 |
APBRmetrics
The statistical revolution will not be televised.
|
| View previous topic :: View next topic |
| Author |
Message |
Dan Rosenbaum
Joined: 03 Jan 2005
Posts: 96
Location: Greensboro, North Carolina
|
 Posted: Wed Mar 02, 2005 4:06 pm Post subject: |
 |
|
| Quote: | | Uncertainty and undervalued aren't the same thing, and Dan Rosenbaum has done some persuasive research that tends to indicate that assists might actually be less valuable than they're generally credited as, at least for point guards, but I've yet to be completely convinced by it. I'd say I operate from the principle that all positions are equally important, and try to rate players according to that ideal. If that means more weight to assists, so be it. |
Another way of reading my work is that I am not completely convinced of it either. As I pointed out earlier in in this thread, my system seems to bend over backwards for Stockton but doesn't find much love for Nash. What that tells me is that to be an elite player, it is important to be productive on both ends of the floor. One end is just not good enough. For a refresher on how much my system likes Stockton, here is a repost.
| Quote: | With the statistical index derived from adjusted plus/minus statistics, here are my rankings for John Stockton dating back to 1993-94.
93-94: #6
94-95: #2
95-96: #6
96-97: #4
97-98: #9
98-99: #4
99-00: #6
00-01: #4
01-02: #8
02-03: #12
So I think it can be said that my system is more than fair to John Stockton-type players.
So where does Steve Nash end up in my ratings this season? He ends up at #40. He surely is better than that, but there just have not been a lot of players like Nash who are so good on the offensive end and so invisible on the defensive end. Just a little over one steal per 40 minutes is incredible for a point guard as quick as Steve Nash. So my system or pretty much anybody's system is going to have a hard time figuring out what to do with Nash.
What seems pretty uncontroversial is that his skills are a perfect fit for what Phoenix is trying to do. He is a bit overrated because his substitutes are pretty bad, but I don't think it is absurd that Dallas can be better without him (and with Dampier and Terry) and he could still be the Most Valuable Player in the league. Most Valuable is different from best.
At this time I will not argue with anyone who calls him an MVP candidate, but I still don't think he is as good as Stockton was during most of his career. |
|
|
| Back to top |
|
 |
KD
Joined: 30 Jan 2005
Posts: 9
|
| Posted: Wed Mar 02, 2005 4:22 pm Post subject: |
|
|
| CrazyFromTheHeat wrote: | | Let's face it. Choosing Nash is sexy because its so easy to see Phoenix before and after, and how they play when he's not there. But is he really more important to his team than Tim Duncan? There aren't many that expect Phoenix to win the title, so he's not likely to make the ultimate difference. The Spurs are the favorites, and Duncan is the main reason. It's a boring, but easy choice. |
Agreed, but:
*1 SanAntonioSpurs 43 13 17 - .768
*2 PhoenixSuns 43 14 16 0.5 .754
And as Pelton's second part explains, it's not exactly a "Nash For MVP" article, it's a "Nash For MVP?" -- without someone messing with the teleprompter.
Still, Nash will have to do a lot more over the next seven weeks to outpace Duncan. |
|
| Back to top |
|
|
admin
Site Admin
Joined: 30 Dec 2004
Posts: 151
Location: Seattle
|
| Posted: Thu Mar 03, 2005 12:20 am Post subject: |
|
|
| KD wrote: | | And as Pelton's second part explains, it's not exactly a "Nash For MVP" article, it's a "Nash For MVP?" -- without someone messing with the teleprompter. |
Yes - Kevin gave me a good chewing out over omitting the question mark when I originally posted this thread.
I hope people realize my intent wasn't to advance any particular position on Nash, though of course I have my opinions, but instead to get us talking and thinking about the issue.
A fear I have about the statistical community is that there we've built up our own sort of "conventional wisdom" and that we're often as resistent to challenges to that CW as anyone else. I certainly feel that same thing myself, and I felt it when DanR made his argument against possession-based linear weights last year.
Obviously, new ideas must be subjected to a certain amount of scrutiny before they become accepted. But I wonder sometimes whether non-CW ideas are even getting that chance. |
|
| Back to top |
|
|
admin
Site Admin
Joined: 30 Dec 2004
Posts: 151
Location: Seattle
|
| Posted: Thu Mar 03, 2005 12:28 am Post subject: |
|
|
Dan, your ratings are naturally exempt from the "abstraction of value" criticism, and I always find them interesting. I hope I didn't imply that you're advocating the "assists are less valuable" thinking as opposed to bringing forward the evidence.
While the "both ends" theory makes sense, I do find it a bit difficult to believe that a play or two makes such a big difference in value. |
|
| Back to top |
|
|
GreggGeth
Joined: 16 Feb 2005
Posts: 8
|
| Posted: Thu Mar 03, 2005 1:15 am Post subject: Nash for MVP |
|
|
I think a big question w/ Nash that needs to be asked is how responsible he is for his team's improvement.
Amare's scoring 57.3% from the field. His first two seasons, he was at about 47%. I do expect there to be a leap in improvement from 2nd to 3rd year in the NBA but a leap like that is unheard of. Has any player of Amare's counterpart ever improved his FG% like that in one season? Shawn Marion's FG% is up from last season, the highest its been since his 2nd year in the league. Both Joe Johnson's FG and 3PT FG are at career highs.
Not sure how to prove this, but I think not only is Nash passing the ball, but he's passing the ball in ways that demonstrably make his team better. Amare is getting the ball in ways that make it easier for him to score. So are some of the other Suns. And I think that has to be somewhat given to Nash.
Does anyone have stats handy of how the Suns have played when Nash is out of the lineup? When he went down early in the season, the difference was off the charts.
These Suns are a fascinating team. It's the right system for the right players. And Nash is the person who makes the offense click. On top of it, he is incredibly efficient himself. So I say, yes, even without playing much defense, he is the player most responsible for the Suns success.
Admitedly, Nash doesn't have the all-around game of Duncan, Nowitzki, Lebron or probably a bunch of other players in the league. Nash's best trait is making it easier for his teammates to score. And he does that better than any player in the league today and arguably better than any player aside from John Stockton. And because of that, he would get my MVP vote. |
|
| Back to top |
|
|
GreggGeth
Joined: 16 Feb 2005
Posts: 8
|
| Posted: Thu Mar 03, 2005 1:21 am Post subject: PER |
|
|
| Are there any stats out there taht keep track of PER year-by-year? I would hedge my bet and say that several of the Suns' players increased their PER from previous seasons. And some of the improved offensive efficiency comes from playing with a superior point guard. |
|
| Back to top |
|
|
Mike G
Joined: 14 Jan 2005
Posts: 74
Location: Delphi, Indiana, USA
|
| Posted: Thu Mar 03, 2005 6:45 am Post subject: Re: Nash for MVP |
|
|
| GreggGeth wrote: | | ... Has any player of Amare's counterpart ever improved his FG% like that in one season? |
Looking at TS%, rather than FG or other %
Amare Stoudemire 2004-05
PPG: 20.6 -- 26.1
TS%: .536 -- .620
Bob McAdoo 1973-74
PPG: 18.0 -- 30.6
TS%: .498 -- .594
Darryl Dawkins 1980-81
PPG: 14.7 -- 14.0
TS%: .548 -- .641
Larry Foust 1954-55
PPG: 15.1 --17.0
TS%: .483 -- .570
Wilt Chamberlain 1966-67
PPG: 33.5 -- 24.1
TS%: .547 -- .637
Dale Ellis 1986-87
PPG: 7.1 -- 24.9
TS%: .502 -- .588
Matt Harpring 2002-03
PPG: 11.8 -- 17.6
TS%: .512 -- .588
Artis Gilmore 1974-75
PPG: 18.7 -- 23.6
TS%: .531 -- .615 |
|
| Back to top |
|
|
Mike G
Joined: 14 Jan 2005
Posts: 74
Location: Delphi, Indiana, USA
|
| Posted: Thu Mar 03, 2005 9:30 am Post subject: |
|
|
I found some more:
+eff% eff% eff% PPG PPG
.127 .484 .612 11.5 21.5 Kiki Vandeweghe 1981 82
.097 .498 .594 18.0 30.6 Bob McAdoo 1973 74
.096 .491 .587 16.8 20.4 Roy Tarpley 1990 91
.090 .547 .637 33.5 24.1 Wilt Chamberlain 1966 67
.086 .502 .588 7.1 24.9 Dale Ellis 1986 87
.080 .457 .537 5.9 21.0 Bob Love 1969 70
.079 .491 .570 18.5 21.6 John Drew 1975 76
.076 .503 .579 21.0 21.3 Dick Van Arsdale 1969 70
.074 .514 .587 19.3 24.5 Rudy Tomjanovich 1973 74
.073 .459 .533 21.6 24.2 John Havlicek 1969 70
.084 : That's Stoudemire's +eff% so far this year. |
|
| Back to top |
|
|
S.K.
Joined: 18 Feb 2005
Posts: 7
Location: Toronto
|
| Posted: Thu Mar 03, 2005 3:42 pm Post subject: Re: Nash for MVP |
|
|
| GreggGeth wrote: | I think a big question w/ Nash that needs to be asked is how responsible he is for his team's improvement.
|
Well, let's take a look at how Stoudemire has improved from last year in some areas that don't fall in the "look at those great feeds from Nash!" category. Using, of course, knickerblogger.net and my copy of the 04-05 Basketball Forecast.
---------03-04-05
FT%: .661-.713-.724
RebRt: 15.9-14.4-12.8
AssRt: 6.0-5.9-6.3
TORt: 14.4-13.5-9.9
Usg: 18.3-22.9-25.1
Interesting, not the slam dunk I'd expected - in particular, his rebound rate is down significantly, which is partly a result of his move to C but still not what you'd expect from a big man having a breakout season.
The signs are there, though - he cut his TO rate while significantly increasing the number of possessions he's using, which is pretty impressive. Could this be attributable to getting passes from Nash in better spots, enabling Amare to simply shoot instead of being forced to create on his own? It can't be ENTIRELY from Nash, I'd say - some of that has to come from Stoudemire himself. Let's take a trip over to 82games.com.
Here we see that Stoudemire's % of baskets assisted is up to 60% from 46% last year. Also, he's getting to the line on 23.5% of his FGA, up from 19.7% in 2004. Interestingly, both of these totals are nearly identical to his rookie year (55% and 23.1%, to be exact), but this probably reflects Stoudemire's improved ability to look for his own shot between years 1 and 2. These numbers seem to support the Nash thesis.
Looking at Amaré's shot breakdown, however, we see a significant change. In 02-03, 40% of Stoudemire's shots were jumpers, which he converted at a terrible .307%. Last year, he took 53% of his shots from outside, and though the number rose to .356% it's still pretty bad. This year, though, a breakthrough: jumper % stays the same, 51%, but now he's converting them at a .433 clip, and only 50% of these are assisted. This is significant progress, and can't (I don't think) be attributed to Nash.
On "inside" shots, of course, the expected change: from .588 FG% to .615 to .718. Fewer blocked, and far more assisted.
It should be noted, though, that not all of these improvements can be necessarily laid at Nash's door - how many of these open shots come simply from being the lynchpin of a lineup full of shooters playing at full-tilt? I'm not sure what conclusions can really be drawn from these numbers, but I think I've proven to my own satisfaction that Stoudemire has made significant enough strides in his game on his own (TO rate, improved jumper) that to chalk up his improvement simply to playing with Steve Nash is ludicrous.
_________________
No books - no articles - no website.
Just opinions.
Ill-informed opinions. |
|
| Back to top |
|
|
Carlos
Joined: 21 Jan 2005
Posts: 6
Location: Montevideo, Uruguay
|
| Posted: Thu Mar 03, 2005 4:22 pm Post subject: |
|
|
| A small point about the value of assists. One of the things that confuses the issue is that we are treating the same way assists that are very different. A pass that leads to a dunk "creates" a high percentage shot while a pass that leads to a 18 foot jump shot "creates" a much lower percentage shot. But if the shots go in, both passes are treated the same. I did a quick check comparing Nash and Tony Parker and noticed that 46% of Nash assists lead to close shots while only 29% of Parker assists lead to close shots. Maybe we should assign value to assists differently according to shot type. |
|
| Back to top |
|
|
WizardsKev
Joined: 03 Jan 2005
Posts: 98
Location: Washington, DC
|
| Posted: Thu Mar 03, 2005 5:10 pm Post subject: |
|
|
| Carlos wrote: | | A small point about the value of assists. One of the things that confuses the issue is that we are treating the same way assists that are very different. A pass that leads to a dunk "creates" a high percentage shot while a pass that leads to a 18 foot jump shot "creates" a much lower percentage shot. But if the shots go in, both passes are treated the same. I did a quick check comparing Nash and Tony Parker and noticed that 46% of Nash assists lead to close shots while only 29% of Parker assists lead to close shots. Maybe we should assign value to assists differently according to shot type. |
But then there's a chicken-egg thing. Did Nash's passing "create" the close shot, or are those close shots the product of his teammates making great cuts to get open? Is that a function of system, or the value of the assist?
_________________
If you can't explain it simply, you don't understand it well enough.
-- Albert Einstein |
|
| Back to top |
|
|
Carlos
Joined: 21 Jan 2005
Posts: 6
Location: Montevideo, Uruguay
|
| Posted: Thu Mar 03, 2005 8:38 pm Post subject: |
|
|
Wizardzkev wrote | Quote: | | But then there's a chicken-egg thing. Did Nash's passing "create" the close shot, or are those close shots the product of his teammates making great cuts to get open? Is that a function of system, or the value of the assist? |
I'm not sure if it's the chicken or the egg, but if guys have very different rates of "close shot assists" season after season, I think we can make a reasonable assumption that their ability in making that kind of pass is different. I think it may be a way to try to separate "system assists" as Dan calls them, from "real assists". |
|
| Back to top |
|
|
WizardsKev
Joined: 03 Jan 2005
Posts: 98
Location: Washington, DC
|
| Posted: Fri Mar 04, 2005 6:07 am Post subject: |
|
|
| Carlos wrote: | Wizardzkev wrote | Quote: | | But then there's a chicken-egg thing. Did Nash's passing "create" the close shot, or are those close shots the product of his teammates making great cuts to get open? Is that a function of system, or the value of the assist? |
I'm not sure if it's the chicken or the egg, but if guys have very different rates of "close shot assists" season after season, I think we can make a reasonable assumption that their ability in making that kind of pass is different. I think it may be a way to try to separate "system assists" as Dan calls them, from "real assists". |
I'm not sure that's a reasonable assumption, though. 
I looked back at previous years numbers. Last season, Nash's inside assist rate was about 40% -- Parker's was about 32%.
In 02-03, Nash's inside assist rate was 34%, but Parker's was 38%.
It could be that Nash has become a better passer the past couple seasons while Parker has gotten worse. There could be other factors at work too. This year, for example, the Spurs are running a good portion of their offense for Ginobili (when he's on the court), and less for Duncan. I don't have time today to look at all the stat indicators I'd be interested in examining, but I'd need some more convincing.
Basically, I think "inside assists" can be system assists as much as a jumper assist can be.
For example, Dwyane Wade's inside assist rate is 40%. How many of those are because their system calls for him to feed Shaq? Ridnour's inside assist rate is about 31%. Is he a worse passer than Wade, or is Seattle's system geared more for perimeter shooters?
_________________
If you can't explain it simply, you don't understand it well enough.
-- Albert Einstein |
|
| Back to top |
|
|
Mike G
Joined: 14 Jan 2005
Posts: 74
Location: Delphi, Indiana, USA
|
| Posted: Fri Mar 04, 2005 8:46 am Post subject: |
|
|
Let's not forget that there are (counted) Assists, and there are potential assists: passes that would be an assist if the shot is made.
If Shaq makes 10 passes out to the arc, and guys make 4 shots, he gets 4 assists on his 10 passes. If Wade dishes to guys closer to the basket (some of whom are named Shaq), he's likely to get 8 assists from 10 passes.
The FG% self-corrects for the effectiveness of the pass. |
|
| 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
