So, after all is said and done, is Marcus a "hit" draft pick

[saltlover]

Put Smart on Warren's Suns or Lavine's Wolves / Bulls, and his winning plays, which are very valuable here, become an alternate version of the empty calories you speak of. 

Smart most likely adds very little to those clubs in terms of wins.  The market confirms this, in the form of 0 bad teams with cap space making a run at him.

For the average awful team picking #6 in the draft, a go-to-scorer is immensely more valuable than a bench defensive specialist.  Smart isn't a 'hit' at #6, but I'd say we got 'solid value.'

I don't think I really agree with this, either.

High volume scorers, especially ones that are not especially efficient and give up a ton on the other end, are overrated commodities.  They don't actually really help you win games.

I think most bad teams would win a lot more games if they had more players like Marcus instead of revolving around players like Lavine. 

But most bad teams don't want to win more games now, they want to assemble valuable assets and win later.

The difference is those teams are hoping that the bad young players like Lavine eventually become good at defense, more efficient at scoring, and provide more overall value.  The seeming upside is what persuades them to go for those guys and give them lots of touches.  They hope the Lavines and Parkers turn into stars or serve as significant parts of trades for stars (like when the Wolves traded Lavine for Butler).

Smart is never going to score 20 points per game, but he was helping the Celts win games pretty much as soon as he entered the league (even though the Celts were bad).

Warren's all time high win shares in his 3rd season (4.2) and Lavine's (3) are right there with Smart's high in year 3 (3.2).  Smart has 3 years with WS at 2.6+, Warren has 2, Lavine 2.

Those guys' teams were awful and that makes it hard to get win shares, but they're right there with Smart.  You're also calling these scorers 'inefficient', but they each shot 55%+ TS% in a 2 year stretch over their 2nd and 3rd years, while Smart's career TS% is 48%. 

His best season was his rookie year, 49.1%.  The win shares are close, and the guys you're calling inefficient look like Steve Nash compared to Smart.  Lavine has considerably more upside than Smart and Smart's highest PER is lower than Warren's lowest, and Warren's top 2 WS seasons beat Smart's best.  There is ample evidence to support my stance in stats and the market.

I also like Smart.  You said 'solid' in your other post- I said that previously on the same page.

Basketball Win Shares do not work like baseball win shares, in that they aren’t based on a team’s record.  For example, as a team, the Celtics had 51 win shares compared to 55 wins.  If you want to use Win Shares as an example, you can’t disqualify them in this manner.

If you want to look at other gestalt stats, Smart dominates both in VORP.  In most of the stats, all three had their worst year since their rookie season last year, so none can make a “trending upward” case.

they are still mostly based on actual team wins though so a guy on a bad team is absolutely hurt in the win shares department by playing on a bad team.

They’re not based on actual team wins at all.  The person who created the metric for basketball found that the total win shares accumulated by the individual players on a team was highly correlated with the total wins a team earned, but that was an after the fact test of his metric to see that it did a good job.  But an individual player’s win shares is team neutral.

[saltlover]

For what it’s worth, CARMELO has TJ Warren at negative value with little upside left (labeled a scrub), LaVine also at negative value, with barely any more growth than Warren (offensive specialist), and Smart at positive value with a little growth left (Up-and-Comer).  This is a downgrade for Smart, who was a Future All-Star last year, thanks to a step back in offense (I don’t know what LaVine and Warren looked like).

https://projects.fivethirtyeight.com/carmelo/tj-warren/
https://projects.fivethirtyeight.com/carmelo/zach-lavine/
https://projects.fivethirtyeight.com/carmelo/marcus-smart/

[mmmmm]

Well, that may be true.  But Win Shares 'favored' Smart in his first couple of seasons relative to his draft classmates.   He was ranked very high by that metric his first two years compared to his draft class.

Was he not being 'accurately assessed' by Win Shares then?

Whether Win Shares are an accurate measure of Smart's value or not, the real problem seems to be that very few metrics of any kind have changed very much in a positive direction for Smart over the 4 years.   Whereas for many of his fellow 2014 draftees, they have started to show large improvements in many measurable areas the last couple of seasons.

I would say no, Win Shares has never really captured his value.

So WS perhaps overrated him then and underrates him now?  That may be true.

At the same time, I agree with you that Smart hasn't really changed much as a player during his time in the league so far.  I'd say he's gotten better as a decision maker passing out of the pick and roll and as a ballhandler he's gotten more controlled I think.  But he hasn't shown any lasting or definite improvement in scoring generally or shooting more specifically.  That is disappointing.

You might think, ... except his turnover rate has increased each of the last two years.  And it was really, really especially bad on pick-&-roll plays this last year.   Horrifically bad, in fact.   Combined with his very poor points-per-play production on P&R plays, there is pretty much zero evidence that he was anything but awful as a ball handler in pick & roll plays these last two years.

Smart desperately needs to show improvement in two key statistics to really turn his value around:  Reduce his missed shots and reduce his turnovers.   The former he can do by either shooting better or taking fewer shots.  The latter he just has to get better in his handle and decision making.   But he needs to improve these two things or he's never going to be more than he is right now.   He's more than good enough in almost every other aspect of the game.  But those two things really drag on him.  It completely nullifies the value of him getting a defensive stop if he turns around and gives the ball back via a missed shot or turnover.

Some other players in the same draft class have shown more improvement, that's true too.  Though I'd say part of that is they started out like most young players do ... rough, not particularly helpful to winning.

Anyways, Smart's upside compared to his peers and whether Smart has shown as much improvement year on year as you'd hope for from a top pick are not really the same discussion as whether the Celts "hit" on the pick, which I interpret as a question of whether the Smart pick has provided value commensurate with the expected yield of a #6 overall pick.

Yeah, as I stated a while ago in this thread, I see Smart as so far being about the 10th most productive player out of this draft which is fine for a #6 pick.   And I totally agree with saltlover's characterization that he was the right choice by Danny given the alternatives at the time (assuming no good trade out deal was workable).   And I do think Smart is a wonderful kid.  I really like what he does for charity work off the court and all his teammates seem to really like and respect him.

But I am very concerned about his lack of developmental progress in the two areas I mentioned.

[mmmmm]

For what it’s worth, CARMELO has TJ Warren at negative value with little upside left (labeled a scrub), LaVine also at negative value, with barely any more growth than Warren (offensive specialist), and Smart at positive value with a little growth left (Up-and-Comer).  This is a downgrade for Smart, who was a Future All-Star last year, thanks to a step back in offense (I don’t know what LaVine and Warren looked like).

https://projects.fivethirtyeight.com/carmelo/tj-warren/
https://projects.fivethirtyeight.com/carmelo/zach-lavine/
https://projects.fivethirtyeight.com/carmelo/marcus-smart/

To be brutally honest, it isn't worth much.  If folks are not feeling good about Win Shares as an accurate measure for Smart's recorded production, they should run far away from CARMELO.  CARMELO every year has missed their projections on Smart by a country mile.

[saltlover]

For what it’s worth, CARMELO has TJ Warren at negative value with little upside left (labeled a scrub), LaVine also at negative value, with barely any more growth than Warren (offensive specialist), and Smart at positive value with a little growth left (Up-and-Comer).  This is a downgrade for Smart, who was a Future All-Star last year, thanks to a step back in offense (I don’t know what LaVine and Warren looked like).

https://projects.fivethirtyeight.com/carmelo/tj-warren/
https://projects.fivethirtyeight.com/carmelo/zach-lavine/
https://projects.fivethirtyeight.com/carmelo/marcus-smart/

To be brutally honest, it isn't worth much.  If folks are not feeling good about Win Shares as an accurate measure for Smart's recorded production, they should run far away from CARMELO.  CARMELO every year has missed their projections on Smart by a country mile.

I’m more using it to show the value it has assigned the past three seasons as another datapoint that it is at best arguable whether one should favor LaVine or Warren.  In general CARMELO has tested well with projections.  Smart took an unexpected (from most everyone) step backwards on offense last season, so sure, it missed that part of the projection.  One could argue that it’s projection offensively is again too rosy, but the point is that Smart has consistently had significantly positive value, whereas LaVine and Warren are negative or near zero each of the past three seasons.  CARMELO’s WARP is a combo of BPM and RPM, in an effort to mute the effects of one of those system being biased towards a certain player or class of players.

[Moranis]

Put Smart on Warren's Suns or Lavine's Wolves / Bulls, and his winning plays, which are very valuable here, become an alternate version of the empty calories you speak of. 

Smart most likely adds very little to those clubs in terms of wins.  The market confirms this, in the form of 0 bad teams with cap space making a run at him.

For the average awful team picking #6 in the draft, a go-to-scorer is immensely more valuable than a bench defensive specialist.  Smart isn't a 'hit' at #6, but I'd say we got 'solid value.'

I don't think I really agree with this, either.

High volume scorers, especially ones that are not especially efficient and give up a ton on the other end, are overrated commodities.  They don't actually really help you win games.

I think most bad teams would win a lot more games if they had more players like Marcus instead of revolving around players like Lavine. 

But most bad teams don't want to win more games now, they want to assemble valuable assets and win later.

The difference is those teams are hoping that the bad young players like Lavine eventually become good at defense, more efficient at scoring, and provide more overall value.  The seeming upside is what persuades them to go for those guys and give them lots of touches.  They hope the Lavines and Parkers turn into stars or serve as significant parts of trades for stars (like when the Wolves traded Lavine for Butler).

Smart is never going to score 20 points per game, but he was helping the Celts win games pretty much as soon as he entered the league (even though the Celts were bad).

Warren's all time high win shares in his 3rd season (4.2) and Lavine's (3) are right there with Smart's high in year 3 (3.2).  Smart has 3 years with WS at 2.6+, Warren has 2, Lavine 2.

Those guys' teams were awful and that makes it hard to get win shares, but they're right there with Smart.  You're also calling these scorers 'inefficient', but they each shot 55%+ TS% in a 2 year stretch over their 2nd and 3rd years, while Smart's career TS% is 48%. 

His best season was his rookie year, 49.1%.  The win shares are close, and the guys you're calling inefficient look like Steve Nash compared to Smart.  Lavine has considerably more upside than Smart and Smart's highest PER is lower than Warren's lowest, and Warren's top 2 WS seasons beat Smart's best.  There is ample evidence to support my stance in stats and the market.

I also like Smart.  You said 'solid' in your other post- I said that previously on the same page.

Basketball Win Shares do not work like baseball win shares, in that they aren’t based on a team’s record.  For example, as a team, the Celtics had 51 win shares compared to 55 wins.  If you want to use Win Shares as an example, you can’t disqualify them in this manner.

If you want to look at other gestalt stats, Smart dominates both in VORP.  In most of the stats, all three had their worst year since their rookie season last year, so none can make a “trending upward” case.

they are still mostly based on actual team wins though so a guy on a bad team is absolutely hurt in the win shares department by playing on a bad team.

They’re not based on actual team wins at all.  The person who created the metric for basketball found that the total win shares accumulated by the individual players on a team was highly correlated with the total wins a team earned, but that was an after the fact test of his metric to see that it did a good job.  But an individual player’s win shares is team neutral.

and yet players with very similar stats have vastly different win shares if one is on a good team and one is on a bad team.

BTW, basketball-reference flat out states they are generally tied to team wins

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

[saltlover]

Put Smart on Warren's Suns or Lavine's Wolves / Bulls, and his winning plays, which are very valuable here, become an alternate version of the empty calories you speak of. 

Smart most likely adds very little to those clubs in terms of wins.  The market confirms this, in the form of 0 bad teams with cap space making a run at him.

For the average awful team picking #6 in the draft, a go-to-scorer is immensely more valuable than a bench defensive specialist.  Smart isn't a 'hit' at #6, but I'd say we got 'solid value.'

I don't think I really agree with this, either.

High volume scorers, especially ones that are not especially efficient and give up a ton on the other end, are overrated commodities.  They don't actually really help you win games.

I think most bad teams would win a lot more games if they had more players like Marcus instead of revolving around players like Lavine. 

But most bad teams don't want to win more games now, they want to assemble valuable assets and win later.

The difference is those teams are hoping that the bad young players like Lavine eventually become good at defense, more efficient at scoring, and provide more overall value.  The seeming upside is what persuades them to go for those guys and give them lots of touches.  They hope the Lavines and Parkers turn into stars or serve as significant parts of trades for stars (like when the Wolves traded Lavine for Butler).

Smart is never going to score 20 points per game, but he was helping the Celts win games pretty much as soon as he entered the league (even though the Celts were bad).

Warren's all time high win shares in his 3rd season (4.2) and Lavine's (3) are right there with Smart's high in year 3 (3.2).  Smart has 3 years with WS at 2.6+, Warren has 2, Lavine 2.

Those guys' teams were awful and that makes it hard to get win shares, but they're right there with Smart.  You're also calling these scorers 'inefficient', but they each shot 55%+ TS% in a 2 year stretch over their 2nd and 3rd years, while Smart's career TS% is 48%. 

His best season was his rookie year, 49.1%.  The win shares are close, and the guys you're calling inefficient look like Steve Nash compared to Smart.  Lavine has considerably more upside than Smart and Smart's highest PER is lower than Warren's lowest, and Warren's top 2 WS seasons beat Smart's best.  There is ample evidence to support my stance in stats and the market.

I also like Smart.  You said 'solid' in your other post- I said that previously on the same page.

Basketball Win Shares do not work like baseball win shares, in that they aren’t based on a team’s record.  For example, as a team, the Celtics had 51 win shares compared to 55 wins.  If you want to use Win Shares as an example, you can’t disqualify them in this manner.

If you want to look at other gestalt stats, Smart dominates both in VORP.  In most of the stats, all three had their worst year since their rookie season last year, so none can make a “trending upward” case.

they are still mostly based on actual team wins though so a guy on a bad team is absolutely hurt in the win shares department by playing on a bad team.

They’re not based on actual team wins at all.  The person who created the metric for basketball found that the total win shares accumulated by the individual players on a team was highly correlated with the total wins a team earned, but that was an after the fact test of his metric to see that it did a good job.  But an individual player’s win shares is team neutral.

and yet players with very similar stats have vastly different win shares if one is on a good team and one is on a bad team.

BTW, basketball-reference flat out states they are generally tied to team wins

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

From your own link:

Because this metric is designed to estimate a player's contribution in terms of wins, it makes sense to see if the sum of player Win Shares for a particular team closely matches the team win total. For the 2008-09 Cavaliers the sum of player Win Shares is 67.9, while the team win total is 66, an error of 66 - 67.9 = -1.9 wins. For the 1964-65 Royals the sum of player Win Shares is 43.5, while the team total is 48, an error of 48 - 43.5 = 4.5 wins. These errors are actually close to the "typical" error; looking at all NBA teams since the 1962-63 season (the last season we have complete player splits), the average absolute error is 2.74 wins and the root mean squared error is 3.41 wins.

They don’t start with a team’s wins and then assign players portions of those wins, which is what Bill James does for baseball.  They are not “tied” in any way to team wins.  It’s just been tested to show that you get very close to team wins, which means the statistic is doing its job.

I can’t explain to you exactly how players with similar stats on will have different win shares, but you’re welcome to buy a book that does. 

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details.

Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details).

[KungPoweChicken]

Smart is a role player. At his draft position, my expectation is a starter. I can't say he was a "hit." Not quite a bust either, though.

[Green-18]

Smart is a role player. At his draft position, my expectation is a starter. I can't say he was a "hit." Not quite a bust either, though.

In a vacuum I can't disagree with you at all.  However, how do we view things if Smart is a clear difference maker in a Finals match up with Golden State at some point over the next few years?  I know I'm projecting quite a bit, but I'm curious how people would feel if we win banner 18 with Smart playing a key role in the 4th quarter of some of those wins.

I care more about his fit with our team as opposed to comparisons that are devoid of context.  Smart's contract should be evaluated strictly on his ability to contribute to big time playoff wins.  The grit, toughness, and intangibles need to be on full display in the most important games.




   

[Big333223]

Marcus is my favorite player in the league so I'm bringing that bias with me, but I think he has DPOY potential. If you get a DPOY at #6, I think you have to be happy with that.

[PhoSita]

Smart desperately needs to show improvement in two key statistics to really turn his value around:  Reduce his missed shots and reduce his turnovers.   The former he can do by either shooting better or taking fewer shots.  The latter he just has to get better in his handle and decision making.   But he needs to improve these two things or he's never going to be more than he is right now.   He's more than good enough in almost every other aspect of the game.  But those two things really drag on him.  It completely nullifies the value of him getting a defensive stop if he turns around and gives the ball back via a missed shot or turnover.

....


But I am very concerned about his lack of developmental progress in the two areas I mentioned.

To be perfectly honest, I don't really understand the level of concern or hand-wringing here.


I would love it if Smart would become a better shooter.  I would love it if he became a better pure ball-handler.  I'm mystified by how poor a finisher he is given how good he was in college.


All of that said, in my mind Smart has been a valuable player pretty much from day one.  He has consistently helped the team win games, he's a major part of the team culture, and a core piece of their defensive game plan (the defense that was the best in the league last year). 

To me, it would be great if Smart showed improvement in some other areas, especially those that show up more in the box score.

But even if he continues to be basically the guy he's been, I think he'll be worth his contract.


The main cause for concern I would have for Smart and his value moving forward is that his style of play will lead to more frequent injuries.  I also worry what happens if he loses a step and isn't quite as agile moving his feet laterally, getting around picks, etc.



Your post makes it sound like Smart isn't actually a particularly valuable player right now and he needs to show major improvement in order to become one.  I don't agree with that perspective.

I genuinely think Smart is in the 99th percentile or whatever in terms of players whose value is not well captured by the stats we most commonly use to measure player value.

That's not to say he doesn't have games where he hurts the team with missed shots or turnovers, but I think his impact is decidedly in the positive category, much more so than your typical ("replacement level") combo guard.

[bellerephon]

Smart is a role player. At his draft position, my expectation is a starter. I can't say he was a "hit." Not quite a bust either, though.

As has been said, it depends on how you define hit. A guy who plays starter minutes, plays in big moments and is usually in there at the end of games, and a guy who really impacts winning is a hit to me at six. I also disagree with the idea that he's not quite a bust, that suggests he is close to being a bust. I totally disagree with that. Smart is the first guy off the bench and a finisher on a team that could contend for a trip to the finals, that is very far from being a bust.

[Moranis]

Put Smart on Warren's Suns or Lavine's Wolves / Bulls, and his winning plays, which are very valuable here, become an alternate version of the empty calories you speak of. 

Smart most likely adds very little to those clubs in terms of wins.  The market confirms this, in the form of 0 bad teams with cap space making a run at him.

For the average awful team picking #6 in the draft, a go-to-scorer is immensely more valuable than a bench defensive specialist.  Smart isn't a 'hit' at #6, but I'd say we got 'solid value.'

I don't think I really agree with this, either.

High volume scorers, especially ones that are not especially efficient and give up a ton on the other end, are overrated commodities.  They don't actually really help you win games.

I think most bad teams would win a lot more games if they had more players like Marcus instead of revolving around players like Lavine. 

But most bad teams don't want to win more games now, they want to assemble valuable assets and win later.

The difference is those teams are hoping that the bad young players like Lavine eventually become good at defense, more efficient at scoring, and provide more overall value.  The seeming upside is what persuades them to go for those guys and give them lots of touches.  They hope the Lavines and Parkers turn into stars or serve as significant parts of trades for stars (like when the Wolves traded Lavine for Butler).

Smart is never going to score 20 points per game, but he was helping the Celts win games pretty much as soon as he entered the league (even though the Celts were bad).

Warren's all time high win shares in his 3rd season (4.2) and Lavine's (3) are right there with Smart's high in year 3 (3.2).  Smart has 3 years with WS at 2.6+, Warren has 2, Lavine 2.

Those guys' teams were awful and that makes it hard to get win shares, but they're right there with Smart.  You're also calling these scorers 'inefficient', but they each shot 55%+ TS% in a 2 year stretch over their 2nd and 3rd years, while Smart's career TS% is 48%. 

His best season was his rookie year, 49.1%.  The win shares are close, and the guys you're calling inefficient look like Steve Nash compared to Smart.  Lavine has considerably more upside than Smart and Smart's highest PER is lower than Warren's lowest, and Warren's top 2 WS seasons beat Smart's best.  There is ample evidence to support my stance in stats and the market.

I also like Smart.  You said 'solid' in your other post- I said that previously on the same page.

Basketball Win Shares do not work like baseball win shares, in that they aren’t based on a team’s record.  For example, as a team, the Celtics had 51 win shares compared to 55 wins.  If you want to use Win Shares as an example, you can’t disqualify them in this manner.

If you want to look at other gestalt stats, Smart dominates both in VORP.  In most of the stats, all three had their worst year since their rookie season last year, so none can make a “trending upward” case.

they are still mostly based on actual team wins though so a guy on a bad team is absolutely hurt in the win shares department by playing on a bad team.

They’re not based on actual team wins at all.  The person who created the metric for basketball found that the total win shares accumulated by the individual players on a team was highly correlated with the total wins a team earned, but that was an after the fact test of his metric to see that it did a good job.  But an individual player’s win shares is team neutral.

and yet players with very similar stats have vastly different win shares if one is on a good team and one is on a bad team.

BTW, basketball-reference flat out states they are generally tied to team wins

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

From your own link:

Because this metric is designed to estimate a player's contribution in terms of wins, it makes sense to see if the sum of player Win Shares for a particular team closely matches the team win total. For the 2008-09 Cavaliers the sum of player Win Shares is 67.9, while the team win total is 66, an error of 66 - 67.9 = -1.9 wins. For the 1964-65 Royals the sum of player Win Shares is 43.5, while the team total is 48, an error of 48 - 43.5 = 4.5 wins. These errors are actually close to the "typical" error; looking at all NBA teams since the 1962-63 season (the last season we have complete player splits), the average absolute error is 2.74 wins and the root mean squared error is 3.41 wins.

They don’t start with a team’s wins and then assign players portions of those wins, which is what Bill James does for baseball.  They are not “tied” in any way to team wins.  It’s just been tested to show that you get very close to team wins, which means the statistic is doing its job.

I can’t explain to you exactly how players with similar stats on will have different win shares, but you’re welcome to buy a book that does. 

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details.

Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details).

I never said that you started from teams wins and worked back.  They absolutely correlate to actual team wins though.  That is why a player like Lebron James can have worse stats in 08-09 and yet have significantly more win shares then he did in 07-08 when he had better stats.  Same player, similar stats, the only difference is one year the Cavs won 45 games and the next they won 66 games.  Team wins matter or else you wouldn't see the correlation that you do, even from season to season of the same player who has similar stats.   

[mmmmm]

I never said that you started from teams wins and worked back.  They absolutely correlate to actual team wins though.  That is why a player like Lebron James can have worse stats in 08-09 and yet have significantly more win shares then he did in 07-08 when he had better stats.  Same player, similar stats, the only difference is one year the Cavs won 45 games and the next they won 66 games.  Team wins matter or else you wouldn't see the correlation that you do, even from season to season of the same player who has similar stats.

In 08-09, James had better stats than 07-08.  He scored at an elite efficiency of .591 TS% compared to (very good but not elite) .568 TS in 07-08.  He played a tiny bit more minutes.  He had a higher DRB% and a higher overall TRB%.  He also had a slightly higher AST% and BLK%, the same STL% and USG% to go along with a slightly lower TOV%.    He scored, rebounded, assisted, blocked at higher per-36 rates while turning the ball over and committing fouls at lower per-36 rates.

In short, pretty much every single rate or efficiency stat was better in 08-09 and he played a handful more minutes.  Thus, more Win Shares.

Individual Win Shares do correlate with team wins (because individually they should add up to close to the team total) and there IS some cross-influence because having better shooting / passing teammates will tend to help your own passing / shooting numbers.

[saltlover]

Put Smart on Warren's Suns or Lavine's Wolves / Bulls, and his winning plays, which are very valuable here, become an alternate version of the empty calories you speak of. 

Smart most likely adds very little to those clubs in terms of wins.  The market confirms this, in the form of 0 bad teams with cap space making a run at him.

For the average awful team picking #6 in the draft, a go-to-scorer is immensely more valuable than a bench defensive specialist.  Smart isn't a 'hit' at #6, but I'd say we got 'solid value.'

I don't think I really agree with this, either.

High volume scorers, especially ones that are not especially efficient and give up a ton on the other end, are overrated commodities.  They don't actually really help you win games.

I think most bad teams would win a lot more games if they had more players like Marcus instead of revolving around players like Lavine. 

But most bad teams don't want to win more games now, they want to assemble valuable assets and win later.

The difference is those teams are hoping that the bad young players like Lavine eventually become good at defense, more efficient at scoring, and provide more overall value.  The seeming upside is what persuades them to go for those guys and give them lots of touches.  They hope the Lavines and Parkers turn into stars or serve as significant parts of trades for stars (like when the Wolves traded Lavine for Butler).

Smart is never going to score 20 points per game, but he was helping the Celts win games pretty much as soon as he entered the league (even though the Celts were bad).

Warren's all time high win shares in his 3rd season (4.2) and Lavine's (3) are right there with Smart's high in year 3 (3.2).  Smart has 3 years with WS at 2.6+, Warren has 2, Lavine 2.

Those guys' teams were awful and that makes it hard to get win shares, but they're right there with Smart.  You're also calling these scorers 'inefficient', but they each shot 55%+ TS% in a 2 year stretch over their 2nd and 3rd years, while Smart's career TS% is 48%. 

His best season was his rookie year, 49.1%.  The win shares are close, and the guys you're calling inefficient look like Steve Nash compared to Smart.  Lavine has considerably more upside than Smart and Smart's highest PER is lower than Warren's lowest, and Warren's top 2 WS seasons beat Smart's best.  There is ample evidence to support my stance in stats and the market.

I also like Smart.  You said 'solid' in your other post- I said that previously on the same page.

Basketball Win Shares do not work like baseball win shares, in that they aren’t based on a team’s record.  For example, as a team, the Celtics had 51 win shares compared to 55 wins.  If you want to use Win Shares as an example, you can’t disqualify them in this manner.

If you want to look at other gestalt stats, Smart dominates both in VORP.  In most of the stats, all three had their worst year since their rookie season last year, so none can make a “trending upward” case.

they are still mostly based on actual team wins though so a guy on a bad team is absolutely hurt in the win shares department by playing on a bad team.

They’re not based on actual team wins at all.  The person who created the metric for basketball found that the total win shares accumulated by the individual players on a team was highly correlated with the total wins a team earned, but that was an after the fact test of his metric to see that it did a good job.  But an individual player’s win shares is team neutral.

and yet players with very similar stats have vastly different win shares if one is on a good team and one is on a bad team.

BTW, basketball-reference flat out states they are generally tied to team wins

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

From your own link:

Because this metric is designed to estimate a player's contribution in terms of wins, it makes sense to see if the sum of player Win Shares for a particular team closely matches the team win total. For the 2008-09 Cavaliers the sum of player Win Shares is 67.9, while the team win total is 66, an error of 66 - 67.9 = -1.9 wins. For the 1964-65 Royals the sum of player Win Shares is 43.5, while the team total is 48, an error of 48 - 43.5 = 4.5 wins. These errors are actually close to the "typical" error; looking at all NBA teams since the 1962-63 season (the last season we have complete player splits), the average absolute error is 2.74 wins and the root mean squared error is 3.41 wins.

They don’t start with a team’s wins and then assign players portions of those wins, which is what Bill James does for baseball.  They are not “tied” in any way to team wins.  It’s just been tested to show that you get very close to team wins, which means the statistic is doing its job.

I can’t explain to you exactly how players with similar stats on will have different win shares, but you’re welcome to buy a book that does. 

Offensive Win Shares are credited to players based on Dean Oliver's points produced and offensive possessions. The formulas are quite detailed, so I would point you to Oliver's book Basketball on Paper for complete details.

Crediting Defensive Win Shares to players is based on Dean Oliver's Defensive Rating. Defensive Rating is an estimate of the player's points allowed per 100 defensive possessions (please see Oliver's book for further details).

I never said that you started from teams wins and worked back.  They absolutely correlate to actual team wins though.  That is why a player like Lebron James can have worse stats in 08-09 and yet have significantly more win shares then he did in 07-08 when he had better stats.  Same player, similar stats, the only difference is one year the Cavs won 45 games and the next they won 66 games.  Team wins matter or else you wouldn't see the correlation that you do, even from season to season of the same player who has similar stats.

Honestly, Moranis, I’m not sure what you’re saying at this point.  “Based on team wins”, which you said, means that wins are a starting point for the calculation.  Otherwise they are not based on them.  Plain and simple.

As for this point:

That is why a player like Lebron James can have worse stats in 08-09 and yet have significantly more win shares then he did in 07-08 when he had better stats.  Same player, similar stats, the only difference is one year the Cavs won 45 games and the next they won 66 games.

In the calculations for win shares, team wins are absent.  It’s a function of “points produced”, a team’s relative pace, average points per possession league wide, and marginal points per win league wide (for offensive win shares). As I said before, go give Dean Oliver some money and buy his book where he comes up with his “points produced” metric, and maybe somewhere in there you’ll see a formula where team success is a factor.  I frankly doubt you’ll find that tho.  You keep harping that LeBron had similar stats, but without knowing how points produced is calculated, you can’t say what different stats mattered, not to mention that changes in team pace, average league offensive efficiency, and the marginal value of a point can fluctuate from year to year.  Saying “the only difference is win totals” is hogwash, and completely belied by the formula at the very link you presented, in which win totals are 100% absent.