Tennessee Volunteer Plus/Minus and Roland Ratings
by Joel Hollingsworth on Mar 3, 2010 7:34 AM EST
in Tennessee Volunteer Basketball
Email
Print
In his latest statements to the media, Wayne Chism noted that despite his 0-7 shooting from behind the arc against Kentucky, his presence on the perimeter opened up the lane for point guards Bobby Maze and Melvin Goins. He also mentioned that his leadership on defense helped the team beat the Wildcats. He wasn't bragging, he was simply stating that there are contributions to a team's success that aren't traditionally found in the boxscore.
That got me thinking back to something I'd first read in Dane Bradshaw's book Vertical Leap. Bradshaw mentioned how Bruce Pearl and his staff used something called plus/minus to measure the unmeasureables, to somehow gauge the impact of a player's overall contribution to the team while he's on the floor. The concept is to analyze indirect contribution by awarding a player a point for each point the entire team scores while that player is on the floor and to subtract a point for each point the opposition scores while that player is on the floor. The net result is that player's plus/minus for the game. Positive is good; negative is bad.
I have no idea to what extent Pearl relies on plus/minus to judge his players. There are certainly some deficiencies with the stat, namely that the best players are generally on the floor with the best players, so they are all helping each other. On the other hand, opponents are pretty good about finding weak links, and you'd think that that would show up in a team's overall performance. Regardless, if Dane Bradshaw says that it's important to Pearl, well, I believe Bradshaw and defer to Pearl.
I also don't know just how deep Pearl takes the concept. Some sites track not only a player's on-floor plus/minus, but also their on-bench plus/minus. Take a look at this plus/minus table from the Tennessee-Kentucky game, culled from the always fantastic StatSheet.com:
| Plus/Minus | Off Court | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| MIN | RR | +/- | + | - | +/- | + | - | ||
 |
W. Chism | 32 | 7 | 8 | 60 | -52 | 1 | 14 | -13 |
|
B. Williams | 22 | 23 | 16 | 39 | -23 | -7 | 35 | -42 |
|
B. Maze | 26 | -7 | 1 | 47 | -46 | 8 | 27 | -19 |
|
J. Prince | 26 | 17 | 13 | 49 | -36 | -4 | 25 | -29 |
|
S. Hopson | 34 | 23 | 16 | 65 | -49 | -7 | 9 | -16 |
|
S. McBee | 8 | -19 | -5 | 14 | -19 | 14 | 60 | -46 |
|
K. Hall | 12 | -27 | -9 | 22 | -31 | 18 | 52 | -34 |
|
M. Goins | 14 | 7 | 8 | 27 | -19 | 1 | 47 | -46 |
|
S. Pearl | 14 | -7 | 1 | 25 | -24 | 8 | 49 | -41 |
|
C. Tatum | 12 | -17 | -4 | 22 | -26 | 13 | 52 | -39 |
In the Plus/Minus section, the plus column shows the number of points the team scored while that player was on the floor. The minus column shows the number of points the opponent scored while that player was on the floor. The +/- is the net of the two.
The Off Court section is exactly what you'd think: the plus and minus columns show the team's and opponent's points while that player is on the bench, and the +/- column nets the two.
So what is the RR column under the Plus/Minus section? That's "Roland Rating," and it nets the on court and off court numbers, thereby giving a kind of overall plus/minus grade.
So, against Kentucky, Scotty Hopson and Brian Williams made the most positive overall contributions with an RR of 23, and J.P. Prince wasn't far behind with 17. Weezy, despite his awful three-point shooting, still managed a postive contribution. Kenny Hall, Skylar McBee, and Cameron Tatum, not so much.
But that'st just one game. How do the numbers look for the entire SEC season? Glad you asked:
| KY | FL | SC | GA | KY | Van. | SC | LSU | FL | GA | AL | Miss. | Au. | AVG | |
| Chism | 7 | -1 | 14 | 11 | 11 | -7 | 28 | 5 | 19 | -7 | 3 | 14 | 0 | 7.5 |
| Pearl | -7 | -7 | -18 | -5 | 5 | 31 | 8 | 15 | -1 | 23 | -11 | 16 | 16 | 5 |
| Williams | 23 | -17 | -2 | 13 | 7 | 1 | 4.2 | |||||||
| Goins | 7 | 7 | -8 | -3 | 9 | -5 | -10 | 11 | 1 | 23 | 3.2 | |||
| Prince | 17 | 3 | 20 | 21 | -23 | -15 | -4 | 9 | -1 | 5 | 5 | -8 | 8 | 2.8 |
| Hopson | 23 | 13 | 8 | 5 | -17 | 13 | 12 | -1 | -13 | -5 | -13 | 6 | -22 | 0.7 |
| McBee | -19 | 1 | -8 | -25 | 29 | -19 | -2 | 5 | 13 | -13 | 7 | 0 | 4 | -2.1 |
| Maze | -7 | -17 | 8 | 9 | -9 | 1 | 4 | -7 | -1 | -1 | -19 | -12 | 4 | -3.6 |
| Hall | -27 | 35 | 4 | -17 | 11 | -7 | -18 | -5 | -13 | -23 | 1 | 0 | -6 | -5 |
| Tatum | -17 | 5 | -18 | -12 | -9 | 1 | 13 | -5.3 | ||||||
| Woolridge | 3 | -23 | 5 | -18 | -23 | -5 | 7 | 7 | -30 | -10 | -8.7 |
Chism appears to have been the most consistent over SEC play, averaging the highest RR and getting only three negative contribution games. The second-best average -- surprise! -- goes to Pearl, but he's fallen off in the last four games. Fortunately, that's just when Pajamas and Hopson, who have both been on four game tears, have picked it up.
On the down side, Woolridge does not have a good RR rating, and you can see why Pearl hasn't been playing him much recently. Hopefully, he can get back on track and make the most of any late season opportunities that come his way.
Again, I have no idea the extent to which Pearl uses Roland Rating or even more standard plus/minus stats to judge his players' performances, but it is interesting to see how the team performs when any particular player is on or off the floor.
What do y'all think of RR and plus/minus generally? What are the strengths and weaknesses of the concept?
31 comments
|
0 recs |
Do you like this story?
Comments
Just from the Kentucky game stats...
Seems as good of an indicator of team perfromance as anything else.
I told someone right after that game that the reason we won was becuase Brian Williams played like a man instead of the 8th grader he usually looks like out there.
He’s been looking better and better in recent games and making me realize how important to the team he really is.
"I condone fun things" ~~ Cortland Finnegan
yes, it's supposed to measure how a player contributed in the overrall team performance
and regardless of the the number + or – it’s all relative in relation to other players on the team.
also, what do you mean Williams normally looks like an 8th grader? How so?
I just mean he had often played immaturely in the past
imo at least. And it seems to me that he has been unwilling to use his size to create things instead of just fouls.
He’s a different player than he has been recently. He’s getting stronger rebounds, blocking more shots, and making the putbacks he should be making for a post player.
"I condone fun things" ~~ Cortland Finnegan
he wasn't utilizing his size to his fullest potential
yeah, i guess you could say that’s immaturity – basketball-wise. but even before he got suspended, i thought that he was a better defensive presence in the paint than Kenny Hall, and probably a better rebounder too. Also, considering the games he has been back for are part of the more difficult end of our schedule, and he was playing against good big men – Oglivy, Patterson, and Cousins (and it was those games that he had his best +/-), i think his average +/- ranking 3rd on the team is even more impressive. Brian is the target of some pretty heated criticism by vol fans, I think mostly because of his poor offensive production, but considering our options, I think he is more valuable than a lot of people realize.
Definitely agree on Kenny Hall
He’s got to get some consistency and composure in his game this offseason
"I condone fun things" ~~ Cortland Finnegan
i wonder how they would look if you weighted them based on minutes played.
or did them on a per-minute basis.
if you look at Kentucky's +/- for that game
DeMarcus Cousins was their worst. But I don’t think that means UK would have been better with him on the bench.
Great post, Joel
My thoughts on the +/- stat is that it’s highly flawed when served in small portions. But over a larger sample size, I think it can be a useful and insightful statistic.
...just apologize for not thanking me.
it can be useful when looking at an individual game
if you watched the game and understand the situations each players’ minutes came under
I think +/- is much more valuable for the minus
for instance, the most telling stat on that board is probably Kenny Hall’s number in comparison to the other post players on the team. It gives you a better idea of how much of a dropoff you have when the best players are on the bench.
you mean the points given up while that particular player is in?
or an overall +/- that is in the negative?
the overall +/-
the top three guys on the board are all post players (though the numbers are less reliable for guys like Brian Williams and Cam Tatum who’ve played fewer games) – the difference between their average and Hall’s is significant, and it gives extra weight to the reasons behind the fact that Steven Pearl (not Hall) was the first one off the bench against Kentucky
by Will Shelton on Mar 3, 2010 11:23 AM EST up reply actions
There are some really surprising things though
For instance, Cam Tatum had one of the worst +/- on the team despite generally filling out the stat sheet. 9 points and 5 boards in only 12 minutes.
In a game of runs, it basically just tells you who was on the court and on the bench when the big runs happened. If someone had a bad stretch during a run, that will really hurt them. For instance, I think Tatum played very well in limited time in the first half. I bet he really filled out the stat sheet. But in the second half, although he was only on for probably 5-7 minutes, he kept jacking up threes at the end of the shot clock, and UK went on their big run. Because of that run, his 5 bad minutes hurt him a lot more than his 7 good minutes helped him. A lot of times that really tells you how people have played. Sometimes it’s just luck of the draw (I would imagine the second Vandy game is also like this). Like kidbourbon said, the large sample size is good. It’s definitely right on Chism being our best player, and it’s right on Pearl being a good contributor. It’s also clearly right on Woolridge being a HUGE liability in SEC play.
by Incipient_Senescence on Mar 3, 2010 11:07 AM EST reply actions
In addition to Tatum
Another example is Maze’s low rating in the last game. He was basically the only player on the floor who kept us in the game when they went on their big run. But that means he was on the floor when they went on their big run. So I think that his poor +/- is a reflection on how the team was playing at that time, despite the fact that he really played well individually.
Obviously, you want guys who will make the team better, but sometimes there’s just one guy playing well, and there’s really nothing he can do. I think that’s what happened with Maze for a little while in the second half. That said, Williams, Hopson, and Prince were definitely the three biggest reasons for that win. Williams’ stats look unimpressive, but I think he played much better. This shows why I (and many others here) thought that.
by Incipient_Senescence on Mar 3, 2010 11:10 AM EST up reply actions
Only the guys who were in the game
The guys who were on the bench have their +/- aided during the run. I’m not convinced that, for instance, having Melvin Goins in ahead of Bobby Maze would’ve changed anything for that stretch of the game.
by Incipient_Senescence on Mar 3, 2010 11:33 AM EST up reply actions
their roland rating is aided
it’s true that there is more to the game than statistics can encapsulate. but as far as statistical analysis goes, i think this is a pretty decent measure
that's why in single game stats, you need to know the situations in which each player played
but over the long run, as kidbourbon points out, it can be significant.
I keep thinking of more things
Also, there are some situations where a high +/- just comes from a player being on the bench when the opponent went on their one big run. I would expect this is the reason for high numbers for Hall against Florida and McBee against Kentucky. Another reason that single game samples aren’t good unless you really know what the stats are telling you.
by Incipient_Senescence on Mar 3, 2010 11:14 AM EST reply actions
so why did a team go on that big run?
could it be bad defense? should that not be accounted for in +/-?
But that's the opposite of the point
The players in the game during a big run for the opponents are penalized, for obvious reasons. But the players on the bench have their +/- aided by being on the bench during the opponent’s big run. Sometimes this makes sense. If we go to pot when Chism gets in foul trouble, then that says something about Chism as a player. But I think it’s less accurate when talking about players who don’t play a whole lot. Do you really think Hall being on the bench was the reason Florida destroyed us in the second half? Was McBee being on the bench the reason that Kentucky killed us in the last ten minutes @Rupp?
by Incipient_Senescence on Mar 3, 2010 11:36 AM EST up reply actions
I'm sorry
I’ve been talking about the RR the whole time. Sorry if that was confusing.
by Incipient_Senescence on Mar 3, 2010 1:06 PM EST up reply actions
I'm floored by McBee's +29 in the first Kentucky game.
Kentucky was feasting on him whenever he was on defense, but I suppose he was also in for a few runs.
Simulated Gameday Experience - just like the real thing, only we have smoke machines.
Chism's comments
Wayne Chism noted that despite his 0-7 shooting from behind the arc against Kentucky, his presence on the perimeter opened up the lane for point guards Bobby Maze and Melvin Goins
Because that’s what basketball is all about: your best (and only effective) post player continually hoisting up three-pointers to open up the paint for your two unders-sized point guards.
Whatever works, I guess…
Lou Brock loves Lamp.
when UK is playing man, and it draws Cousins or Patterson out of the lane
then it’s not such a bad plan. Pearl has been very successful in the past with that strategy.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recent FanPosts
SHOP THE ROCKY TOP TALK STORE
Animated Drive Charts
RTT Classics
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
