ImageImageImage

TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP)

Moderators: bisme37, Froob, Darthlukey, Shak_Celts, Parliament10, canman1971, shackles10, snowman

humblebum
Banned User
Posts: 11,727
And1: 1,755
Joined: Jan 20, 2005

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#21 » by humblebum » Fri Dec 5, 2008 2:14 pm

Pete, I think it's clear that a player's fit amongst his teammates is key to his individual success. But all teams are attempting to bring cohesive rosters together which build on individual players strengths. I think you could put Rondo on plenty of teams and they would get better, especially with the way he's been playing of late. Rondo is the ultimate example of making his teammates better... and that's almost always the main criteria of what a Superstar really is. Just look at that game against the Pacers... Ray, KG, and Perkins all played brilliantly and never had to work particularly hard in order to do so. So when Rondo is playing this way surrounded by his current cast the team is truly Great (with a capital G), if he was playing with a lesser team he'd make them better... maybe not championship worthy but he'd knock them up a couple of notches.

I don't think it's been done yet... can we get a sticky on this thread. Great stuff Pogue!
Gant
RealGM
Posts: 10,929
And1: 15,221
Joined: Mar 16, 2006

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#22 » by Gant » Fri Dec 5, 2008 2:42 pm

GuyClinch wrote:
With what we've seen lately from Rondo I'm not surprised at all by these statisics. He's been the best player and most impactful player on the team for the past 6-7 games. He's probably been our most vital and important piece since last year. The team's success seems to be firmly based on Rondo's ability to initiate and dominate at the point of attack on both ends of the court.


Well obviously he has played great lately. I do think this statistic overvalues him though. It rates him as a superstar. But on a team with poor offensive players (like that 24 win team) he isn't that useful. Rondo is a great fit for our team but on other teams he wouldn't be a superstar. Whereas guys like KG and PP would be great on any team, IMHO.

The big problem is he doesn't shoot very well at the PG position. A guy like that is never a true superstar as far as I am concerned. I think that a team like the Cavs wouldn't be as good with Rondo as with Mo Williams for example. Whereas if you swapped KG for big Z you know they would be better.



I think this does overvalue Rondo somewhat, but regarding that 24 win team, Rondo was the only guy that had a positive +/- on that squad. (And he's gotten much better since then.) So even on a bad team he was very useful.

Cousy wasn't a pure shooter either, but he was a superstar. Magic wasn't a good shooter when he entered the league. Rondo offsets his bad outside percentage by only taking smart shots and is #4 in the league in FG% at his position right now.

I'd much rather have a supposedly bad shooter like Rondo at 51.3% than a "great" shooter like Calderon at 45.1%, or a "great" shooter like Mike Bibby at 47.5%, or a scorer like Baron Davis at 38.4%.
humblebum
Banned User
Posts: 11,727
And1: 1,755
Joined: Jan 20, 2005

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#23 » by humblebum » Fri Dec 5, 2008 3:05 pm

I think it's unfair that Rondo's game is always qualified with, "but he can't shoot." It's unfair because he does everything else at such a high level (ballhandling, great vision, an array of passing techniques to get the ball to the open man, rebounding!, dominating the transition game both ways, steals off his man and in the passing lanes, dribble penetration, drawing fouls) his BBIQ is off the charts, he's freakishly athletic, and he has about the perfect demeanor you could ask for in a floor general. He's possibly one of the most versatile, impact the game in all the different areas, players in the NBA. Outside of shooting I can't really think of one thing he doesn't do really well. He's on his way to becoming a Great player and I don't even think he needs the jumpshot to be good (just around slightly below average would be fine) to get there.
GuyClinch
RealGM
Posts: 13,345
And1: 1,478
Joined: Jul 19, 2004

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#24 » by GuyClinch » Fri Dec 5, 2008 6:52 pm

I think it's unfair that Rondo's game is always qualified with, "but he can't shoot." It's unfair because he does everything else at such a high level (ballhandling, great vision, an array of passing techniques to get the ball to the open man, rebounding!,


It's not unfair if you understand the value of shooting. We aren't watching tennis where you can win with speed and defense alone. In that game you could just force people into mistakes and win matches.

In basketball though you have to get the ball in the basket to score any points. It's fundamentally an offensive game. That's why for all the lip service defense gets a team like the Celtics that wins is notable because has not one but two amazing scorers on it. You want to get scorers to play defense. Its much easier and more effective then get defenders to score.

Your never going to see a team of guys that can't shoot win a championship for this reason. Basketball's core skill is shooting. I'd give Rondo credit for making his mediocre shooting work for him and his teamates but I still think it disqualifies him from greatness at the PG spot.

I'd say TA has a bit of the same problem. I am a big TA fan - and he has been incredibly helpful. I happen to really like athletic players. But without guys to fill up the basket their utility is diminished.

That's why a guy like Chris Paul is not only better then JKidd now - he is better right now then Jason Kidd has ever been. Even without a touch of the speed that Rondo has - or the crazy long arms his shooting ability just elevates him into a superstar.

I'll give Danny some credit for wanting to move Pierce and draft him. That would have been an excellent move as well. OTOH I will point out that Rondo CAN shoot - to some extent. He just isn't an NBA level shooter. JKidd is a good comparison.
User avatar
Pogue Mahone
Head Coach
Posts: 6,006
And1: 738
Joined: Aug 09, 2003
Location: In the Sun
Contact:
     

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#25 » by Pogue Mahone » Fri Dec 5, 2008 7:43 pm

Top-30 PER (Min 250 MP), TEN+ Rating, Fair $alary

1. LeBron James, PER: 33.0, TEN+: 217, Fair$: $34,944,802.02
2. Dwyane Wade, PER: 31.2, TEN+: 215, Fair$: $35,797,521.21
3. Chris Paul, PER: 30.7, TEN+: 210, Fair$: $39,451,556.12
4. Dwight Howard, PER: 28.4, TEN+: 140, Fair$: $14,976,687.70
5. Devin Harris, PER: 27.6, TEN+: 181, Fair$: $24,676,645.58
6. Tony Parker, PER: 27.1, TEN+: 186, Fair$: $23,491,710.15
7. Carlos Boozer, PER: 25.3, TEN+: 161, Fair$: $17,729,067.06
8. Chris Bosh, PER: 25.1, TEN+: 149, Fair$: $19,028,399.32
9. KobeStopKobeDon't, PER: 25.1, TEN+: 171, Fair$: $20,051,948.11
10. Tim Duncan, PER: 25.0, TEN+: 157, Fair$: $17,099,307.38
11. Zydrunas Ilgauskas, PER: 24.5, TEN+: 141, Fair$: $11,292,077.34
12. Dirk Nowitzki, PER: 23.6, TEN+: 138, Fair$: $14,402,474.78
13. Vince Carter, PER: 23.4, TEN+: 145, Fair$: $16,421,913.47
14. Antawn Jamison, PER: 23.3, TEN+: 138, Fair$: $15,047,370.82
15. Brandon Roy, PER: 22.8, TEN+: 159, Fair$: $19,043,701.81
16. Amare Stoudemire, PER: 22.8, TEN+: 132, Fair$: $13,494,125.75
17. Al Jefferson, PER: 22.2, TEN+: 121, Fair$: $11,080,886.64
18. Pau Gasol, PER: 21.9, TEN+: 134, Fair$: $12,713,870.66
19. Andris Biedrins, PER: 21.6, TEN+: 127, Fair$: $11,269,740.41
20. Yao Ming, PER: 21.6, TEN+: 132, Fair$: $11,579,763.31
21. Chauncey Billups, PER: 21.5, TEN+: 177, Fair$: $20,738,314.56
22. Andrei Kirilenko, PER: 21.4, TEN+: 164, Fair$: $15,720,215.39
23. Shaquille O'Neal, PER: 21.3, TEN+: 128, Fair$: $9,242,810.98
24. Caron Butler, PER: 21.1, TEN+: 140, Fair$: $15,533,794.16
25. Andrew Bynum, PER: 21.0, TEN+: 116, Fair$: $8,604,609.34
26. Trevor Ariza, PER: 20.8, TEN+: 141, Fair$: $11,158,638.84
27. Kevin Garnett, PER: 20.8, TEN+: 160, Fair$: $17,780,832.02
28. Jason Terry, PER: 20.5, TEN+: 124, Fair$: $12,071,322.84
29. Joe Johnson, PER: 20.4, TEN+: 147, Fair$: $15,594,302.42
30. Carl Landry, PER: 20.4, TEN+: 115, Fair$: $6,491,264.65
GuyClinch
RealGM
Posts: 13,345
And1: 1,478
Joined: Jul 19, 2004

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#26 » by GuyClinch » Fri Dec 5, 2008 8:20 pm

Well I give it to ya - eyeball wise Ten+ seems a little better then PER. Some of those high PER guys that we know are dogs defensively (Al Jefferson, Bosh) score much lower on TEN then PER.
User avatar
Pogue Mahone
Head Coach
Posts: 6,006
And1: 738
Joined: Aug 09, 2003
Location: In the Sun
Contact:
     

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#27 » by Pogue Mahone » Fri Jan 16, 2009 12:34 am

Through 41 games ...

Allen, Ray - Scr: 21 Reb: 5 Bhd: -1 Dsrp: 2
TEN + : 114
FAIR $: $8.044M
OPPST%: .728

Allen, Tony - Scr: 14 Reb: 5 Bhd: -3 Dsrp: 5
TEN + : 69
FAIR $: $1.82M
OPPST%: .507

Davis, Glen - Scr: 6 Reb: 9 Bhd: -1 Dsrp: 1
TEN + : 58
FAIR $: $1.39M
OPPST%: .492

Garnett, Kevin - Scr: 20 Reb: 13 Bhd: 1 Dsrp: 6
TEN + : 163
FAIR $: $16.578M
OPPST%: .752

House, Eddie - Scr: 16 Reb: 5 Bhd: 0 Dsrp: 4
TEN + : 89
FAIR $: $3.919M
OPPST%: .493

O'Bryant, Patrick - Scr: 19 Reb: 17 Bhd: -10 Dsrp: 7
TEN + : 79
FAIR $: $2.265M
OPPST%: .426

Perkins, Kendrick - Scr: 12 Reb: 16 Bhd: -3 Dsrp: 5
TEN + : 110
FAIR $: $6.418M
OPPST%: .779

Pierce, Paul - Scr: 20 Reb: 7 Bhd: 0 Dsrp: 4
TEN + : 149
FAIR $: $15.557M
OPPST%: .726

Powe, Leon - Scr: 15 Reb: 16 Bhd: -4 Dsrp: 5
TEN + : 94
FAIR $: $3.568M
OPPST%: .495

Pruitt, Gabe - Scr: 11 Reb: 6 Bhd: 2 Dsrp: 4
TEN + : 83
FAIR $: $2.711M
OPPST%: .441

Rondo, Rajon - Scr: 12 Reb: 8 Bhd: 8 Dsrp: 7
TEN + : 145
FAIR $: $14.102M
OPPST%: .756

Scalabrine, Brian - Scr: 11 Reb: 5 Bhd: 0 Dsrp: 3
TEN + : 73
FAIR $: $1.783M
OPPST%: .538

Advanced Statistical Scouting (A.S.S.) Stats
Scr: Scoring (Lg. Avg. = 15) - Gives neutral weight to the ability to score (ie. balances out volume shooters vs low-usage efficiency.)
Reb: Rebounding (Lg. Avg. = 9) - Additional weight is given to those who procure the offensive variety.
Bhd: Ballhandling (Lg. Avg. = 0) - A negative rating isn't a death sentence but it is a sign that someone else on the team needs to be handling the ball and playmaking.
Dsrp: Disruptive Defense (Lg. Avg. = 4) - Players with a rating </= 4 aren't necessarily bad defenders, they just need to compensate for their inability to make consistent defensive plays with very good team defense.

TEN+ - Rating system that is similar in concept to OPS+ or ERA+ in baseball. Attempts to incorporate defense by using regression of on-court/off-court data. TEN+, at the team level, has consistently outperformed pythag in determining expected winning percentage.
190+ : MVP Canidate
150+ : Superstar
130+ : All-Star
100+ : Starter
_75+ : Rotation
_60+ : Deep Bench
</=59: Spot minutes/Freely Available Talent*

*To put it in simpler terms, a 60 TEN+ performance is roughly replacement level. For those familiar with PER, a 9.00 PER is also considered replacement level.

FAIR $: Estimated monetary value of the player's performance that is arrived at via TEN+. Roughly 2/3rds of a player's performance value is derived from his offensive contributions. The other 1/3rd is arrived at via defensive impact and intangibles (with the help of regression analysis.) It an not an exact .667/.333 split (over the last three years it has averaged out to be approximately .685/.315) but it generally holds true.

OPPST%: Opponent Starters Faced, as a percentage. If a player starts every game, playing only one minute each game, against all five of each team's starters, he would have a opponent's starters faced of 5.00 and a percentage of 1.00%.
elrod enchilada
Sixth Man
Posts: 1,729
And1: 0
Joined: Feb 19, 2004
Location: Land of Lincoln
Contact:

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#28 » by elrod enchilada » Fri Jan 16, 2009 2:16 am

Thanks, Pogue. Looks like Rondo has slipped from his superstar status, and that seems pretty accurate.

Just wondering, what is your take on Powe and on O'Bryant?
elrod enchilada
Sixth Man
Posts: 1,729
And1: 0
Joined: Feb 19, 2004
Location: Land of Lincoln
Contact:

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#29 » by elrod enchilada » Fri Jan 16, 2009 2:21 am

One other question: Do you have a list of, say, the top 25 or 50 players in the league according to TEN+? It would be interesting to see.
User avatar
ParticleMan
Retired Mod
Retired Mod
Posts: 15,070
And1: 9,071
Joined: Sep 16, 2004
     

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#30 » by ParticleMan » Fri Jan 16, 2009 5:34 pm

I like these stats because they show that KG is really our best player. PP is close. That reflects my impressions as well. I still think Rondo is a tad overrated here, and Ray underrated. The rest are about right. Powe and House are the best bench guys.
User avatar
Pogue Mahone
Head Coach
Posts: 6,006
And1: 738
Joined: Aug 09, 2003
Location: In the Sun
Contact:
     

Re: TEN+, A.S.S. Stats, Opposition Starters (20 GP SSS) 

Post#31 » by Pogue Mahone » Fri Jan 16, 2009 10:13 pm

ParticleMan wrote:I like these stats because they show that KG is really our best player. PP is close. That reflects my impressions as well. I still think Rondo is a tad overrated here, and Ray underrated. The rest are about right. Powe and House are the best bench guys.


The main reason for the drop off in Rondo's numbers are 1. He was in the middle of a hot streak and at the previous 20 games was a small sample size and 2. He is not using scoring possessions. All his secondary numbers (or off the ball numbers, whatever you want to call it) have held constant.

This system, like every system, has some problems with the efficiency vs quantity conundrum. I think it does a much better job than PER, for example, in that regard.

Secondly, in regards to Ray, the reason he is ranked as he is, imo, is that he uses a lot of possessions at excellent efficiencies. What does he do, on a consistent level, outside of that? If you were to look at #20 and see the name George Smith and knew that he was a 15-20 mpg scorer off of the bench for the South Dakota Calumets and not Ray Allen, the rating makes a lot of sense.

Everything in my system is measured on a per minute basis, adjusted for pace. With that in mind, consider the following:

Of all NBA guards, 6-5 and under, who have played at least 250 MP, here are some of Ray Allen's secondary statistics ( Rank among the 99 players queried)

BLK% 0.2 (63rd) - Block King Chris Quinn is at 0.2% in 2009
STL% 1.7 (78th) - Intergalactic ball-thief Jose Calderon is at 1.3% in 2009
ORB% 3.2 (20th) - Ray has actually been solid in this regard
DRB% 8.0 (80th) - Noted Windex man Allen Iverson is at 8.6% in 2009
AST% 12.7 (76th) - Professional gunner Flip Murrary is at 12.4% in 2009

Long story short, while Ray's fantastic individual efficiencies and high usage rate certainly provide good value, his off-the-ball activity is severely lacking. In other words, Ray has become a glorified gunner (albeit with great efficiency.) His defensive impact is roughly equivalent to what Dan Freakin Dickau has done in his career. And that may seem like hyperbole but I say it in all seriousness.

Really, there is not a lot that separates Eddie House and Ray Allen, right now, in terms of on-court value. In fact, Eddie House secures more defensive rebounds (11.4%), steals the ball at a much greater rate (2.3%), blocks a greater amount of shots (0.6%) and assists at slightly a lesser rate (11.8%.) So what separates House and Allen is Ray's shooting prowess. Both are gunners but at least House shows some off the ball activity.

Now for Rondo, I will do the same deal as I did with Ray:

BLK% 0.4 (39th)
STL% 3.6 (2nd)
DRB% 12.7 (10th)
ORB% 4.9 (6th)
AST% 38.8 (7th) - And the six above him are all first or second options -- not fourth

So Rondo is clearly contributing in more ways than Ray. I can count four elite level skills for his position that have manifested themselves into production just from those simple measurements. The argument, as I understand it, is that Ray should count more because he is putting the ball in the basket more and with greater frequency.

My counter-argument is that Ray's teammates are responsible for much of Ray's offense. Yes, Ray has to make his shots but 68% of Ray's conversion are assisted. For Rondo, 31%. Rondo creates large amounts of offense, both for teammates and himself, and despite teammates making those shots that he assists on, he does get a share of the credit.

Now, role and opportunity certainly play into each player's production and performance. I think that goes without saying. That being said, any system, from simple shooting percentages to more advanced metrics, is only measuring performance and not skill level.

FWIW, Ray Allen is rated by PER to be approximately 15.33% above average. In my system, he is 14%.

By PER, Rajon Rondo is rated ~24.67% above average. In my system, he is 45%.

Now consider that my system at least makes an attempt to quantify defensive impact by using regression and PER doesn't outside of blocks, steals and defensive rebounds. So, my system values Ray's offensive impact more than PER but penalizes him more because of his defense and off the ball contributions (or lack thereof.)

If Ray was even an average defender, imo, he would likely rank at a high All-Star level. Remember, he is not creating for others, he is a very poor defensive rebounder and he doesn't force turnovers. The fact of the matter is that he isn't even average as a defender. Which is fine but it stands to reason that a system that attempts to quantify performance is going to penalize Ray in areas that he is lacking and reward him in the areas he is good in.

I hope that clarifies things a bit.

Bob,

I will try to get something up as time permits.
floyd
Bench Warmer
Posts: 1,414
And1: 649
Joined: Aug 04, 2006

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#32 » by floyd » Fri Jan 16, 2009 10:49 pm

What's your regression model look like?
User avatar
Pogue Mahone
Head Coach
Posts: 6,006
And1: 738
Joined: Aug 09, 2003
Location: In the Sun
Contact:
     

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#33 » by Pogue Mahone » Sat Jan 17, 2009 12:19 am

I use a GLM.
Jammer
General Manager
Posts: 8,792
And1: 3,314
Joined: Mar 06, 2001
Contact:
 

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#34 » by Jammer » Sat Jan 17, 2009 12:47 am

Pogue, with 41 games of data available for this season,
why would anyone use PER over NET PER.

On a small sample size I can see a problem with NET PER,
with 41 games available,
I think it's far superior to straight PER.
floyd
Bench Warmer
Posts: 1,414
And1: 649
Joined: Aug 04, 2006

Re: TEN+, A.S.S. Stats, Opposition Starters (Through 41 GP) 

Post#35 » by floyd » Sat Jan 17, 2009 1:50 am

Pogue Mahone wrote:I use a GLM.


Poisson or binomial data? Oy, I always try to steer clear if I can.

I actually meant the parameters in your model. You mentioned +/-. Is it just player vs. +/- with 41 replicates? Just thinking about the obvious lack of independence between players but I have no idea how you'd deal with it.

Return to Boston Celtics