Professional and College Basketball Forums banner
1 - 20 of 24 Posts

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #1 · (Edited)
Below is an interesting (and extrememly telling) stat called Roland Rating. Essentially, it's the measure of how well a player's team does with a particular player on the court. The link below shows the Roland Ratings for every player drafted in the first round of last spring's draft. Outside of guys who only get mopup minutes in blowouts (Darko, Bell, Jones, etc.), there is one rookie that dominates this particular rating, especially given the minutes he's played all season. Guess what? It's this guy:



Roland Rating is the far right column:
http://www.82games.com/comm17.htm

Roland Rating -- the difference between the player's on court and off court "Net48" ratings, which is just plus/minus scaled to 48 minutes. In other words does the team play better (a positive rating) with the player on the court, or worse (a negative rating)
 

·
Registered
Joined
·
2,725 Posts
GO TO THE EFFICIENCY THREAD

you'll get all the reasons and explanations why this isnt "extremely telling"
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #3 ·
Originally posted by <b>tpb2</b>!
GO TO THE EFFICIENCY THREAD

you'll get all the reasons and explanations why this isnt "extremely telling"
Hate to break it to you, but Roland Ratings are completely different than efficiency ratings. Effeciency ratings entail a bit of subjectivity (ie, rebounds count as much as assists, which is debatable), while Roland Ratings tell it how it is: how much better or worse a team does with a particular player on the floor.

It's not only "extremely telling", it's also very simple.
 

·
Premium Member
Joined
·
6,992 Posts
Originally posted by <b>VincentVega</b>!


Hate to break it to you, but Roland Ratings are completely different than efficiency ratings. Effeciency ratings entail a bit of subjectivity (ie, rebounds count as much as assists, which is debatable), while Roland Ratings tell it how it is: how much better or worse a team does with a particular player on the floor.

It's not only "extremely telling", it's also very simple.
They discuss more than just efficency ratings. Why not check the thread out instead of putting it down first.
 

·
Nobody
Joined
·
6,373 Posts
The efficiency thread contains some great discussion on the value of +/- based ratings...

I think the bottom line is, +/- in the NBA doesn't correspond accurately enough with what you see when you watch the games, and it's hard to put a lot of value on stats that are so contradictory to subjective experience.
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #7 ·
Originally posted by <b>R-Star</b>!
When Pietrus is leading you know these rankings are garbage.
Pietrus gets most of his minutes in garbage time.
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #8 ·
Originally posted by <b>rynobot</b>!

Why not check the thread out instead of putting it down first.
Because I do other things.

But if some people can't discern or appreciate the patent value of Roland Ratings, then I guess it's useless to argue with them.

But I have to ask this question: why do college and pro teams employ statisticians who formulate data such as +/-, Roland Ratings and efficiency ratings? Seems to me that if these statistical ratings aren't accurate, then these teams are wasting a lot of time and money on a couple of staff members and worthless data they don't need.

Then again, maybe certain people only listen to what they want to hear.
 

·
Registered
Joined
·
14,070 Posts
heres my question...if the Bulls are doin so good while he is in, and he plays around 40 mins a game lately, why arent the Bulls winning any games
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #10 ·
This is hilarious.

tpb2 wrote in the efficiency rating thread:
There needs to be some way to compile many statistics in order to rank players. The best way would be to examine how much a particular stat contributes to the win.
This is the description of the Roland Rating:
Roland Rating -- the difference between the player's on court and off court "Net48" ratings, which is just plus/minus scaled to 48 minutes. In other words does the team play better (a positive rating) with the player on the court, or worse (a negative rating)
Tell me, tpb or rynobot, what part of Roland Rating criteria doesn't meet tpb2's request in the aforementioned effeciency rating thread?
 

·
Nobody
Joined
·
6,373 Posts
I like Kirk Hinrich a lot; I think he's our future Olympic team PG in 2008. However, I don't believe that +/- ratings can prove that he's the most effective rookie.

Here is a post from the Efficiency thread:

Major Problems with Plus-Minus Ratings

1. Defensive stoppers get subbed-in on the top offensive players. This hurts their +/-. (You see this quite a bit in the NHL)

2. A good "second unit" player will often play against opposing second unit players, and can end up with a better +/- than the starter, even if the starter is actually a better all-round player. However, if you started the second unit player, and played the starter in the second unit, the original starter (the better player) would have a much better +/-.

3. You can't fairly compare +/- numbers for players on teams that don't have similar TEAM +/- numbers. A superb player on a terrible team is likely still going to be overall MINUS. A weak player on a championship team is still going to be PLUS.

4. Players who routinely get late-game garbage time minutes may benefit from easy buckets against disinterested defence, closing a score form a 30pt margin to a 15 pt margin.

There are other examples. Confounding variables, indeed.
 

·
Oladipo for the people
Joined
·
48,150 Posts
Originally posted by <b>VincentVega</b>!


Pietrus gets most of his minutes in garbage time.
If he gets his minutes in garbage time and leads the rookies in Roland then like its been said, how is this accurate?
 

·
Nobody
Joined
·
6,373 Posts
I hate to be harsh on the Bulls here, but don't they have a whole lot of garbage time? I don't know what exactly it means to have a great +/- number on the worst team in the league.
 

·
Registered
Joined
·
2,725 Posts
They hire people to use better, statistically researched rankings. They dont need to employ people for regular efficiency and roland. I could do those myself. THOSE two stats have major flaws, some are obvious as stated, some have to do with finding more confounding patterns, which is left to statistics PhDs.
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #15 ·
Originally posted by <b>R-Star</b>!


If he gets his minutes in garbage time and leads the rookies in Roland then like its been said, how is this accurate?
Take out the garbage time players (the extremes), and the ratings become more accurate, just like in any formula.
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #16 ·
Originally posted by <b>tpb2</b>!
They hire people to use better, statistically researched rankings. They dont need to employ people for regular efficiency and roland. I could do those myself. THOSE two stats have major flaws, some are obvious as stated, some have to do with finding more confounding patterns, which is left to statistics PhDs.
Agreed. Roland Ratings are relatively crude, but I do think it shows something of merit.
 

·
Registered
Joined
·
2,725 Posts
Originally posted by <b>VincentVega</b>!


Take out the garbage time players (the extremes), and the ratings become more accurate, just like in any formula.
they became more accurate but still not close to accurate

It is difficult to eliminate the countless confoundings of the TEAM INTERACTIONS to get accurate INDIVIDUAL CONTRIBUTIONS.

I think it's easier to start with the individual and get complex.
 

·
Registered
Joined
·
8,045 Posts
Discussion Starter · #18 ·
1. Defensive stoppers get subbed-in on the top offensive players. This hurts their +/-. (You see this quite a bit in the NHL)
Aren't the top offensive players usually in the starting unit? If so, wouldn't the defensive stopper -- a player with more defensive proficiency than his starting teammate at the same position -- actually increase his +/- relative to his first-team teammate if he does his job and limits the offensive production of the opposing top offensive player?

2. A good "second unit" player will often play against opposing second unit players, and can end up with a better +/- than the starter, even if the starter is actually a better all-round player.
Agreed, but...

However, if you started the second unit player, and played the starter in the second unit, the original starter (the better player) would have a much better +/-.
...this assumes that the starter thoroughly carries the second team and overcomes any deficiencies of the second team...and/or the "good second unit player" is so much of a downgrade from the normal starter that he actually creates a conspicuous rift in +/-, even though he's playing with the first (best) team.

3. You can't fairly compare +/- numbers for players on teams that don't have similar TEAM +/- numbers. A superb player on a terrible team is likely still going to be overall MINUS. A weak player on a championship team is still going to be PLUS.
Definitely.

4. Players who routinely get late-game garbage time minutes may benefit from easy buckets against disinterested defence, closing a score form a 30pt margin to a 15 pt margin.
ie, Pietrus.
 

·
Registered
Joined
·
10,839 Posts
So according to Roland:

Amare Stoudemire had a crappy season last year? and Dunleavy was better then the ROY?

No matter what you do there is yet to be an equation that will truely value a player accurately just by one number.
 

·
Registered
Joined
·
9,527 Posts
Rank Player Team On Court +/- Off Court +/- Roland Rating
1 Garnett MIN +9.1 -13.5 +22.5
2 Kidd NJN +7.1 -15.1 +22.2
3 Jefferson NJN +5.3 -11.3 +16.6
4 Cassell MIN +8.9 -6.6 +15.5
5 Cardinal GSW +7.3 -7.0 +14.2
6 Davis NOH +3.2 -11.0 +14.2
7 Collins NJN +7.9 -6.2 +14.1
8 Miller SAC +10.0 -2.7 +12.6
9 O'Neal IND +8.1 -4.4 +12.5
10 Kirilenko UTA +3.2 -9.0 +12.2
11 Sprewell MIN +7.3 -3.9 +11.2
12 Martin NJN +6.0 -5.1 +11.1
13 Foster IND +9.9 +0.2 +9.7
14 O'Neal LAL +8.7 -0.2 +8.9
15 Artest IND +7.1 -1.8 +8.9
16 Howard ORL -2.9 -11.8 +8.8
17 Nowitzki DAL +5.5 -2.8 +8.4
18 Alston MIA +0.6 -7.5 +8.2
19 Butler MIA +2.1 -6.0 +8.1
20 Hassell MIN +8.7 +0.6 +8.1
21 Boozer CLE +0.4 -7.5 +7.9
22 Miller IND +7.9 +0.4 +7.5
23 Kittles NJN +4.0 -3.5 +7.5
24 Claxton GSW +2.7 -4.7 +7.5
25 McGrady ORL -3.8 -11.2 +7.4
26 Divac SAC +10.3 +2.9 +7.4
27 Brown NOH +1.9 -5.3 +7.3
28 Turkoglu SAS +10.5 +3.2 +7.3
29 Jackson HOU +4.7 -2.3 +7.0
30 Payton LAL +6.2 -0.8 +6.9

here's the current rankings - can someone tell me what they mean, other than the obvious, which is the players teams do better with them on the court then when they're (or at least, they're units) are not on the court.

there's obviously value to the roland ratings. but there are severe limitations in using it to determine how good players are or what they are contributing. a players roland rating will almost automatically go down if his backup gets better.
 
1 - 20 of 24 Posts
Top