 |
APBRmetrics
The statistical revolution will not be televised.
|
| View previous topic :: View next topic |
| Author |
Message |
94by50
Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
|
 Posted: Tue Mar 28, 2006 2:41 pm Post subject: |
 |
|
| Analyze This wrote: |
with Shaq:
33-11 won/loss record (.750)
28.84 pt/g
6.82 rb/g
5.55 as/g
1.73 st/g
0.82 bl/g
3.39 to/g
53.2% FG%
78.5% FT%
11.18 FTA/g
18.07 FGA/g
19.0% 3FG%
without Shaq:
9-9 won/loss record (.500)
25.00 pt/g
6.50 rb/g
6.50 as/g
2.28 st/g
0.72 bl/g
3.39 to/g
44.0% FG%
76.1% FT%
10.94 FTA/g
20.72 FGA/g
05.0% 3FG%[/i]
I find this difference remarkable. |
I don't. What this says to me is that Wade shoots worse when Shaq doesn't play (presumably because the defense can focus on Wade, and not have to worry about anyone else). It also says that the Heat play worse without Shaq.
Of course the Heat play worse without Shaq! He's still as good as any center in the league, even if he's not the force that he was in his prime. So if he can't play, the Heat have to sub in a worse player for him. Not only that, but because Mourning would start in that case (ideally), the Heat have to replace Mourning's role off the bench with someone like Doleac.
This is why I don't pay much attention to "with x player/without x player" splits. EDIT: This kind of split may be meaningful sometimes, but I don't see a very large difference here. Wade's TO/G actually stay the same even though he has to take on a larger role in the offense, and his Ast/G and Stl/G increase. |
|
| Back to top |
|
 |
Analyze This
Joined: 17 May 2005
Posts: 364
|
| Posted: Tue Mar 28, 2006 2:59 pm Post subject: |
|
|
| Yeah, it's normal that they do worse because of the remarks that you are making. But that much? I should have suspected that the difference was a lot smaller. I ask myself this, if Shaq had played all those games and Wade was hurt that long, how would Shaq his win lost record would have looked? Wade has the highest PER of the league and the second biggest difference between per and opponent per (30,3 versus 13, 7/ Lebron his rating is 0,1 better: so you could say they are both at the number 1 spot). But does that mean that Wade has the highest value concerning winning games? Well I don't know about that. There are other methods of measuring this of course. But I'm not convinced that we can measure the complete impact of a player. |
|
| Back to top |
|
|
94by50
Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
|
| Posted: Tue Mar 28, 2006 3:20 pm Post subject: |
|
|
| Analyze This wrote: | | Yeah, it's normal that they do worse because of the remarks that are you making. But that much? I should have suspected that the difference was a lot smaller. I ask myself this, if Shaq had played all those games and Wade was hurt that long, how would Shaq his win lost record would have looked? Wade has the highest PER of the league and the second biggest difference between per and opponent per (30,3 versus 13, 7/ Lebron his ratig is 0,1 better: so you could say they are both at the number 1 spot). But does that mean that Wade has the highest value concerning winning games? Well I don't know about that. I'm not convinced that we can measure the complete impact of a player. |
Then what does the split tell you? Do you believe that it is relevant in some way, and if so, how is it relevant? What do you think it means?
In your view, PER does not fully capture a player's value. I agree, although I feel confident that it provides a basis for distinction - that is, for example, a player with a 28 PER is most likely more valuable than a player with a 22 PER.
You have cited Wade's split as evidence of... something. What? Does the split indicate that Wade is not worth considering for MVP? Does it indicate that Shaq is, or that they both are, or that neither one is? |
|
| Back to top |
|
|
Mike G
Joined: 14 Jan 2005
Posts: 3597
Location: Hendersonville, NC
|
| Posted: Tue Mar 28, 2006 4:43 pm Post subject: |
|
|
| 94by50 wrote: | | Charles wrote: | | [Kobe and] Lebron have had MVP qualitry, years, but their team's haven't won enough. |
OK, the Cavaliers are going to be a top-4 seed in the playoffs, and they're probably going to end up near 50 wins on the year, and the only other players who've made any significant contribution are Ilgauskas and Gooden ..
How much more do you want James to do? And why are you penalizing him because Hughes went down for the year? |
He should be penalized because that's the American way. It's just like changing your religion in Afghanistan. If he's not a major winner, he's a loser.
Cle is heading for 48-34, while LA is looking at 43-39. That's a pretty big difference. Conference be damned.
Kevin Garnett is punished every day for 'his' organization. He should also be unrecognized as an mvp-candidate. Loser.
_________________
`
36% of all statistics are wrong |
|
| Back to top |
|
|
Analyze This
Joined: 17 May 2005
Posts: 364
|
| Posted: Wed Mar 29, 2006 12:48 am Post subject: |
|
|
| 94by50 wrote: | | Analyze This wrote: | | Yeah, it's normal that they do worse because of the remarks that are you making. But that much? I should have suspected that the difference was a lot smaller. I ask myself this, if Shaq had played all those games and Wade was hurt that long, how would Shaq his win lost record would have looked? Wade has the highest PER of the league and the second biggest difference between per and opponent per (30,3 versus 13, 7/ Lebron his ratig is 0,1 better: so you could say they are both at the number 1 spot). But does that mean that Wade has the highest value concerning winning games? Well I don't know about that. I'm not convinced that we can measure the complete impact of a player. |
Then what does the split tell you? Do you believe that it is relevant in some way, and if so, how is it relevant? What do you think it means?
In your view, PER does not fully capture a player's value. I agree, although I feel confident that it provides a basis for distinction - that is, for example, a player with a 28 PER is most likely more valuable than a player with a 22 PER.
You have cited Wade's split as evidence of... something. What? Does the split indicate that Wade is not worth considering for MVP? Does it indicate that Shaq is, or that they both are, or that neither one is? |
Well I don't know. I agreed that Wade was an mvp candidate and then I saw the split and I'm not sure anymore. It could be an indication that Wade his value is not so high as some other players with lower PER rankings or +/- rankings. In basketball you have constant interaction. If you want to measure what players do, you must break the flow of the game in small pieces. e fg% measures the efficiency of fg's, ts% also takes ft into the bad, rebound rate, as ratio, to ratio and so on. You can measure most parts of the offense pretty well. In defense it becomes more difficult. That's problem number one. If you want to give a value to what a player did for his team you must put all the smal pieces back together. In doing so you give a value to the different parts. And in doing so you are influencing the outcome. Other values= another outcome. What's the value of an assist? How much credit goes to the scorer and how much credit to the man who gives the assist? Do you must value all assists the same way? I don't think so. A difficult assist, a player who finds a small opening by for example giving a player the ball in the paint could be worth more than an easy assist. But again that's interpretation. And you have the problem of defense. How do you value a screen, deny the ball defense, and so. And if a player misses is it because of good defense or just because the player missed regardless the defense. How do you value those kind of things? I think sometimes it's a case of interpretation. You can say well if the player doesn't score that's well defended, if he scores that's not well defended. We will measure it that way. But that is not always true. If I want to know if player x is a good defender in a certain game or serie, I want to know if his opponent didn't do well because of his defense or because he just didn't play good regardless the defense in a particular game. Or perhaps it was a mix of the two. And what about help defense? And how do you split the credit? How do you determine that? That matters if you want to give the defender credit for his defense. And again how do you value deny the ball defense, help defense, a screen; and so on. I know the work of Oliver and the defensive box score project. I know also his other system, the estimate, the defensive rating. But with the last system you again are giving values.
With + - rating the quality of the back up plays a role, the opponents, the moment in the game. So how can you measure the complete value that a player has on a team with all that kind of stuff going on. What happens if you break the game up in small pieces? Can you do that without missing something?
Why is it that Nash seems to have had a big impact on the Suns concerning win % but when he left Dallas, that team won even more than when he was there? Because he was replaced well? Because they changed perhaps the system? Because he has even a bigger impact on the Phoenix system than that of Dallas? Because Nash his talents went even better with the Phoenix teammates than with the Dallas teammates?
A mix of some of the above reasons, if so how do you split the credit?
Can someone give a 100% clear explanation that is not open for discussion, or not, and if not why not? Could Nash his Dallas replacement do the same for the Phoenix offense? And so on...
I'm not convinced that we can measure very good the complete impact that a player has on winning games. |
|
| Back to top |
|
|
benji
Joined: 31 Dec 2004
Posts: 32
|
| Posted: Wed Mar 29, 2006 1:58 am Post subject: |
|
|
One quick measure I sometimes I look at, is a players PER * Minutes. Then I look at the next highest player on the team. I'm using week old numbers here because I'm lazy, please forgive me.
LeBron-Zydrunas: +34790
Wade-Shaq: +33271
Kobe-Odom: +31192
Dirk-Terry: +26730
Brand-Cassell: +26308
Gasol-Battier: +23593
Arenas-Jamison: +16251
Anthony-Miller: +12204
Billups-Hamilton: +9964
Marion-Nash: +7207
It's just a quick way to see who's had the biggest impact on his team. But I personally consider an MVP someone who leads the team with the least the farthest. It's why I would've voted for Dirk last year because he was so much better than the teams second best player. I suppose you could get even fancier and go like to the top five or top eight minute receivers.
Just my two cents on considering MVP. |
|
| Back to top |
|
|
Kevin Pelton
Site Admin
Joined: 30 Dec 2004
Posts: 979
Location: Seattle
|
| Posted: Wed Mar 29, 2006 2:42 am Post subject: |
|
|
| Analyze This wrote: | | I'm not convinced that we can measure very good the complete impact that a player has on winning games. |
If it weren't for the fact that I only got it from Aaron Schatz, I would take one of his favorite quotes, "The best is the enemy of better," as my motto.
Are our methods perfect? Of course not, and thank goodness. How boring would that be, if all you had to do was look up a player's PER to know his inherent God-given value? There would be no purpose to this message board, because what would we debate? Would would consultants tell teams? What would columnists write?
These are supposed to be tough questions. That's the whole point! That's the fun!
Are our methods good enough that they can inform the discussion and help guide it a more intelligent, comprehensive manner? I do believe so.
But everything is inherently very subjective.
Charles terms the Heat's .500 record without Shaq in the lineup as "mediocre"; I prefer average, becaue mediocre is so loaded. But Bryant's Lakers, mostly healthy, are not far above .500. Iverson's Sixers are below that mark. Garnett's Timberwolves well below it.
Are these players not good enough? Are their teams negligent, as MikeG suggests? Or is it that we're holding superstars to too high of a standard? I think Jordan kinda screwed things up for every swingman that follows him because it's not enough to win or succeed individually at a high level; they must do both, and heaven forbid they do so with the luxury of a quality 7-footer as a teammate. They'd better carry Bill Cartwright and Luc Longley to six championships, or else.
94x50 argues only three Cavaliers have made "significant" contributions to the team this year, which I see as a slap in Donyell Marshall's face. No, Marshall hasn't really lived up to expectations and he's only shooting 32% from 3-point range. But teams still seem to be fearing those 3s enough that the Cavs Offensive Rating is 5.9 points better with Marshall on the court, despite the fact that he's backing up two of the significant contributors.
So, in this discussion as in life, it's all about perspective. |
|
| Back to top |
|
|
Analyze This
Joined: 17 May 2005
Posts: 364
|
| Posted: Wed Mar 29, 2006 3:07 am Post subject: |
|
|
I agree with your methods/perspective remark. A lot of people are trying to get that perfect picture. The most of them know that perhaps it never will be 100% but they are trying to get as close as possible to perfection.
I actually wouldn't mind if we could tell for certain for example who is the mvp of this season. I don't think that if you can dissect and understand something completely it will get boring. |
|
| Back to top |
|
|
94by50
Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
|
| Posted: Wed Mar 29, 2006 9:47 am Post subject: |
|
|
| admin wrote: | | Are these players not good enough? Are their teams negligent, as MikeG suggests? Or is it that we're holding superstars to too high of a standard? I think Jordan kinda screwed things up for every swingman that follows him because it's not enough to win or succeed individually at a high level; they must do both, and heaven forbid they do so with the luxury of a quality 7-footer as a teammate. They'd better carry Bill Cartwright and Luc Longley to six championships, or else. |
I think you're right.
| admin wrote: | | 94x50 argues only three Cavaliers have made "significant" contributions to the team this year, which I see as a slap in Donyell Marshall's face. No, Marshall hasn't really lived up to expectations and he's only shooting 32% from 3-point range. But teams still seem to be fearing those 3s enough that the Cavs Offensive Rating is 5.9 points better with Marshall on the court, despite the fact that he's backing up two of the significant contributors. |
You're right... although I really meant that only Big Z had been a significant contributor for the entire year. Two months of Larry Hughes and one month of Flip Murray didn't seem to pass muster, I dismissed Eric Snow fairly quickly, and I didn't consider anyone on the Cavs' bench. I dismissed Marshall because he has played below expectation, although that was probably insufficient reason to do so. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
