2014 NBA Draft Thread (No Spoilers You Clowns)

[Kliq]

For Embiid to be the top pick, he is going to have to have some incredible game with the nation watching. If he has a game in the tournament where he goes like, 30-20 with 8 blocks or something. To pick him over Wiggins or Parker or Smart, you have to be REALLY sure that he is going to be the next great big man.
 
I still think Wiggins is the likely number one pick. I think GMs are not that worried about his slow (for what we expected) start.

 

[Cellar-Door]

So from what I've gathered:
 
Embiid
Parker
Smart
Wiggins (all upside until he learns how to shoot?)
Randle (tweener concerns?)
Exum (only plays against high schoolers? broken jumper?)
Hezonja (attitude problems?)
Hood (unathletic?)
 
Objections?

Gordon should be on that list no lower than after exum. Personally I'd have him before Randle.

 

[bowiac]

Is Embiid really that a controversial a pick? DraftExpress ranks him 4th ahead of Smart, Exum or Gordon already. I mean, we just saw Anthony Bennett go first overall. Teams seem to draft what their scouts tell them.

 

[The X Man Cometh]

For Embiid to be the top pick, he is going to have to have some incredible game with the nation watching. If he has a game in the tournament where he goes like, 30-20 with 8 blocks or something. To pick him over Wiggins or Parker or Smart, you have to be REALLY sure that he is going to be the next great big man.
 
I still think Wiggins is the likely number one pick. I think GMs are not that worried about his slow (for what we expected) start.

 
Respectfully disagree. Its not like the other guys are slam dunk prospects.
 
Wiggins has absurd physical talent, but from the games that have been on TV, he's a major work in progress offensively. His jumper isn't consistent enough to justify taking a lot of shots, his handle isn't tight enough to justify high usage. If Wiggins/Embiid is a Jordan/Bowie or a Durant/Oden situation, we have zero reason to expect it at this stage.
 
I mean, people keep talking about how a GM wouldn't have the stones to take Embiid over Wiggins. But isn't it Embiid who's the safe pick and not Wiggins? With Embiid you at least have a shotblocker and board cleaning defensive presence out of the box, and anything else is gravy. Wiggins can't just settle into being an excellent wing defender and not be a disappointment.

 

[The X Man Cometh]

Gordon should be on that list no lower than after exum. Personally I'd have him before Randle.

 
Gordon didn't impress me in the two Arizona games I saw. I get that the draw with him is that he's kind of an elite glue guy moreso than a conventional "star", but unless he has the ability to improve into more of a focal point offensively I wouldn't draft him very high.
 
That said, from what we've seen of Brad Stevens and what we know about Rondo, I think Gordon would fit well on the Celtics. If he were to slide due to a lack of neatly defined role, I'd love to see the Celtics scoop him up. Have him defend 3 positions, pass the ball, and run around like his hair is on fire.

 

[Brickowski]

I've been looking at film of Embiid and he's going to be the number 1 pick unless he injures himself.  He's been playing ball a lot longer than advertised: he was reportedly trained by the coach of Cameroon's national team before coming to the states at age 16.  It shows.  This kid does a lot more than just dunk the ball and block shots.

 

[SpacemanzGerbil]

I've been looking at film of Embiid and he's going to be the number 1 pick unless he injures himself.  

 If you haven't, you need to watch the Kansas-Florida game. He was horrendous.

 

[Brickowski]

No I did not see that game.  He'll probably have more bad games; he's young.  But from what I have seen, I would not take Wiggins, Parker or Randle ahead of him.

 

[Cellar-Door]

 
Gordon didn't impress me in the two Arizona games I saw. I get that the draw with him is that he's kind of an elite glue guy moreso than a conventional "star", but unless he has the ability to improve into more of a focal point offensively I wouldn't draft him very high.
 
That said, from what we've seen of Brad Stevens and what we know about Rondo, I think Gordon would fit well on the Celtics. If he were to slide due to a lack of neatly defined role, I'd love to see the Celtics scoop him up. Have him defend 3 positions, pass the ball, and run around like his hair is on fire.

I think he's developing nicely in part because he doesn't need to carry his offense. I also think he's the best defender of his peer group. He absolutely shut down Robinson and made Parker look like Rudy Gay on a bad night. He needs to be better on pick and rolls and midrange. But he shoots the three pretty well.

 

[bowiac]

I agree with that BSF, but my understanding is that Embiid is thought to be a top upside guy as well. He's an "upside, but raw" guy who has played better than "raw" so far. I think that's the source of the excitement. 

 

[Brickowski]

Embiid is potentially the one true franchise player in the draft (again, with the injury caveat).  Some of the other guys are going to be very good NBA players, but not franchise changers.  There are times where Embiid looks bad, but others where he looks like a young Kevin Garnett.  And even if he becomes the next Serge Ibaka (as opposed to KG), it's still not a blown pick.
 
I suppose my evaluation is tainted by the fact that I was expecting the next Fab Melo, but-- as Bowiac suggests-- I've seen a much, much more polished player.   

 

[Julius.R]

So, Philadelphia is now a bottom 3 team and also owns New Orleans' 2014 pick as well. With a young core of MCW, Noel, Evan Turner, somebody in the Wiggins/Parker/Embiid/Smart/Gordon group, Hawes, Wroten, and another late lottery pick, they are going to have a great foundation for the future. You have to wonder what their plans are for Young, Hawes, and Turner are. Could be an opportunity for a contender to add a useful piece.

Might be interesting to see if the 76ers and Suns work out a trade. One of Phoenix's 1sts for one of the three, Young might be the best fit. Suns get a 'Young' player to help them compete and make the playoffs, while the 76ers add to their stock of draft picks.

 

[The X Man Cometh]

So, Philadelphia is now a bottom 3 team and also owns New Orleans' 2014 pick as well. With a young core of MCW, Noel, Evan Turner, somebody in the Wiggins/Parker/Embiid/Smart/Gordon group, Hawes, Wroten, and another late lottery pick, they are going to have a great foundation for the future. You have to wonder what their plans are for Young, Hawes, and Turner are. Could be an opportunity for a contender to add a useful piece.

 
If I'm Philadelphia, I'm looking into trading Evan Turner. Too ball-dominant of a player to coexist with a star player. Not worth the money he's going to want to be extended at.
 
As for Wroten... that guy is an awful chucker. Lots of players can score when given a blank card to do it. I'd include 2nd year PF Arnett Moultrie (blew out his knee in the offseason) as a building block before him.

 

[Brickowski]

Why would the Suns want Turner?  They already have three guards wth eFG% in the mid 50's.  Turner was 45.6% in his best year, and he's currently down at 43%.
 
The Bledsoe Dragic combo has worked really well for the Suns, who certainly don't need Turner to make the playoffs. If I were McDonough, the Sixer I'd be targeting is Hawes.  Maybe I would offer Frye and a protected first round pick for him.

 

[The X Man Cometh]

I agree that Turner could be a likely trade candidate, especially since they'll have a pretty good chance of landing either Wiggins or Parker. Wroten is definitely a chucker but he's been a semi-valuable piece for them this year with MCW out. And since they got him for basically nothing, I consider him to be an asset to their bench. His value still probably isn't very high but he's not a guy I'd give away for free anymore like the Grizzlies did.

 
Fair enough. Don't sleep on Arnett Moultrie. He's a very good rebounder who has the tools to defend pretty much any power forward and developing offensive game. Getting hurt this year probably stunted his progress, but he could be a very starter in the NBA.
 
If I'm Philly, I'm not trading Thad Young or Spencer Hawes unless the return is really hard to pass up. Neither is exactly an all-defense type of player but they play hard and are active, and shoot the lights out of the ball. Low maintenance and productive. Perfect fits for a young team. Dragging bigs out of the paint to shadow them at the 3-pt line is going to be great for Wiggins/Parker/Exum/Smart/MCW/whomever as well.

 

[beezer]

One thing I like about Embiid is the fact that he can finish around the rim in traffic.  I feel like it has been a long time since the C's had someone who could grab an offensive rebound and bring the ball right back up for a dunk.  I know I'm completely overrating this need, but I've just at times become frustrated with the constant parade of undersized centers playing below the rim for the Celts.

 

[EL Jeffe]

Kyle Anderson doesn't get nearly enough love from the draft gurus. He's like a longer, more athletic Luke Walton. Maybe more of a Nicholas Batum type? A skinny Boris Diaw? Great passer with a very high basketball IQ. He's not explosive enough to ever be much of a scorer at the next level, so he lacks the proverbial upside label. The draft sites don't like him but I see definite first round value and a long, solid career from him. He strikes me as the type of player the Spurs typically steal.

 

[radsoxfan]

Kyle Anderson doesn't get nearly enough love from the draft gurus. He's like a longer, more athletic Luke Walton. Maybe more of a Nicholas Batum type? A skinny Boris Diaw? Great passer with a very high basketball IQ. He's not explosive enough to ever be much of a scorer at the next level, so he lacks the proverbial upside label. The draft sites don't like him but I see definite first round value and a long, solid career from him. He strikes me as the type of player the Spurs typically steal.

 
Anderson is an interesting prospect, not sure I've seen anyone quite like him.
 
But he is not "more athletic" than Luke Walton, or anyone else.  Probably one of the least athletic NBA prospects of all time.  When he tries to jump, he actually somehow moves closer to the ground. 
 
His shooting numbers have improved this year, but his shot still looks flat and not very good so I'm skeptical in that respect.  I'm honestly not sure he can guard anyone in the NBA to even give himself a chance to show his offensive skill.

 

[Cellar-Door]

Playing a bad Southern team Arizona decided to get Gordon some work.
The Good- 21 points on 11 shots is spectacular, no turnovers etc.
 
The Bad- only 4 REB is a bit disappointing
 
The Ugly 4-10 on FTs. UGH

 

[oumbi]

this is from celticsblog.com by a poster named celtis18. It looks at the past 30 years of drafts and shows win-shares by draft slot. Interesting information really and I wanted to share it. Please forgive the copy and paste.

http://forums.celticsblog.com/index.php?topic=68700.0#msg1592337

I set about trying to find the answer to this question.  So, I went to BasketballReference and did a little research.

I looked at the win share totals for all players selected at given spots in the 20 drafts between 1991 and 2010.  I then added up the total win shares for each draft spot from 1 to 28 to see which spots have yielded the most successful players over that time period.

Here are the results, listed in order of most successful draft spot to least from picks 1 to 28:

Rank           Draft position         total win shares

1                       1st                                  1,354.1

2                       5th                                  1,056.8

3                       3rd                                  1,055.1

4                       4th                                  1,030.8

5                       2nd                                    987.8

6                       9th                                     837.9

7                      10th                                   815.4

8                      13th                                   658.4

9                       7th                                     596.5

10                   14th                                   523.8

11                     8th                                     499.6

12                   24th                                   484.1

13                   15th                                   460.2

14                   21st                                   457.0

15                   17th                                  418.5

16                     6th                                    413.2

17                   11th                                  368.1

18                   23rd                                  353.5

19                   18th                                   343.6

20                   16th                                   334.2

21                   26th                                  315.5

22                   12th                                  296.9

23                   20th                                  293.3

24                   25th                                  258.6

25                   19th                                  246.2

26                   27th                                   203.3

27                   28th                                  203,1

28                   22nd                                  202.1

 

[snowmanny]

Well the oddity there is that while 1's did best (not surprising) and there wasn't much difference between the 2,3,4 and 5 picks (also not all that surprising) the 6th picks generally bombed.

 

[Grin&MartyBarret]

Isn't the pretty obvious flaw in that methodology that it emphasizes the value of veterans? For instance, Kevin Garnett is 8th all time in WS, and has produced more Win Shares than any player drafted since '91. So he is far more valuable in something like this than Kevin Durant, who has played 11 fewer seasons. Guys like Antawn Jameson and Jason Terry, because of longevity alone, are weighted far more than somebody like Anthony Davis, Kevin Love, or Steph Curry.

 

[nighthob]

Isn't the pretty obvious flaw in that methodology that it emphasizes the value of veterans? For instance, Kevin Garnett is 8th all time in WS, and has produced more Win Shares than any player drafted since '91. So he is far more valuable in something like this than Kevin Durant, who has played 11 fewer seasons. Guys like Antawn Jameson and Jason Terry, because of longevity alone, are weighted far more than somebody like Anthony Davis, Kevin Love, or Steph Curry.

 
Yeah, it makes far more sense, but takes far more work, to use PER to roughly measure the productivity of the various draft spots.

 

[bowiac]

 
Yeah, it makes far more sense, but takes far more work, to use PER to roughly measure the productivity of the various draft spots.

It never makes sense to use PER. Win shares is flawed, but PER is just a fundamentally garbage stat. But it also wouldn't take much more work. If you want a rate stat however, I would suggest either win shares/year, or WS/48 minutes.
 
With respect to overvaluing veterans, it's valuing career contributions. I don't know that that's a flaw. It's not entirely crazy to say that Jason Terry's 1100 game career has been more valuable than Anthony Davis's 81 game career. That effect will also balance out, unless there's something about later picks that makes them more likely to stick around and have long journeyman-style careers. 

 

[nighthob]

It never makes sense to use PER. Win shares is flawed, but PER is just a fundamentally garbage stat. But it also wouldn't take much more work. If you want a rate stat however, I would suggest either win shares/year, or WS/48 minutes.

 
All of the magic formulae are garbage. Win Shares and w/48 included.

 

[bowiac]

All of the magic formulae are garbage. Win Shares and w/48 included.

None of them capture everything, but I'd argue win shares does a surprisingly good job. It misses on some guys with unusual skillsets, but has real value overall.
 
PER is just fundamentally broken in that it rewards chucking. Every additional 2 point shot you take at over a 30.4% rate, or every additional 3 point shot you take at over a 21.4% rate will increase your PER. That's why JR Smith looks like a good player by PER. Regardless of the inherent flaws in trying to boil down as complex a game as basketball to an all-in-one metric, the math in PER is just wrong.

 

[nighthob]

No, they all do suck. Basketball doesn't really break down in that fashion. On the micro-level the stats are solid, but the all-in-ones all reflect the subjective preferences of the people that create them. Also PER doesn't "reward chucking". It attempts to deal with the reality that, strictly speaking, points scored has the greatest correlation to winning and people that produce them have some value (if PER really "rewarded chucking" Rudy Gay would still be suiting up for Hollinger's team). But both win shares and w/48 reward guys for not shooting at all, even though the inability to score actively hurts a team's ability to win. So, no, they aren't "surprisingly good" unless by "surprisingly good" you mean "they're 3% better than my old Magic 8 Ball".

 

[The X Man Cometh]

I'd rather dispense of the stats and just watch the games. Baseball it is not.

As for the "chucking" argument, efficiency has gone from being an underrated trait to an overrated one in my opinion

 

[ishmael]

 
All of the magic formulae are garbage. Win Shares and w/48 included.

For cost/benefit type of analysis, WS/48 is actually a reasonable proxy. For example, over the past 20 years the 31st pick has performed comparably to the 29th pick, but with far lower risk and guaranteed money. That is a predictable pattern that also offers some insight for a GM in terms of where to target asset purchases at draft time.

 

[TroyOLeary]

From that same forum, I think this image of the last 30 years is more interesting:
 
http://i.imgur.com/jsXSlmz.png
 
Some curiousities:

• The 6th pick has been strangely bad - 9 All-NBA players picked 5th only 1 pick 6th 
• No 12th pick has made multiple all-star teams
• No 22nd pick has won a championship

 

[bowiac]

No, they all do suck. Basketball doesn't really break down in that fashion. On the micro-level the stats are solid, but the all-in-ones all reflect the subjective preferences of the people that create them. Also PER doesn't "reward chucking". It attempts to deal with the reality that, strictly speaking, points scored has the greatest correlation to winning and people that produce them have some value (if PER really "rewarded chucking" Rudy Gay would still be suiting up for Hollinger's team). But both win shares and w/48 reward guys for not shooting at all, even though the inability to score actively hurts a team's ability to win. So, no, they aren't "surprisingly good" unless by "surprisingly good" you mean "they're 3% better than my old Magic 8 Ball".

The bolded is inane. Rudy Gay rates as a good player by PER. That has nothing to do with suiting up for Hollinger's team, other than it suggests the Grizzlies don't take PER that seriously. Besides the question of how much decision-making power Hollinger really has, the fact that he invented PER doesn't mean he's dumb enough to take it seriously as a stat. That doesn't mean PER doesn't reward chucking, unless you think taking a a bunch of 32% FGAs isn't chucking. 
 
Win shares does not reward you for not shooting. It does reward efficiency too much I agree, but usage (i.e. shooting) factors heavily into the formula. Look at the league leaders in offensive win shares, and you'll see mostly a bunch of guys with high usage rates. I agree it misses some degree of shot creation, and obviously doesn't factor into it much in the way of team dynamics, but that's not the same thing as being worthless.
 
To each his own I suppose.

 

[ivanvamp]

I'd rather dispense of the stats and just watch the games. Baseball it is not.
As for the "chucking" argument, efficiency has gone from being an underrated trait to an overrated one in my opinion

 
As a basketball coach myself (obviously not on a college or NBA level), I greatly value efficiency.  I use a very simple efficiency formula - something that is easy to grasp and calculate:  points per field goal attempt.  There are, I'm sure, far better, more complex ways of calculating it, but this really works for me.
 
So take two guys from the Bulls-Thunder game last night:  Tony Snell (Chi) and Nick Collison (OK).  Here were their respective lines:
 
Snell:  41 min, 4-15 fg, 2-11 3ptfg, 0-0 ft, 10 points
Collison:  20 min, 4-6 fg, 0-0 3ptfg, 1-1 ft, 9 points
 
So Snell scored 10 points on 15 field goal attempts.  His efficiency was therefore 0.667.  Collison scored 9 points on 6 field goal attempts, and thus his efficiency was 1.500.  
 
Snell scored more points in the game, so he helped his team more in that most important department, right?  Not in my book.  He required so many more shots to get those 10 points.  Collison, meanwhile, was brutally efficient.  When he decided to take a shot, he averaged 1.5 points per shot attempt.  Snell?  0.67.  
 
Some guys are volume scorers, but give me efficient guys all day long.  And it's not just about higher shooting percentage either.  You could get 20 points by going 8-12 from the floor with 4 free throws, or you could get to 20 points by going 5-12 from the floor, having 4 of them be 3's, and hit 6 free throws.  Either way works, and the point per field goal attempt number is still the same.

 

[The X Man Cometh]

For cost/benefit type of analysis, WS/48 is actually a reasonable proxy. For example, over the past 20 years the 31st pick has performed comparably to the 29th pick, but with far lower risk and guaranteed money. That is a predictable pattern that also offers some insight for a GM in terms of where to target asset purchases at draft time.

 
I don't think that's a useful conclusion at all.
 
Most of the "win shares" a vet accumulates are on his second contract, third contract, etc. Where the draft slot money isn't a factor.
 
In this analysis you have long-career journeyman/vets contributing most of the "value" that was drafted at each spot. Most of their careers are played on free agent deals.
 
And imo win shares miss the point here, because they ascribe a linear relationship between "win shares" and value. Both Chandler Parsons has contributed 2x as much WS/48 so far in his career as James Anderson. But Parsons is not merely 2x as valuable.
 
 

 
As a basketball coach myself (obviously not on a college or NBA level), I greatly value efficiency.  I use a very simple efficiency formula - something that is easy to grasp and calculate:  points per field goal attempt.  There are, I'm sure, far better, more complex ways of calculating it, but this really works for me.
 
So take two guys from the Bulls-Thunder game last night:  Tony Snell (Chi) and Nick Collison (OK).  Here were their respective lines:
 
Snell:  41 min, 4-15 fg, 2-11 3ptfg, 0-0 ft, 10 points
Collison:  20 min, 4-6 fg, 0-0 3ptfg, 1-1 ft, 9 points
 
So Snell scored 10 points on 15 field goal attempts.  His efficiency was therefore 0.667.  Collison scored 9 points on 6 field goal attempts, and thus his efficiency was 1.500.  
 
Snell scored more points in the game, so he helped his team more in that most important department, right?  Not in my book.  He required so many more shots to get those 10 points.  Collison, meanwhile, was brutally efficient.  When he decided to take a shot, he averaged 1.5 points per shot attempt.  Snell?  0.67.  
 
Some guys are volume scorers, but give me efficient guys all day long.  And it's not just about higher shooting percentage either.  You could get 20 points by going 8-12 from the floor with 4 free throws, or you could get to 20 points by going 5-12 from the floor, having 4 of them be 3's, and hit 6 free throws.  Either way works, and the point per field goal attempt number is still the same.

 
I don't disagree that it is important to consider the opportunity cost of shots - that using a possession takes away the potential for a better take. But sometimes, when you're across from the best basketball players in the world, you end up having to take bad shots or force it. Efficient players who don't take bad shots, aren't statistically punished for their inability to make them when they are needed. Conversely, players who are "better bad shot takers" aren't rewarded for making them.

Like the Celtics this year. You need Crawford, when Bass and Green can't get a good shot, to barrel into the paint and hoist spinning floaters. Because he's more likely to hit some strange contested shot than anyone else on the roster, and a "system shot" isn't creating itself. His reckless style creates offense in a way that percentages don't encompass. Ditto Russell Westbrook, which is why the Thunder collapsed in his absence, when the stats people would contend that the ever efficient Kevin Durant could just take more shots.

 

[nighthob]

For cost/benefit type of analysis, WS/48 is actually a reasonable proxy. 

 
It really isn't. This isn't baseball, no number will ever tell you anything. The micro numbers are great, we can really drill down into even situational evaluations of skills with the SportVu data. But the all-in-one formulae are just plain miserable. I don't have any preference for PER as anything more than a quick and dirty number to measure relative productivity on a per minute basis, but it doesn't really have any value beyond that (and neither do any of the other all-in-ones).

 

[nighthob]

The bolded is inane. Rudy Gay rates as a good player by PER. 

 
Actually he doesn't. He rates a cunt hair over average per PER. Which is probably why Hollinger (you know, the guy responsible for PER?) traded him to Toronto for Ed Davis and Tayshaun Prince's re-animated corpse.

 

[bowiac]

 
Actually he doesn't. He rates a cunt hair over average per PER. Which is probably why Hollinger (you know, the guy responsible for PER?) traded him to Toronto for Ed Davis and Tayshaun Prince's re-animated corpse.

Regardless of the semantics of "good" (a 16.2 career PER is plenty good), Hollinger: 1) didn't trade Rudy Gay - he's neither the GM nor the owner; 2) doesn't necessarily believe PER. I would suggest he probably doesn't give PER much weight, in spite of inventing it, cause by all indications he's a bright guy, and PER doesn't have much value.
 
Are you really saying a stat that thinks every additional shot you take above 30.4% is great does not reward chucking?

 

[ivanvamp]

 
I don't think that's a useful conclusion at all.
 
Most of the "win shares" a vet accumulates are on his second contract, third contract, etc. Where the draft slot money isn't a factor.
 
In this analysis you have long-career journeyman/vets contributing most of the "value" that was drafted at each spot. Most of their careers are played on free agent deals.
 
And imo win shares miss the point here, because they ascribe a linear relationship between "win shares" and value. Both Chandler Parsons has contributed 2x as much WS/48 so far in his career as James Anderson. But Parsons is not merely 2x as valuable.
 
 
 
I don't disagree that it is important to consider the opportunity cost of shots - that using a possession takes away the potential for a better take. But sometimes, when you're across from the best basketball players in the world, you end up having to take bad shots or force it. Efficient players who don't take bad shots, aren't statistically punished for their inability to make them when they are needed. Conversely, players who are "better bad shot takers" aren't rewarded for making them.

Like the Celtics this year. You need Crawford, when Bass and Green can't get a good shot, to barrel into the paint and hoist spinning floaters. Because he's more likely to hit some strange contested shot than anyone else on the roster, and a "system shot" isn't creating itself. His reckless style creates offense in a way that percentages don't encompass. Ditto Russell Westbrook, which is why the Thunder collapsed in his absence, when the stats people would contend that the ever efficient Kevin Durant could just take more shots.

 
That's fine, and I kind of like those players as leaders of a second unit, given that most second-unit players aren't great scorers.  

 

[nighthob]

Regardless of the semantics of "good" (a 16.2 career PER is plenty good), Hollinger: 1) didn't trade Rudy Gay - he's neither the GM nor the owner; 2) doesn't necessarily believe PER. I would suggest he probably doesn't give PER much weight, in spite of inventing it, cause by all indications he's a bright guy, and PER doesn't have much value.
 

He is the VP in charge of basketball operations, if you want to maintain that he has no say in personnel matters and that it's a magical coincidence that a max player that doesn't show well in any of the magic all-in-one formulae, including the one Hollinger is famous for, got shipped out right after his hiring, fine. But the Grizzlies would likely disagree with your contention.
 
I expect that no serious basketball executive uses any of the magic numbers in player evaluation because in a team sport it's pretty easy to see what a player's role is and there are infinitely better ways of evaluating how well a player performs his role than any of the magic numbers.
 
EDIT: Also, I meant to mention, the 30.4% thing is a Dave Berri claim, made when thumping his chest about the superiority of w/48. Unfortunately it's also wrong. The value of shots is calculated relative to year, always, so the value of two point shots, three point shots and missed FGA (because they're in there too) are always changing.

 

[bowiac]

I expect that no serious basketball executive uses any of the magic numbers in player evaluation because in a team sport it's pretty easy to see what a player's role is and there are infinitely better ways of evaluating how well a player performs his role than any of the magic numbers.

I guarantee you this is incorrect. I know the Rockets use adjusted +/- numbers extensively, and Cuban has indicated the Mavericks do as well, although I don't have first hand knowledge there.

 

[nighthob]

Adjusted +/- numbers aren't at all like the magic all-in-ones and don't claim to be a complete numeric valuation of a player (as anyone will tell you they change radically from year to year depending on the overall construction of the player-to-be-evaluated's team).

 

[bowiac]

Adjusted +/- numbers aren't at all like the magic all-in-ones and don't claim to be a complete numeric valuation of a player (as anyone will tell you they change radically from year to year depending on the overall construction of the player-to-be-evaluated's team).

I don't know what you mean by this. They claim to be measuring the same things as win shares claims to, and win shares-type stats also change based on roster construction.

 

[nighthob]

Adjusted +/- measures the point differentials when a player is on the floor. But, unlike w/48, doesn't claim to have predictive power and doesn't claim to represent the totality of a player's contribution. It's very creators present it as an "in addition to" number to go along with the micro-stats. Overall the micro-numbers, especially now in the age of sportvu (which allow us to evaluate situational performance), give you a far superior picture than the magic numbers do.

 

[Nick Kaufman]

Regardless of the semantics of "good" (a 16.2 career PER is plenty good), Hollinger: 1) didn't trade Rudy Gay - he's neither the GM nor the owner; 2) doesn't necessarily believe PER. I would suggest he probably doesn't give PER much weight, in spite of inventing it, cause by all indications he's a bright guy, and PER doesn't have much value.
 
Are you really saying a stat that thinks every additional shot you take above 30.4% is great does not reward chucking?

I disagree. By default a PER of 15 is considered average. Therefore, one would tend to call a PER of 16 a tad above average.

Moreover, you have to consider why PER was invented in the first place. As a casual fan, I might look at Gay's stats and see he's scoring 20 ppg and think to myself that this guy is good on par with the other all star scorers. But then you look at PER and you understand he's just above average. You look what other people with a PER of 16 make and you understand that Gay is massively overpaid.

I guess you could argue that Gay is worse than average. And you can make a pretty good argument at that. But then again, as others have said there is a value in taking shots too. So that particular PER doesn't look to me as way off.

 

[Nick Kaufman]

Btw, the original draft analysis is sort of interesting, though one should take some parts as statistical artifacts. The no 5 isn't better than the no 2 spot nor does it make some teams choose better; it's just variance.

OTOH, you could do stuff like Win Shares per player per season per draft position and calculate the average win share you should expect getting from a player for each season. Or you could calculate what the chances are of getting a player above a certain Win Share threshold per season per draft position.

 

[bowiac]

Adjusted +/- measures the point differentials when a player is on the floor. But, unlike w/48, doesn't claim to have predictive power and doesn't claim to represent the totality of a player's contribution. It's very creators present it as an "in addition to" number to go along with the micro-stats. Overall the micro-numbers, especially now in the age of sportvu (which allow us to evaluate situational performance), give you a far superior picture than the magic numbers do.

I don't know where you're getting any of these claims from (WS/48 claims to have predictive power, APM stuff doesn't, etc...) but it doesn't match up with what I've read from the creators of any of these stats, nor does your understanding prevent one from using WS/48 as an "in addition to" stat. They can be used in much the same ways, and are used in much the same ways. The APM numbers are more "black-box"ish, but probably better overall, which is both good and bad.

 

[bowiac]

I guess you could argue that Gay is worse than average. And you can make a pretty good argument at that. But then again, as others have said there is a value in taking shots too. So that particular PER doesn't look to me as way off.

I didn't bring up Gay - I brought up that PER rewarded a 30.4 FG%, which somehow led to Gay.
 
I do think he's significantly overrated by PER, but he's not the patriarch of that particular group. 

 

[nighthob]

I don't know where you're getting any of these claims from (WS/48 claims to have predictive power, APM stuff doesn't, etc...) but it doesn't match up with what I've read from the creators of any of these stats, nor does your understanding prevent one from using WS/48 as an "in addition to" stat. They can be used in much the same ways, and are used in much the same ways. The APM numbers are more "black-box"ish, but probably better overall, which is both good and bad.

 
There's not really much of anything black boxish about any of them. But adjusted +/- are pretty concrete numbers. PER just claims to measure overall productivity whereas the win shares and w/48 (that last being the worst of all the magic numbers) attempt to tie the creators' subjective preferences to something concrete. They just don't work and the quants themselves complain that their magic numbers aren't taken seriously by actual NBA people. And this proves, to them, that the league people are idiots. The alternative, that the pros are using more meaningful numbers hasn't dawned on them yet (Synergy Sports is years old and now SportVu gives you even better drill down on the micronumbers, and the NBA has now made it universal).

 

[nighthob]

I didn't bring up Gay - I brought up that PER rewarded a 30.4 FG%

 
It doesn't. Berri forgot that PER is relative to season and that missed FGA are treated as a possession (and hence cost you points based on a possession change). Berri maintains that you somehow magically "lose points" even when there's no possession change (which occurs anywhere from 25%-30% of the time in any given season).

 

[bowiac]

There's not really much of anything black boxish about any of them. But adjusted +/- are pretty concrete numbers. 

APM is black boxish in that doesn't tell what exactly a player is doing right/wrong. It just spits out a number telling you his contribution, but it doesn't tell if you a guy is taking too many long 2s, or is turning the ball over too much. It just gives you an end result on each end of the court.
 

PER just claims to measure overall productivity whereas the win shares and w/48 (that last being the worst of all the magic numbers) attempt to tie the creators' subjective preferences to something concrete. They just don't work and the quants themselves complain that their magic numbers aren't taken seriously by actual NBA people. And this proves, to them, that the league people are idiots. The alternative, that the pros are using more meaningful numbers hasn't dawned on them yet (Synergy Sports is years old and now SportVu gives you even better drill down on the micronumbers, and the NBA has now made it universal).
 

I don't know where you're getting any of this ("attempts to tie the creators' subjective preference")? And please find for me Kubatko complaining that his numbers aren't taken seriously by NBA people or that the people in the league are idiots. This is a series of strawman arguments you're making here, insofar as neither I, nor the creators of WS have made these claims. I'm not citing Dave Berri here, nor have I cited a stat of his.

 

[nighthob]

I don't know where you're getting any of these claims from (WS/48 claims to have predictive power, APM stuff doesn't, etc...) but it doesn't match up with what I've read from the creators of any of these stats, nor does your understanding prevent one from using WS/48 as an "in addition to" stat. They can be used in much the same ways, and are used in much the same ways. The APM numbers are more "black-box"ish, but probably better overall, which is both good and bad.

 
Berri himself claims that adjusted +/- formula don't have predictability (and I don't know any advocate of them that maintains that any given player's adjusted +/- will transfer from year to year or team to team) and that that's the reason his magic number is so much better than anyone else's, because it moves with them from team to team.
 
But, honestly, there are far better ways of measuring the defensive and rebounding contributions of players than w/48. Especially in the age of situational micro-stats.