Why can't Smoove learn from Horford?

[Jody23]

Do you think he would stay in Atlanta for not max money, if Milwaukee offers it to him?

Yes, I do. We all know Smoove doesn't like Milwaukee. But moreover, he wants to be apart of a winner here and I think he would and should accept much less than max to stay. As David Aldridge wrote recently, players love Atlanta and there will be free agents who want to come to the ATL due to the new circumstances.

[AHF]

Okay so when you say you're fine with efficiency, you're simply talking about shooting efficiency? You don't put stock in either Player Efficiency or PER, although you feel that PER is the better of the 2.

So now that we have that figured out, what do you consider to be the single best metric for measuring a players overall impact?

The dictionary definition that best relates to sports statistics for me is:

accomplishment of or ability to accomplish a job with a minimum

expenditure of time and effort: The assembly line increased industry's efficiency

Anything that measures efficiency should meet some basic criteria. Any measure of efficiency should relate to how much value you are generating out of your playing time, shot opportunities, etc. Any measure of efficiency should not reward volume except to the extent that the volume represents efficiency. So if you take two people and they do the exact same thing every minute on the floor but one does it over 25 minutes and the other over 20 minutes then the efficiency for both players should be identical and the value of that performance should depend on how efficient the players are.

So PER fails on this standard because it rewards poor execution on a volume basis (the more 35% shooting, the higher the PER).

Player Efficiency fails on this standard because it rewards poor execution on a volume basis and because it rewards volume of minutes (players who play heavy minutes are heavily rewarded even if they are doing the same or worse than players with less minutes).

Of the two, PER comes closer to measuring efficiency but is still flawed (primarily based on rewarding volume production and its inability to meaningfully reflect defensive value).

IMO, there is no perfect number but my favorite for measuring overall value per minute is WS/48.

The best efficiency numbers are probably the %s. For scoring, I think we actually do have a good metric with TS%. For rebounding, rebounding % is similarly useful (TS% is better for what it measures because it is less teammate dependent but I think this is still better than simply rpg or rp36). Etc.

Edited February 27, 2013 by AHF

[Admin]

I agree that volume should not be rewarded and I would say that it should actually penalize the player. I like effective FG% and true shooting % as well as rebound % and things like assist ratio, etc. Just would be nice if there was a single stat that quantified things like that together.

Do you know how win shares are calculated?

[AHF]

I agree that volume should not be rewarded and I would say that it should actually penalize the player. I like effective FG% and true shooting % as well as rebound % and things like assist ratio, etc. Just would be nice if there was a single stat that quantified things like that together.

Do you know how win shares are calculated?

http://www.basketball-reference.com/about/ws.html

[Admin]

Thank you for posting the link to the formula. Here's where I have trouble with win shares. According to them, last year in 58 games Zaza produced 5 win shares (0.086 per game played). This year in 52 games Josh has produced 2.9 wins (0.055 per game played) which comes out to roughly 3.234 win shares over 58 games for Josh. So if you believe in win shares then you believe that last years Zaza produced about 50% more win shares than Josh has this year. It also shows that Korver and Teague have each produced about 50% more win shares than Josh this year as well as Horford producing 2.2 times more win shares than Josh this year. Does all of that sit right with you? Do you believe that last years Zaza was that much better than this years Josh? Do you believe that Korver and Teague are that much better this year than Josh or that Horford is more than 2 times better than Josh?

[drzachary]

AHF has summed it up pretty nicely. Here are my additions:

"We're not saying Josh is an efficient offensive player, we're just saying that he's an efficient player!"

I am not sure how this is possible. What definition of 'efficient' can include 'offensively inefficient?' "Josh Smith is one of the very best defenders in the league" gets no argument from me.

"Josh's offense doesn't invalidate all his defensive contributions!"

This is true, but only inasmuch as his offensive problems don't invalidate *all* of them. His offensive problems are what make him a top 100 player, not a top 25 one. Good, not great. Also-ran, not All-Star. He could literally fix all of these issues by taking less shots, which is sad, and is definitely largely due to coaching.

"Why is WS better than PER or NBA Efficiency?"

Easy: one reflects winning and one does not. PER/"Efficiency" can be generated for all players back to the late 70s pretty easily, and then compared to the actual win totals of the teams they were on. WS/WP gives us a highly correlated result; PER/Efficiency do not.

[Admin]

AHF has summed it up pretty nicely. Here are my additions:

"We're not saying Josh is an efficient offensive player, we're just saying that he's an efficient player!"

I am not sure how this is possible. What definition of 'efficient' can include 'offensively inefficient?' "Josh Smith is one of the very best defenders in the league" gets no argument from me.

"Josh's offense doesn't invalidate all his defensive contributions!"

This is true, but only inasmuch as his offensive problems don't invalidate *all* of them. His offensive problems are what make him a top 100 player, not a top 25 one. Good, not great. Also-ran, not All-Star. He could literally fix all of these issues by taking less shots, which is sad, and is definitely largely due to coaching.

"Why is WS better than PER or NBA Efficiency?"

Easy: one reflects winning and one does not. PER/"Efficiency" can be generated for all players back to the late 70s pretty easily, and then compared to the actual win totals of the teams they were on. WS/WP gives us a highly correlated result; PER/Efficiency do not.

Not sure who you're attributing these quotes too, but I said Josh isn't an efficient SHOOTER, I never said he's not an efficient offensive player. I consider his passing to be a big factor in the help that he brings to the offense.

And win shares has a lot of seemingly arbitrary numbers they are using to derive their numbers and last year and this year they're off by a full 2 games for the Hawks. That might not be a big deal but that along with how much better it tries to say Korver and Zaza have been for the Hawks the past 2 years makes me not put much belief into it.

[drzachary]

Thank you for posting the link to the formula. Here's where I have trouble with win shares. According to them, last year in 58 games Zaza produced 5 win shares (0.086 per game played). This year in 52 games Josh has produced 2.9 wins (0.055 per game played) which comes out to roughly 3.234 win shares over 58 games for Josh. So if you believe in win shares then you believe that last years Zaza produced about 50% more win shares than Josh has this year. It also shows that Korver and Teague have each produced about 50% more win shares than Josh this year as well as Horford producing 2.2 times more win shares than Josh this year. Does all of that sit right with you? Do you believe that last years Zaza was that much better than this years Josh? Do you believe that Korver and Teague are that much better this year than Josh or that Horford is more than 2 times better than Josh?

Yes to all of the above, with the caveat that it's not perfect, so maybe Zaza was only 33% or 40% better. The reason it doesn't "sit well" is that our eyes lie to us. :)

[drzachary]

Not sure who you're attributing these quotes too, but I said Josh isn't an efficient SHOOTER, I never said he's not an efficient offensive player. I consider his passing to be a big factor in the help that he brings to the offense.

And win shares has a lot of seemingly arbitrary numbers they are using to derive their numbers and last year and this year they're off by a full 2 games for the Hawks. That might not be a big deal but that along with how much better it tries to say Korver and Zaza have been for the Hawks the past 2 years makes me not put much belief into it.

They're not quotes, they're summaries of arguments I hear. Sorry for the confusion.

And as for WS being off 2 games, my reaction is: only 2 games! That's awesome. It's not perfect, generating expected wins out of something like PER/Efficiency, you're usually off 30% or so.

And there's nothing arbitrary about it, really. Arbitrary was how PER/Efficiency were created and the various factors weighted. With WS/WP, the weights were derived from actually measuring -- regressing -- what statistical factors were present in the box scores of teams that actually won games! Where it gets tricky is when you break it down to individual players, of course -- but the summing of wins helps bear it out a bit.

Edited February 27, 2013 by drzachary

[AHF]

Win shares is not perfect and will have to be further refined over time, I am sure (just like the WAR debates in MLB). The biggest positive I see for it the correlation that Dr. Z talks about. You should want a number that correlates strongly to actual wins, and Win Shares represent the best option for this of which I am aware.

As for Josh this season, I think it speaks to the fact that one of the worst offensive players in the league this season (see the earlier links showing Josh is in the bottom 3rd) is taking so many shots that it hurts the team.

Josh's offensive win share this season is -0.3, by far his career worst. Considering this is his career worst season for shooting and that the is turning the ball over more this season than in the past 4 years, he should be getting punished with his numbers for this play. Notably, the other guy with a -.3 OWS number for this season is the other guy that I think hurts the team with his craptastic offense - Anthony Tolliver (Tolliver is worse so his OWS/48 is significantly worse than Josh which is right).

Josh is leading the team in defensive win shares and is top 10 in the league, which sounds about right. He is hurting the team with his offense which also sounds right.

This doesn't hurt my head at the end of the day because I don't think replacing Josh with a more limited but much more efficient PF would significantly impact our team's record. I do think Josh taking good shots with the shot selection of someone like Steve Nash (who is patient and works for good % opportunities) would be a star and would really impact our record and I think that is consistent with his WS this season as well.

If Josh keeps up his recently improved play, his numbers will look better at the end of the season and he will deserve the improvement.

[MaceCase]

Bah, spent 20 minutes between filling to type out a response and AHF already beat me to the punch.

[benhillboy]

Go Josh, Go Al, Go Hawks. We're all connected in that sentiment I'm sure.

Edited February 27, 2013 by benhillboy

[drzachary]

Win shares is not perfect and will have to be further refined over time, I am sure (just like the WAR debates in MLB). The biggest positive I see for it the correlation that Dr. Z talks about. You should want a number that correlates strongly to actual wins, and Win Shares represent the best option for this of which I am aware.

As for Josh this season, I think it speaks to the fact that one of the worst offensive players in the league this season (see the earlier links showing Josh is in the bottom 3rd) is taking so many shots that it hurts the team.

Josh's offensive win share this season is -0.3, by far his career worst. Considering this is his career worst season for shooting and that the is turning the ball over more this season than in the past 4 years, he should be getting punished with his numbers for this play. Notably, the other guy with a -.3 OWS number for this season is the other guy that I think hurts the team with his craptastic offense - Anthony Tolliver (Tolliver is worse so his OWS/48 is significantly worse than Josh which is right).

Josh is leading the team in defensive win shares and is top 10 in the league, which sounds about right. He is hurting the team with his offense which also sounds right.

This doesn't hurt my head at the end of the day because I don't think replacing Josh with a more limited but much more efficient PF would significantly impact our team's record. I do think Josh taking good shots with the shot selection of someone like Steve Nash (who is patient and works for good % opportunities) would be a star and would really impact our record and I think that is consistent with his WS this season as well.

If Josh keeps up his recently improved play, his numbers will look better at the end of the season and he will deserve the improvement.

Also worth noting that his offensive prowess has been better in years' past -- this is a legendarily bad season for him (or any player) on the offensive side of the ball. He's a top 100 guy this year, been top 50 in prior seasons, so the 'cusp of an All-Star' judgement is pretty accurate. Definitely not a guy you want to max out, but them's the breaks.

[buckeye242424]

24,14,7... And wasn't the best player on the court! We're peaking too soon... Lol

[Hawkfan89]

Josh is just hard headed

[buckeye242424]

Josh is just hard headed

I don't get it.. It's the same shit. He took bad shots at bad times. Actually, he did both. Par for the course...But is anyone gonna argue we beat Utah at home without him? I hope not.

[DrReality]

Stubbornness short circuits intelligence=dumb

[Wretch]

I don't get it.. It's the same shit. He took bad shots at bad times. Actually, he did both. Par for the course...

But is anyone gonna argue we beat Utah at home without him? I hope not.

I'm gonna try to simplify this for you and the others who are promoting the sitcom "Everybody Hates Josh"

On any given night, any team can beat any team. We've beaten good and great teams without Horford, Smoove, and back in the day without Joe. Yes, we could beat Utah without Josh just like we did on these 3 occasions:

http://espn.go.com/nba/boxscore?id=400277760

http://espn.go.com/nba/boxscore?id=400278158

http://espn.go.com/nba/boxscore?id=400278292

That's 3 for 3 without him. BUT I AM NOT ARGUING THAT WE ARE A BETTER TEAM WITHOUT HIM. Let that sink in.

***I AM NOT ARGUING THAT WE ARE A BETTER TEAM WITHOUT HIM***

Nor am I saying that all of his jumpers negate all of his contributions EVERY single night. Very often they can and do however. You would know that if you pay attention to the key possessions in the game that start and end runs - which is where basketball is won and lost.

Please don't make the "hater" assumption and bring this conversation back around to that point. Very few people actually believe that Josh is a "bad" player. Very few people actually "hate" on him (criticism is not hate and not always bad). Most of us have been following him and pulling for hi for TEN YEARS and we DON'T want to see him in another uniform. We just want to see him change and most of us see this change as a big deal and necessary.

Can you see the big deal here?

What benefit are those jump shots in the red box? NONE. Why are we allowing him to do this? It doesn't help, it's never helped, it's not going to help, and honestly...it doesn't make any sense. What these shots do is accomplish some very specific things:

a) They are keeping Josh from maxing his potential

b) They are taking away opportunities for better shooters (look to the blue box).

c) They are keeping better scorers from assuming more responsibility (Josh is currently our "alpha." He leads the team in shot attempts by a WIDE margin).

d) They are wasting possessions

e) They are HELPING the opposition - this is the shot they want us taking and they GIVE it to us.

Please explain to me how ANY of this is ANY good to our team. Please also consider that this is 8 of his 19 shots last night...again, right on par, 40%. That's a lot of damn jumpshots that we are better off without. Something has to be done about it because it is a serious liability when we play better teams.

Or do you think this team is so good that we can just afford to throw away possessions?

[buckeye242424]

I don't want to waste possessions, but every team does throughout the course of a game through turnovers, fouls, shot clock violations, and bad shots. It's part of basketball. We just have a player that chooses to take bad shots...I've accepted it. And why people call him stupid and dumb is played. He's going to keep doing it...and as long as he does everything else and keeps putting up numbers like he is, a playing at the level he is, we're better for it.

[AHF]

I don't want to waste possessions, but every team does throughout the course of a game through turnovers, fouls, shot clock violations, and bad shots. It's part of basketball. We just have a player that chooses to take bad shots...I've accepted it. And why people call him stupid and dumb is played. He's going to keep doing it...and as long as he does everything else and keeps putting up numbers like he is, a playing at the level he is, we're better for it.

The issue is that it really caps his usefulness and makes it a very complicated question how you price him as a FA. Teams that consistently waste possessions don't win NBA titles without overwhelming talent. A team with Josh as its highest paid player is going to waste possessions and will not have overwhelming talent. So what do you do with him? It isn't a simple question for me.