are overrated and vice versa (if you trust us stat-geeks on the principle of 'luck').

To convert Luck (Win% - Expected Win%) to ranking deviation (Poll rank - my rank), I found the following easy equation:
Expected Ranking Deviation = 140 * Luck% - 0.38

So for each conference, I found the average luck of my top 49 (I stopped at 49 because that's where Murray is ranked) and found the following. I split the rankings into three categories

-Murray State (the only OVC team here, and the most overrated team in the country)
-Conferences (with at least 2 listed teams in the poll)
-Other Conferences (the total of each of the conferences with one bid)

Like so:

#Top 40 Avg Luck
Murray State10.138
Other Mid-Majors5-0.002

Then I used the Expected Ranking Deviation formula to come up with each 'conference's average expected ranking deviation:

Murray State118.9
Other Mid-Majors5-0.7

Then I simply subtracted each team's Expected Ranking Difference from their Actual Ranking Difference to give us an estimate of conference bias. The results are pretty intuitive:

#Top 40 LuckExpRkDiffActualRkDiffBias
Murray State10.13818.933.014.1
Other Mid-Majors5-0.002-0.7-5.4-4.7

The only real surprise to my eyes is the Big 12 having a bias of -2.3. Murray State is overrated by their ridiculously easy schedule which they haven't beaten to a pulp. The only four other overrateds are the rest of the power conferences minus the Pac-12, who is still in "recovery mode."

The Journey of Jeremy Lin

Via InfographicWorld

Click image to enlarge

The Journey of Jeremy Lin
Source: Infographic World

How To Adjust Game Results for Garbage Time (NCAA)

UNC's Blue Steel scrubs. Photo 100% stolen from
There is a fairly straightforward, rule-of-thumb way that I have used in the past to adjust for "garbage time" - basically an effective way of smoothing out the end of games where:

a) the scrubs begin playing
b) teams try and make a comeback despite a mighty deficit...tons of free throw shooting that could skew the point margin in either team's favor
c) the winning team doesn't care about playing anymore, really...
d) etc

Bill James (of baseball stats fame) introduced his own simple metric that uses, and succeeds rather invariably: Lead "Safeness." Bill James also refers to Coach K as "the human typo" in that article. I love him.

Anyways, the math is pretty straightforward (I remove the bit about who has the ball to make it easier to calculate, as it averages to zero):

     (Point Margin - 3)^2  /  Seconds Remaining = % Safe

So we do a little algebra, and we can quickly estimate point margin when the lead became 100% safe:

     Point Margin =Sqrt(seconds remaining when safe) + 3

Which we then use to forecast the rest of the game, ignoring how it actually played out:

     Adjusted Final Margin = [sqrt(Seconds Remaining when safe)+3] x 40 / [40-(Seconds Remaining when safe/60)]

You can easily find "seconds remaining when safe" via For example, last night's UNC-Wake Forest Game was "statistically over" with 3:48 left to go. Plus this in to my handy-dandy spreadsheet, and you get:

Point Margin When Safe: ~18
Adjusted Final Point Margin: 20

So while the Heels only won by 15, they had plenty of room to let the game slide in the final minutes - if they had played with the same quality all the way through they would have won by around 20.

NCAA Teams' Bench Impact

Bench Impact = estimated total Bench Rating per 100 possessions x %Minutes - estimated Starters' Rating per 100 possessions x %Minutes

-Wyoming/Colorado's bench impact is low because of true disparity between their starting 5 and their bench.
-Ohio State's bench is actually decent, but their starting 5 is significantly better....
-Grambling suffers immensely by not playing Quincy Roberts more -- EDIT -- this was because of the transfer rules that prevented Q-Rob from playing the first half of the season, so technically he was on their "bench" when I ranked his minutes played back in January. His presence was definitely the most notable discrepancy in minutes played and overall value because of that.

Bottom 25

teambench impact
2South Carolina Upstate-10.31
4Green Bay-9.14
5Alabama State-9.08
6Florida International-9.01
7Ohio State-8.96
8Northern Iowa-8.90
11Cal Poly-8.62
13Texas State-8.56
14Tennessee State-8.18
15South Carolina-7.94
18Jackson State-7.84
19Southeast Missouri State-7.78
24St. Francis (NY)-7.57

Top 25

teambench impact
2South Alabama2.79
4Colorado State2.50
5Loyola Marymount2.46
6North Carolina-Greensboro1.75
7Texas Southern1.53
8Oklahoma State1.39
9Western Carolina1.03
10William & Mary1.00
13California-Santa Barbara0.85
14Maryland-Eastern Shore0.68
15Western Michigan0.51
17Southern Mississippi0.41
18Alcorn State0.35
19Southeastern Louisiana0.35
20Southern Methodist0.28
22Rhode Island0.26
23Miami (FL)0.24
25East Tennessee State0.13

Top 25 Players as of 1/18/2011

I have adjusted minutes% for teammate quality and injuries (I factor back in minutes played above or below what we would expect a coach to play them with a number I call SQZ: how much more or less a coach squeezes out of a player over the course of a game.)

True Impact per Game = Efficiency Impact x (expected min% while healthy, with average teammates)

 rankplayerteamORTGusage%DRTGTrue Impact per Game
1Kevin JonesWest Virginia129.624.292.59.6
2Jared SullingerOhio State128.926.3769.4
3Damian LillardWeber State136.
4Marcus DenmonMissouri140.523.293.98.6
5Thomas RobinsonKansas117.328.278.78.5
6Anthony DavisKentucky140.51873.68.1
7Doug McDermottCreighton127.731.61027.9
8Kenny BoyntonFlorida135.924.6102.87.6
9Dominique MorrisonOral Roberts13125.4103.17.4
10Will BartonMemphis1192694.47.3
11Cody ZellerIndiana13422.2907.2
12J'Covan BrownTexas122.628.2100.77.1
13John ShurnaNorthwestern117.527.4102.47.0
14Mike ScottVirginia127.128.782.36.9
15Jae CrowderMarquette124.623.384.66.7
16Isaiah CanaanMurray State130.326.495.86.6
17Deshaun ThomasOhio State124.823.889.76.6
18Hollis ThompsonGeorgetown12921.3936.6
19John JenkinsVanderbilt127.126.31016.6
20Nate WoltersSouth Dakota State121.230101.96.4
21Ricardo RatliffeMissouri138.222.491.56.3
22Jeremy LambConnecticut120.923.61006.2
23Drew CrawfordNorthwestern119.125.5104.76.2
24Jordan TaylorWisconsin115.824.586.76.2
25Jason ClarkGeorgetown117.526.189.86.0

Top 25 Freshies, as of 1/11/2012

EDIT: Misleading title. This only includes games through 1/11, not 1/13.

By team impact (estimated efficiency margin impact times % of possessions played).

rankplayerteamconfORTGusage%DRTGEff Imp.Team Imp.
1Anthony DavisKentuckySEC13418719.376.79
2Cody ZellerIndianaBig Ten13422869.636.50
3Kevin PangosGonzagaWCC13221977.755.88
4Michael Kidd-GilchristKentuckySEC12021855.964.60
5Sheldon McClellanTexasBig 1213221967.174.58
6Spencer DinWiddieColoradoPac-1212721965.963.76
7Trey BurkeMichiganBig Ten11025974.303.67
8D'Angelo HarrisonSt. John's (NY)Big East116241023.813.32
9Austin RiversDukeACC108261034.473.32
10Seth TuttleNorthern IowaMVC12319896.143.23
11Kentavious Caldwell-PopeGeorgiaSEC10927994.153.18
12Adam SmithNorth Carolina-WilmingtonCAA116271114.083.09
13Jonathan HolmesTexasBig 1212619945.533.06
14Anthony DrmicBoise StateMWC12324984.663.06
15Omari GrierFlorida AtlanticSun Belt130211056.583.04
16Justin EdwardsMaineAEC12026993.973.01
17Otto PorterGeorgetownBig East11616884.192.85
18Jordan TolbertTexas TechBig 1211331964.682.76
19Rodney HoodMississippi StateSEC125171003.192.69
20Andre DrummondConnecticutBig East11220943.832.59
21B.J. YoungArkansasSEC11727944.322.56
22P.J. HairstonNorth CarolinaACC12426937.682.50
23Quinn CookDukeACC137201027.872.48
24Quincy MillerBaylorBig 1210925884.022.34
25Michael CaffeyLong Beach StateBig West12015984.342.33

Top 51 Overall NCAA Players as of 1/10

Games through 1/10.
Team Impact = (Estimated Efficiency Impact per 100 possessions) x (% of Possessions played in)
Where offensive and defensive efficiency impacts are estimated* by using offensive and defensive rating (& offensive usage%) at, and strength of schedule from

rankplayerteamconfOffensive Impact (100)Defensive Impact (100)Total Impact (100)Team ImpactMin%
1Damian LillardWeber StateBig Sky13.5-0.413.111.084%
2Kevin JonesWest VirginiaBig East7.
3Doug McDermottCreightonMVC11.1-0.610.58.177%
4Dominique MorrisonOral RobertsSummit9.0-
5Jared SullingerOhio StateBig Ten8.04.812.87.861%
6Marcus DenmonMissouriBig
7Isaiah CanaanMurray StateOVC8.
8Will BartonMemphisCUSA6.
9Thomas RobinsonKansasBig
10Kenny BoyntonFloridaSEC10.0-
11Mike ScottVirginiaACC7.
12Jae CrowderMarquetteBig East6.
13Anthony DavisKentuckySEC4.
14Cody ZellerIndianaBig Ten7.
15Nate WoltersSouth Dakota StateSummit8.2-
16Jeremy LambConnecticutBig East6.
17Noah HartsockBrigham YoungWCC5.
18Colt RyanEvansvilleMVC6.
19Jarrod JonesBall StateMAC6.
20Larry AndersonLong Beach StateBig West5.
21John JenkinsVanderbiltSEC7.
22Doron LambKentuckySEC7.
23Ryan BroekhoffValparaisoHorizon7.
24Ryan KellyDukeACC7.
25Hollis ThompsonGeorgetownBig East6.
26Deshaun ThomasOhio StateBig Ten6.
27Kevin PangosGonzagaWCC6.
28Drew CrawfordNorthwesternBig Ten6.9-
29Steven PledgerOklahomaBig
30Kris JosephSyracuseBig East5.
31C.J. McCollumLehighPatriot6.
32Jordan TheodoreSeton HallBig East4.
33Brian ConklinSaint LouisA-
34J'Covan BrownTexasBig
35Reggie HamiltonOaklandSummit7.8-
36Chase TapleySan Diego StateMWC5.
37Langston GallowaySaint Joseph'sA-
38Allen CrabbeCaliforniaPac-
39John ShurnaNorthwesternBig Ten6.
40Deonte BurtonNevadaWAC6.
41Zack RosenPennsylvaniaIvy6.1-
42Erick GreenVirginia TechACC6.
43Julian MavungaMiami (OH)MAC4.
44Draymond GreenMichigan StateBig Ten3.
45Chace StanbackNevada-Las VegasMWC6.
46Anthony RaffaCoastal CarolinaBig South4.
47Robbie HummelPurdueBig Ten5.
48Trevor RelefordAlabamaSEC4.
49Ricardo RatliffeMissouriBig
50Robert CovingtonTennessee StateOVC6.
51Tyler ZellerNorth CarolinaACC4.

So, Lillard is the primary reason that an otherwise-extremely-mediocre offensive team (Pomeroy subscribers only) is #57th in the Pomeroys for offense. Weber's second-best offensive player (Scott Bamforth) uses 10% usage less, and is 10 points per 100 less efficient.

When I sum up all of Weber State's offensive contributions, we get +14.4. Their adjusted offensive rating is only +8 or so on, so it's my guess that (but if we just fit players' performances to their team ratings, we get some extremely ugly results, so I don't do that...I am basing it on last year's fit for Strength of Schedule, however). Even if we multiply each player's offensive value by 0.6 (which is roughly the same as 8/14.4), Lillard is still a top-25 player.

Mike Scott is still also an extremely good player as far as the eye can see, and is even underrated by his minutes (like Sullinger).

* Not going to reveal the regression equation just yet. Or ever :-)

No Apologies, Just Numbers

Okay, some apologies.

My prior post has sparked a little bit of interest/intrigue/outrage/fear among college basketball fans.
Never fear! Happier numbers are on the way.

First, let me reiterate that these are based on advanced stats only, such as offensive rating and defensive rating, from There are no plus-minus stats involved.

I realized that there were some pretty important adjustments that needed to be made. First, that players with extremely high numbers in any one category were being skewed.* Second, that the strength of schedule adjustment was a little counter-intuitive, and a little too strong. I adjusted my ratings thusly.

Unfortunately, the sample size for this season does not produce very intuitive results at the moment, which I should have mentioned in my last post (so please don't destroy me for anything that looks strange...just chew on it). For example, that Weber State's Damian Lillard is number one in the system. While I am a huge fan of Lillard and would have no problem listing him up there at the end of the year if things remain the same, I am wary of the fact that his team impact is two full points higher than the next-highest (and his per-possession impact is higher than Sullinger's). As major-conference teams improve their strength of schedules during conference play and players get more minutes, things will even out.

As proof that things will even out, here is last year's top-10 in terms of overall Team Impact (efficiency impact times minutes-on-the-floor%). We can call this the Holy Grail 3.0

Player School Impact(100)  Team Eff. Impact
Kemba Walker Connecticut 13.1 12.1
Jordan Taylor Wisconsin 12.4 11.2
Talor Battle Penn State 10.3 9.8
Jimmer Fredette Brigham Young 11.0 9.7
Jared Sullinger Ohio State 12.2 9.6
JaJuan Johnson Purdue 10.4 9.2
Derrick Williams Arizona 11.8 8.7
Jon Leuer Wisconsin 10.4 8.7
Ben Hansbrough Notre Dame 9.7 8.4
Jon Diebler Ohio State 9.3 8.3

I hope most people wouldn't have a problem with this list.

So here are the current numbers, which are now split into offense and defense...separated by the Top-26 Major Conference Players and the Top 26 not-so-Major Conferences. (Had to include the older Zeller. It's the law.)



Notice I refrained from ranking players by their per-possession numbers...I think these are more intuitive and represent Player-of-the-Year candidates (if your coach didn't play you enough, you unfortunately didn't impact your team enough).

And look, Hummel is on a list! Now everyone can be happy. Even me!

* My method involved dividing advanced statistics such as Steal%, Assist%, or Offensive Rating by their league average. My NBA numbers had extremely high sample sizes and didn't have any problem with skewed we have that problem a lot more, especially in smaller categories (such as Steal%) rather than larger categories (the more all-encompassing "Offensive Rating").

The Holy Grail 2.5: NCAA Player Ratings (1/4/2012)

EDIT: I have updated my methodology and made things a little easier to understand, I hope:

Now that uses the same advanced stats for NBA and College players, it has become very simple and straightforward to create a statistical +/- based on advanced stats for college players.


In order to adjust for college, I simply use each advanced stat* divided by its league average. Then, applying the same to college works well. I fit the 2010-2011 season's worth of raw statistical +/- data to each team's efficiency margin, and then to their adjusted efficiency margin. I then built a model that combines raw statistical +/- and team strength of schedule (simply Adjusted Efficiency Margin mins Raw Efficiency Margin, via to give us an "adjusted Statistical +/-" that we will call (as we have before) Efficiency Impact.


For each player, we can look at Efficiency Impact per 100-possessions, and Team Efficiency Impact(their efficiency impact times the % of their team's possessions they played in).

Here are the top 100 players in both of those categories. First, efficiency impact per 100 possessions (25% of teams' minutes played to qualify)

rankplayerteamEfficiency Impact/100
1Jared SullingerOhio State19.4
2Russ SmithLouisville16.8
3Thomas RobinsonKansas15.8
4Damian LillardWeber State15.8
5Mike ScottVirginia14.9
6Dion WaitersSyracuse14.7
7James SoutherlandSyracuse14.0
8JaMychal GreenAlabama13.9
9Cody ZellerIndiana13.7
10C.J. McCollumLehigh13.6
11Marcus DenmonMissouri13.4
12Anthony DavisKentucky13.2
13Jae CrowderMarquette13.2
14Jared BerggrenWisconsin13.1
15Isaiah CanaanMurray State12.9
16Jarrod JonesBall State12.8
17Brian ConklinSaint Louis12.5
18Jamaal FranklinSan Diego State12.4
19Ryan PearsonGeorge Mason12.1
20Herb PopeSeton Hall11.7
21Doug McDermottCreighton11.5
22Arsalan KazemiRice11.4
23Kevin JonesWest Virginia11.4
24Anthony RaffaCoastal Carolina11.3
25Miguel PaulEast Carolina11.3
26Chase TapleySan Diego State11.2
27Ricardo RatliffeMissouri11.2
28Ryan EvansWisconsin11.1
29Henry SimsGeorgetown11.1
30Draymond GreenMichigan State10.9
31Cody EllisSaint Louis10.8
32Luke MartinezWyoming10.6
33Jereal ScottStephen F. Austin10.6
34Noah HartsockBrigham Young10.5
35Mike MoserNevada-Las Vegas10.5
36Tyler ZellerNorth Carolina10.4
37Carlos LopezNevada-Las Vegas10.4
38Quinn CookDuke10.3
39Will BartonMemphis10.3
40Kris JosephSyracuse10.3
41Rob JonesSaint Mary's (CA)10.2
42Dominique SuttonNorth Carolina Central10.2
43Jordan TaylorWisconsin10.2
44Jason ClarkGeorgetown10.2
45Ian HummerPrinceton10.2
46Steven WernerSam Houston State10.1
47Leonard WashingtonWyoming10.1
48Robert CovingtonTennessee State10.1
49Drew GordonNew Mexico10.0
50P.J. HairstonNorth Carolina10.0
51Trevor RelefordAlabama9.9
52Tony MitchellAlabama9.9
53Ryan KellyDuke9.8
54William BufordOhio State9.8
55Terrell HollowayXavier9.7
56Kenny BoyntonFlorida9.7
57Harrison BarnesNorth Carolina9.7
58John HensonNorth Carolina9.7
59Terell ParksWestern Illinois9.6
60Victor OladipoIndiana9.6
61Carl HallWichita State9.5
62Deshaun ThomasOhio State9.4
63Davante GardnerMarquette9.4
64Ken HortonCentral Connecticut State9.4
65Javon McCreaBuffalo9.4
66Joe RaglandWichita State9.4
67D'Aundray BrownCleveland State9.4
68Michael Kidd-GilchristKentucky9.3
69Steven PledgerOklahoma9.3
70Donte PooleMurray State9.3
71Robbie HummelPurdue9.3
72Brandon FortenberrySoutheastern Louisiana9.3
73Doron LambKentucky9.3
74Khris MiddletonTexas A&M9.2
75Evan SmotryczMichigan9.1
76Jackie CarmichaelIllinois State9.1
77Jorge GutierrezCalifornia9.0
78Chace StanbackNevada-Las Vegas9.0
79Velton JonesRobert Morris9.0
80Jaquon ParkerCincinnati9.0
81Kenton WalkerSaint Mary's (CA)8.9
82Kevin PangosGonzaga8.9
83Erick GreenVirginia Tech8.9
84Quincy AcyBaylor8.8
85Nate WoltersSouth Dakota State8.8
86Mark LyonsXavier8.8
87Arnett MoultrieMississippi State8.8
88D.J. CooperOhio8.7
89Sheldon McClellanTexas8.7
90Jamar SamuelsKansas State8.7
91Hollis ThompsonGeorgetown8.7
92Trevor MbakweMinnesota8.7
93Reggie BullockNorth Carolina8.7
94Tony SnellNew Mexico8.7
95Jamal FentonNew Mexico8.7
96Brad WaldowSaint Mary's (CA)8.6
97Scott SaundersBelmont8.6
98Mike Dixon, Missouri8.6
99Maurice KempEast Carolina8.5
100Jamar GulleyMissouri State8.5

rankplayerteamTeam Efficiency ImpactEfficiency Impact/100Min%
1Thomas RobinsonKansas12.315.877.7%
2Damian LillardWeber State12.115.876.8%
3Jared SullingerOhio State11.819.461.0%
4Jarrod JonesBall State11.312.888.6%
5Mike ScottVirginia10.714.972.2%
6Marcus DenmonMissouri10.513.478.6%
7C.J. McCollumLehigh10.413.676.3%
8Kevin JonesWest Virginia10.111.488.8%
9Isaiah CanaanMurray State9.812.976.1%
10Herb PopeSeton Hall9.711.783.2%
11Anthony DavisKentucky9.513.272.3%
12Anthony RaffaCoastal Carolina9.511.384.1%
13Cody ZellerIndiana9.213.767.1%
14Chase TapleySan Diego State9.011.280.7%
15Jae CrowderMarquette9.013.268.2%
16Draymond GreenMichigan State8.810.980.9%
17Jordan TaylorWisconsin8.810.286.4%
18Ryan PearsonGeorge Mason8.812.172.9%
19Brian ConklinSaint Louis8.712.569.6%
20Jared BerggrenWisconsin8.713.166.7%
21Ian HummerPrinceton8.610.284.5%
22Will BartonMemphis8.610.383.2%
23Miguel PaulEast Carolina8.511.374.7%
24Rob JonesSaint Mary's (CA)8.410.282.2%
25Noah HartsockBrigham Young8.310.578.6%
26Doug McDermottCreighton8.211.571.1%
27William BufordOhio State8.29.883.5%
28Russ SmithLouisville8.216.848.4%
29Dominique SuttonNorth Carolina Central8.110.278.9%
30Tony MitchellAlabama7.99.980.0%
31Dion WaitersSyracuse7.814.753.2%
32Nate WoltersSouth Dakota State7.88.888.6%
33Ryan EvansWisconsin7.711.168.8%
34Robert CovingtonTennessee State7.610.175.5%
35Mike MoserNevada-Las Vegas7.610.572.5%
36Kenny BoyntonFlorida7.69.778.0%
37Ken HortonCentral Connecticut State7.69.480.4%
38JaMychal GreenAlabama7.513.954.0%
39D'Aundray BrownCleveland State7.59.480.4%
40Jason ClarkGeorgetown7.510.273.6%
41Arsalan KazemiRice7.511.465.2%
42Drew GordonNew Mexico7.410.074.0%
43Kris JosephSyracuse7.410.372.2%
44Terrell HollowayXavier7.49.776.0%
45Jordan TheodoreSeton Hall7.47.795.9%
46Jereal ScottStephen F. Austin7.410.669.8%
47Jamaal FranklinSan Diego State7.212.457.7%
48Trevor RelefordAlabama7.19.971.9%
49Michael Kidd-GilchristKentucky7.19.376.2%
50Doron LambKentucky7.19.376.2%
51Robbie HummelPurdue7.09.375.5%
52Deshaun ThomasOhio State7.09.474.2%
53Larry AndersonLong Beach State6.97.987.1%
54Tim FrazierPenn State6.97.789.2%
55Terell ParksWestern Illinois6.99.671.8%
56D.J. CooperOhio6.88.778.2%
57Jeremy LambConnecticut6.88.184.1%
58Kevin PangosGonzaga6.88.976.7%
59Langston GallowaySaint Joseph's6.87.985.8%
60Edwin FuquanSeton Hall6.87.689.7%
61Greg ManganoYale6.88.085.2%
62J'Covan BrownTexas6.88.084.6%
63Ricardo RatliffeMissouri6.811.260.5%
64Henry SimsGeorgetown6.811.161.0%
65Ryan BroekhoffValparaiso6.78.579.2%
66Erick GreenVirginia Tech6.78.975.4%
67Tyler ZellerNorth Carolina6.710.464.3%
68Velton JonesRobert Morris6.79.074.0%
69John HensonNorth Carolina6.69.768.8%
70Luke MartinezWyoming6.610.662.9%
71Donte PooleMurray State6.69.371.3%
72Isaiah WilkersonNJIT6.67.884.6%
73Jared CunninghamOregon State6.57.784.5%
74Dominique MorrisonOral Roberts6.57.388.6%
75Julian MavungaMiami (OH)6.45.9108.9%
76Steven PledgerOklahoma6.49.368.7%
77Drew CrawfordNorthwestern6.47.683.8%
78Quincy AcyBaylor6.38.871.9%
79Javon McCreaBuffalo6.39.467.4%
80Bradford BurgessVirginia Commonwealth6.37.980.0%
81Hollis ThompsonGeorgetown6.38.772.5%
82Leonard WashingtonWyoming6.310.162.3%
83Harrison BarnesNorth Carolina6.39.765.0%
84Jorge GutierrezCalifornia6.29.068.8%
85John JenkinsVanderbilt6.28.276.0%
86Ryan KellyDuke6.29.862.9%
87Aaron CraftOhio State6.28.077.7%
88Victor OladipoIndiana6.29.664.6%
89Matthew DellavedovaSaint Mary's (CA)6.26.891.1%
90Reggie HamiltonOakland6.17.285.9%
91Andre RobersonColorado6.18.571.8%
92Joe Harris Virginia6.18.075.9%
93Colt RyanEvansville6.07.580.7%
94Sean KilpatrickCincinnati6.07.184.4%
95Mike MuscalaBucknell6.07.678.8%
96Kerron JohnsonBelmont6.08.471.4%
97Tony SnellNew Mexico5.98.768.3%
98Chace StanbackNevada-Las Vegas5.99.065.8%
99Jeffery TaylorVanderbilt5.97.777.0%
100DeAndre KaneMarshall5.97.281.9%

Quite a few mid-majors at the top here, but the strength-of-schedule adjustment is well-calibrated. These will probably match somewhat closely to Pomeroy's "KPOY."

* - My raw regression involves Defensive Rating, Offensive Rating, Usage%, Assist%, and Steal%.

NCAA Power Ratings

I now have my own usually-updated NCAA power ratings. (Located at the top of the page).

These are simply based on point margin, and I adjust for consistency, recency, and which teams appear to play up or down to their opponents (i.e. fixing the issue of cupcake-killers), in the spirit of the work of DSMok1 aka Daniel.

"Original Rating" is basically SRS, simply adjusting a team's point differential by their opponent's point differential several layers down.

"Quality Rating" weighs each opponent/game by their "original rating" and whether or not a game was at home or away. The final result of a very difficult game is worth the most in this average, and the final result of a very easy game impacts this the least.

"Cupcake Rating" does the exact opposite of "quality rating."

Finally, each team is ranked by using my expected outcome (from Quality/Cupcake/Original ratings) against the top-25 "Original" teams. This gives us a basic idea of how teams might fare come March.


Blog Archive

About Me

I wish my heart were as often large as my hands.