PDA

View Full Version : Team and Players rated -- the Utah Index (a slight return)



kester99
12-31-2009, 11:19 PM
In the last couple of years, I've posted some of the Pacers player ratings as determined by the Utah Index, a simple player rating system used by the late Larry Miller, of the Utah Jazz.

This is a comparison of our players' individual performance, and the Pacers' performance this year vs last, and vs the rest of the league, and our opponents.....kind of a performance / stats overview for the non-statistically inclined.

The Utah Index adds points, rebounds, assists, steals and blocks; subtracts FGAs, TOs and PFs; then divides the whole mess by minutes played. So: (Points+As+Rbs+Stls+Blks) - (FGAs+TOs+PFs)
......................Minutes Played.............................

example -- Granger thru first 30 games this year:
(24.4pts + 6.1 Rbs +2.6 ast + 1.6 stls + 1.0 blks) = 35.7
minus (18.6 FGAs + 2.9 TOs + 3.3 PFs)................= 24.8
.................................................. .......................10.9,
............................... divided by 36.6 mins/gm= .298 rating

Player.........09/10 .....08/09 season
Granger...........298........312
Murphy............411........429
Ford................216........262
Dunleavy.........155........200
Foster.............207.........287
Hibbert............220........194
Rush...............124........112
McRoberts.......148.........212
D. Jones..........098
S. Jones..........145
Head..............128
Watson...........240
Hansbrough.....247
Price...............219
Diener...........................274

Pacers...........204.........243
Opponents.....261.........273
NBA..............235 .........241

So what do we see? We are, as a team, below the index average, although we were just above the average last year.

Our opponents are significantly above the average each year, although not as much this year vs 08-09.

Using the .235 NBA avg this year, we have four above average players.

Almost all the returning players indexes are down...Roy and Rush being the exceptions.

---------------------------------------------
Interesting to see the relative numbers. Obviously, a purely defensive specialist could get short-changed by this index, if his defensive efforts don't result in steals or blocks. Conversely, (I'll say it before someone else does, even if I don't buy the spin personally) a stats-hungry fantasy stud that doesn't play serious defense could come off very well.

It should be noted also, this index looks at performance vs floor minutes, so someone that played two minutes really well would have a better number that an average 30-minute player.

count55
12-31-2009, 11:31 PM
And if I don't meet you no more in this world
Then ill, I'll meet you in the next one and don't be late, don't be late.

Putnam
01-01-2010, 08:22 AM
We have four above average players

(all of whom are rated worse than last year)



No sun coming through my windows,
feel like I'm sitting at the bottom of a grave.
No sun coming through my windows,
feel like I'm sitting at the bottom of a grave.

I wish you'd hurry up 'n' rescue me
so I can be on my mis'rable way

Anthem
01-01-2010, 09:57 AM
Man, we need to trade Murph to Utah. They'd love him there.

BlueNGold
01-01-2010, 10:34 AM
Man, we need to trade Murph to Utah. They'd love him there.

No doubt. We need to print this stat out and show how much better Murphy is than Granger. No question about it...they'd give us Deron for Troy.

Better yet, I calculated Kobe's Utah Index for this year. It's only .362. Why Larry is not on the phone right now with the Lakers...I just don't understand.

Putnam
01-01-2010, 11:28 AM
Here the players' Utah Index during the . . . let's just say during a period of a couple of weeks during early to mid November:


Hibbert.....451
Granger....399
Ford.........350
Watson.....303
DJones.....266
Rush........213
Price........211
SJones.....155
Head........119



EDIT: Where's McKeyFan? This post has been here nearly five hours and he hasn't yet jumped on it.

.

Brad8888
01-01-2010, 11:39 AM
Here the players' Utah Index during the . . . let's just say during a period of a couple of weeks during early to mid November:


Hibbert.....451
Granger....399
Ford.........350
Watson.....303
DJones.....266
Rush........213
Price........211
SJones.....155
Head........119



.

So, is it safe to assume that Jim O'Brien is not from Utah?

At least it is as valid as +/- as a tool, isn't it?

:)

Anthem
01-01-2010, 11:46 AM
So, is it safe to assume that Jim O'Brien is not from Utah?
He might be. It would explain why he thinks Troy is our best player.

Brad8888
01-01-2010, 12:16 PM
He might be. It would explain why he thinks Troy is our best player.

Great catch...:o

Naptown_Seth
01-01-2010, 12:51 PM
So, is it safe to assume that Jim O'Brien is not from Utah?

At least it is as valid as +/- as a tool, isn't it?

:)
Not even remotely close to being as valid.

1) Watch games
2) Ford and Troy at the bottom of the +/- list, and that's per minute not just total
3) Troy best player on team by Utah, Ford better than Head and Rush and not that far off of Watson even, DJones as an epically terrible player
3b) See things like Kobe's Utah index, now look at Kobe's +/- for the Lakers

Not only that, but 82games.com has the ON-OFF court +/- which is even better, and Troy and Ford are the ONLY Pacers who have a POSITIVE off court number to go with their large on court negative number. The only guys who the numbers say the team actually wins when they don't play.



Here's why the Utah index is terrible, the stats aren't normalized.

If you are a great defender and the only way we measure that is by blocks and steals here, then what does a GREAT, MEGA STAR defender put up for those stats? Lets say 3 blocks and 2.5 steals per game. This is a guy crushing FG% and chasing people away from the rim constantly, tipping passes and picking pockets. It's 50% of the game and he's a hero on that end, a combo of Dwight Howard (leading with 2.5) and Rondo (leading with 2.5).

Your total "score" for the defensive end play is 5.

Now look at any run of the mill PG, a guy the team could or could not live without. He drops 4-5 assists per game, okay but not great. He doesn't score ANY points or make any defensive plays at all. None.

This system says he's just as good as the superstar defender.

In +/- the 4.5 assist player could go -10 while letting his man waltz to the rim and killing the shot clock at the other end before a bail out pass to a teammate results in a missed FG. The Defensive ace doesn't do anything on offense, just stands to the side while his team eek out 6-7 points, but at the other end he literally doesn't allow the other team a single point. He's +7.

So the +/- shows this, the Utah index says these guys are identical.



ALL FORMULAS have these issues, they are based on already questionable meaning stats like rebounds (they aren't all the same) and assists (it's a subjective stat where an individual decides when to give it, and even then they aren't all of the same quality). Then they typically make little to no effort to normalize what is considered "good". Is 2 blocks really the same as 2 uncontested rebounds? Or even 2 assists? 2 blocks on the first 5 trips and the other team might not even bother coming at you after that, not to mention the changed shots to avoid the block.

The fouls and TOs are meant to counter the defensive play and assists, but that great defender I mentioned could easily have 2-3 fouls in a game. So his Utah score isn't even 5.5 probably, it's more like 2.

The shooting stat is really just the Points Per Shot, except instead of division it's done as a total, so the more you shoot the better your score (if you have a positive PPS, which any marginally good player does).

Yet if you don't make 3s or draw FTAs, but do shoot well, ie 50% from the floor, you suck by the Utah index. 13-26 for 26 points is worthless. It's a non-contribution.

To keep up with a marginal rebounder of say 7 rebounds per game, that 13-26 shooter would need to go 4-5 from the FT line AND 3 of 7 from deep for a 33 point night.

33 point night on 50% from the field and you are IDENTICAL to the 7 rebound guy who literally did nothing else.

Then to make it worse you get 2 steals, a block and 3 assists, but turn it over 3 times and have 3 fouls, and you are STILL identical to the guy that stands on the sideline on defensive possessions.



The Utah system is notorious for overemphasizing rebounds and under-appreciating defensive play.

Oh, and like the +/- it's based 100% on WHO YOU PLAY WITH AND AGAINST, just like every single stat kept.



*I'm not trying to be a huge +/- fan, but it's clearly being dismissed as a stat by fans who have never taken the time to really analyze the already used and massively flawed "traditional" stats.

Midcoasted
01-01-2010, 03:02 PM
*I'm not trying to be a huge +/- fan, but it's clearly being dismissed as a stat by fans who have never taken the time to really analyze the already used and massively flawed "traditional" stats.

I think if you take out fans and insert O'Brien you would have it right. Ford and Murphys plus/minus is tell all IMO. We are better without them. I hear him talk about the plus/minus all the time but obviously he isn't living by it, or by whats clearly visible on the court. The season is lost. Bench the vets and play the young guys now. We just might be better in the future for it. Instead of being a mediocore to bad team.

And I agree that the Utah Index is flawed for player to player comparisons. Obviously Murphy isn't better than Granger or Kobe. But what you have to look at here is the trends from year to year and compare them.

I agree with the Index that Murphy, Ford, Granger, and Dunleavy have all regressed this year. I also agree Rush and Hibbert have progressed. It is spot on. I think the Utah Index is 100 percent right in this aspect.

speakout4
01-01-2010, 03:47 PM
I also agree Rush and Hibbert have progressed.
Rush has progressed? What am I missing? Because he has been so inconsistent on offense all of a sudden people are talking about his brilliant defense. I am not convinced even that is in the very good category since I haven't seen numbers comparing how he defends said player vs. how others defend the same player.

kester99
01-01-2010, 04:29 PM
Rush has progressed? What am I missing?

I think we're looking at maybe three distinct periods so far with him. One was early last year (most of the year); two was when he came out of his shell the last few games last year (which everybody remembers, but it was really a short period); and three is this year.

To me, he's better so far this year than the majority of last year...still the learning curve seems to be long. His flashes of effectiveness don't seem quite so far apart, though.

Putnam
01-01-2010, 04:34 PM
I think we're looking at maybe three distinct periods so far with him. One was early last year (most of the year); two was when he came out of his shell the last few games last year (which everybody remembers, but it was really a short period); and three is this year.



Four distinct periods. Four!






.

kester99
01-01-2010, 04:40 PM
I said 'maybe' three. That's the same as four.

speakout4
01-01-2010, 05:12 PM
I said 'maybe' three. That's the same as four.
I would have no problem with Rush if his trajectory was up regardless of how poorly he started.

Brad8888
01-01-2010, 07:03 PM
Not even remotely close to being as valid.

1) Watch games
2) Ford and Troy at the bottom of the +/- list, and that's per minute not just total
3) Troy best player on team by Utah, Ford better than Head and Rush and not that far off of Watson even, DJones as an epically terrible player
3b) See things like Kobe's Utah index, now look at Kobe's +/- for the Lakers

Not only that, but 82games.com has the ON-OFF court +/- which is even better, and Troy and Ford are the ONLY Pacers who have a POSITIVE off court number to go with their large on court negative number. The only guys who the numbers say the team actually wins when they don't play.



Here's why the Utah index is terrible, the stats aren't normalized.

If you are a great defender and the only way we measure that is by blocks and steals here, then what does a GREAT, MEGA STAR defender put up for those stats? Lets say 3 blocks and 2.5 steals per game. This is a guy crushing FG% and chasing people away from the rim constantly, tipping passes and picking pockets. It's 50% of the game and he's a hero on that end, a combo of Dwight Howard (leading with 2.5) and Rondo (leading with 2.5).

Your total "score" for the defensive end play is 5.

Now look at any run of the mill PG, a guy the team could or could not live without. He drops 4-5 assists per game, okay but not great. He doesn't score ANY points or make any defensive plays at all. None.

This system says he's just as good as the superstar defender.

In +/- the 4.5 assist player could go -10 while letting his man waltz to the rim and killing the shot clock at the other end before a bail out pass to a teammate results in a missed FG. The Defensive ace doesn't do anything on offense, just stands to the side while his team eek out 6-7 points, but at the other end he literally doesn't allow the other team a single point. He's +7.

So the +/- shows this, the Utah index says these guys are identical.



ALL FORMULAS have these issues, they are based on already questionable meaning stats like rebounds (they aren't all the same) and assists (it's a subjective stat where an individual decides when to give it, and even then they aren't all of the same quality). Then they typically make little to no effort to normalize what is considered "good". Is 2 blocks really the same as 2 uncontested rebounds? Or even 2 assists? 2 blocks on the first 5 trips and the other team might not even bother coming at you after that, not to mention the changed shots to avoid the block.

The fouls and TOs are meant to counter the defensive play and assists, but that great defender I mentioned could easily have 2-3 fouls in a game. So his Utah score isn't even 5.5 probably, it's more like 2.

The shooting stat is really just the Points Per Shot, except instead of division it's done as a total, so the more you shoot the better your score (if you have a positive PPS, which any marginally good player does).

Yet if you don't make 3s or draw FTAs, but do shoot well, ie 50% from the floor, you suck by the Utah index. 13-26 for 26 points is worthless. It's a non-contribution.

To keep up with a marginal rebounder of say 7 rebounds per game, that 13-26 shooter would need to go 4-5 from the FT line AND 3 of 7 from deep for a 33 point night.

33 point night on 50% from the field and you are IDENTICAL to the 7 rebound guy who literally did nothing else.

Then to make it worse you get 2 steals, a block and 3 assists, but turn it over 3 times and have 3 fouls, and you are STILL identical to the guy that stands on the sideline on defensive possessions.



The Utah system is notorious for overemphasizing rebounds and under-appreciating defensive play.

Oh, and like the +/- it's based 100% on WHO YOU PLAY WITH AND AGAINST, just like every single stat kept.



*I'm not trying to be a huge +/- fan, but it's clearly being dismissed as a stat by fans who have never taken the time to really analyze the already used and massively flawed "traditional" stats.

I 100% agree with your post. I should have placed a Sarcasm Alert warning at the beginning and end of my post.

I actually have no interest in either +/- OR this "Utah System" as a true indicator of anything. They are happenstance curiosities that are either totally dependent on everything that a player cannot control in the case of +/-, being the performance (or lack of) of the other players who happen to be on the court at the same time, or, in the case of the "Utah System" only a portion of the overall stat line that does not take many important factors into account.

Woe be unto any GM who would rely much on this system in making personnel decisions based on its' own merits. It does surprise me that it is from Utah, but I would doubt that Sloan would put much, if any, faith into it if he were to be involved with personnel decisions. He is way too smart for that.

bulldog
01-01-2010, 09:44 PM
I actually have no interest in either +/- OR this "Utah System" as a true indicator of anything. They are happenstance curiosities

I don't think you're actually agreeing Naptown_seth. My reading of his post, and correct me if I'm wrong, Naptown, was that ANY statistical system has flaws, but that's not a reason to "have no interest" in it or just throw it out. You watching the game has a lot of flaws (you might not pay attention while your significant other is making dinner plans with you, you might not know a lot about basketball, you might have a bias for players that remind you of a guy you used to play with in high school) but you don't completely distrust your eyes and opinions. Both the "Utah System" and +/- have a lot of weaknesses (the Utah System, particularly, has a whole lot of them), but when you look at the game through multiple viewpoints, and acknowledge the weaknesses and strengths of each, you get a very useful way of looking at the game. So useful that you can start building a team with that knowledge. I guarantee that +/-, and various adjusted +/-'s, are in the toolkits of a variety of the more stat-headed organizations in the league, like Houston and Dallas.

Brad8888
01-02-2010, 08:44 AM
I don't think you're actually agreeing Naptown_seth. My reading of his post, and correct me if I'm wrong, Naptown, was that ANY statistical system has flaws, but that's not a reason to "have no interest" in it or just throw it out. You watching the game has a lot of flaws (you might not pay attention while your significant other is making dinner plans with you, you might not know a lot about basketball, you might have a bias for players that remind you of a guy you used to play with in high school) but you don't completely distrust your eyes and opinions. Both the "Utah System" and +/- have a lot of weaknesses (the Utah System, particularly, has a whole lot of them), but when you look at the game through multiple viewpoints, and acknowledge the weaknesses and strengths of each, you get a very useful way of looking at the game. So useful that you can start building a team with that knowledge. I guarantee that +/-, and various adjusted +/-'s, are in the toolkits of a variety of the more stat-headed organizations in the league, like Houston and Dallas.

Sorry to have offended you Bulldog. Yes, Rick Carlisle definitely uses +/- and other analytic methods and always has, I am sure Adelman does, too, and I am also sure that nearly any team at any level of prominence does also.

Thank you for putting your two cents in here. Any time that there are statistics available, they can have value. I don't like either +/- or the Utah system because, as I posted previously, they both have, in my view, significant weaknesses, and should not be utilized in making personnel decisions due to their individual weaknesses, and probably should have been more clear and said that, as a fan of the actual game I have no use for either in figuring out the contributions of individual players in a given game because at that level their effectiveness is extremely limited due to the size of the data set that they draw from. This year's Pacer team is one of the better examples of the weaknesses of both +/- and the Utah system, in my opinion, and, while I glance at the +/- after watching individual games, I often laugh out loud at what it shows, which is basically who played with the part of the rotation that hit its threes that night vs. those who didn't. It shows nothing about why, and cannot, and neither does the more limited Utah system that takes defensive effort, and some offensive stats, out of the equation entirely.

My guess is, and I am sure your guess is, as well, that if teams at any level use these two analytic methods, they are only a small portion of the overall set of analytic methodologies that they use, and that they use all of the analytic methods that they choose to use as only a portion of the criteria when making personnel decisions. If stats were the end all be all for personnel decisions, why would any team at any level have entire scouting departments for the purposes of both game preparation and personnel decisions such as drafting, trading, and signing future free agents? Why waste all of that time and money? Obviously it is because ultimately any sport is too complex to completely break down in a reliable fashion through the use of statistics, let alone arguably the most fluid sport on the planet, basketball.

I do agree with Naptown, whether you think I do or not. +/- is the better of the two, easily, and ultimately neither are all that effective due to each of their inherent weaknesses.

Hopefully this clears my position up for you, Bulldog, and further for you, Naptown, if that was required, which I doubt.

Putnam
01-02-2010, 08:58 AM
If I were to say, "AJ Price is an all-star point guard because he leads the NBA in assists this season!" I would be wrong. He doesn't. He has only 9, while Steve Nash has 368. So the statement is wrong in fact.

If I were to say, "Peyton Manning is overrated. Far from being the best in the NFL, he is not even the top scorer on his own team!" I would again be wrong, but for a different reason. Addai, Clark and Wayne have all scored 10 touchdowns each, while Manning has not scored a single one all season. So I would be correct as to fact, but wrong as to interpretation -- it is not the quarterback's job to score.

Now, my point is that statistics are almost never wrong in the first sense, and are never, ever to be blamed in the second sense. In other words:

1. They are almost always true.*
2. The interpretation is up to you and is not implicit in the number.

It is wrong, childish, stupid to deliberately misinterpret a data value and then declare it inaccurate or misleading. The +/- doesn't say Troy Murphy is the worst player on the Pacers. The Utah doesn't say he is the best. What they say about him is true -- and it is up to you to explain how, being what he is, he can produce such outstandingly good or bad stats for himself.

Now, as to Seth's point about this, that and the other statistic being influenced by exogenous variables. Yes! They are.

Consider: if you go over to Kegboy's Rate the PaceMate thread, I'm sure that some of the data presented there will generate some interest. And I'm sure that some of the pixels on those young ladies will draw more attention than others do. There's nothing wrong with one person focusing on navels, while another studies ankles, another eyes, and another smiles. (Yes, yes, I know perfectly well that you're all looking just at T&A, but my point stands.) It is possible to measure which of those dancers has the biggest bust, the longest legs, the firmest tummy, the phattest bubble (did I get that right?) or what have you. But there is no measuring tape that will unerringly show which is the hottest PaceMate, or even which is "practically perfect in every way" like Mary Poppins.

It is OK to prefer blondes. It is not OK to say that, because her hair is brown, she is uglier than she truly is.

Similarly, it is OK to look at the Utah rating and say, "Murphy gets this high a rating because he is a stat stuffer." Or to look at the +/- and say, "Murphy gets into the box score, but he doesn't make his team better." It is also OK to look at them and say, "Yes, this is true, but the game was won or lost at the free throw line," or "The refs cheated us," or "You just can't expect to beat that team on their floor," or whatever. There are lots of other factors (as Brad says just above) that matter as much as those contained in a particular statistic.

But it is plain dumb to look at a statistical value and say it is wrong or useless.





* Allowing for the matter of interpretation. We know that assists and rebounds are subjective, and that rogue scorers have appeared from time to time and those issues do add a bit of muzziness.

BlueNGold
01-02-2010, 11:28 AM
While this stat is correct for what it is, I don't see how it's practically useful.

You might as well multiply Troy's rebounds by the DOW Jones average by the distance to the moon by the number of ants that have crossed my driveway in the last 6 months.

Sure, that stat will tell you a little bit about each of these things, but it doesn't tell you anything else.

McKeyFan
01-02-2010, 11:30 AM
Here the players' Utah Index during the . . . let's just say during a period of a couple of weeks during early to mid November:


Hibbert.....451
Granger....399
Ford.........350
Watson.....303
DJones.....266
Rush........213
Price........211
SJones.....155
Head........119



EDIT: Where's McKeyFan? This post has been here nearly five hours and he hasn't yet jumped on it.

.

I typically avoid threads with words like Index. The plural Indices is a deal breaker, and doesn't even makes sense to have the i before the e.

Anyway, no I am not surprised that the numbers are high for players during that time of the Fall of 09 that we dare not mention lest the Gestapo hear us.

We played very well during that stretch, especially on defense, and that defense doesn't really show up that well on the Utah index.

But I do appreciate your posting those numbers to confirm the obvious.

bulldog
01-02-2010, 04:29 PM
Sorry to have offended you Bulldog.

Hopefully this clears my position up for you, Bulldog, and further for you, Naptown, if that was required, which I doubt.

You definitely didn't offend me, and your points were for the most part good to me from the beginning. Sorry for the misunderstanding.

Anthem
01-02-2010, 09:39 PM
"practically perfect in every way" like Mary Poppins.
:drool: