Players mislabeled as "busts?" Who am I missing?

[@docoolstuff]

I just compiled a list of a few players who have been labeled as "busts" or disappointments despite putting up some pretty damn good seasons simply because they were drafted early. My list with my explanations are here: http://lifeisapeachbasket.blogspot.com/ ... ought.html

"Busts"
Shawn Bradley
Tony Battie
Eddie Griffin
Josh Childress
Martell Webster
Brandan Wright
Al-Farouq Aminu

Honorable Mentions:
Raef Lafrentz
Nick Collison
Drew Gooden
Jordan Hill
Tristan Thompson

Always appreciate feedback! Who am I missing?

[bchaikin]

agreed - shawn bradley was far from being a bust. compare the first 11 years of the careers of shawn bradley and mark eaton - statistically you'll find little difference...

[Mike G]

I don't know much about who might have been called a 'bust', and I don't know that there'd be a way to quantify it.
But I am pretty sure you'll not get much serious interest in a 'study' that uses WP as though it means anything.

[Dr Positivity]

I think one of the reasons is for a lot of these players, the team that drafted them didn't get much out of the pick. So from the team perspective the pick was a bust, even if the player went on to be good

[Mike G]

Yes; it's likely a drafting team's fan base and media would apply the label to a player who is seen to have been a waste of a high pick.
Meanwhile, looking at Win Shares at this page: http://bkref.com/tiny/vIzCj
... it's seen that in the lottery era (1985- ), Bradley ranks 17th of 29 players picked at #2. Below him are Marvin and Derrick Williams, Beasley, Evan Turner, Kidd-Gilchrist, and Oladipo among active players. Some of these could pass him before they're done.
Ferry, Swift, Darko, Jay Williams (and Bias) won't be passing anyone.
And there's Thabeet.

So of these 29, Bradley will presumably wind up 17th to 20th. The term 'bust' might be reserved for the bottom 20% or so?
Judging by the Median WS for a given draft pick is a lot kinder than judging relative to the Mean. But a surprise like David Lee at #30 -- he already has the career WS (64) of the median #1 pick -- shouldn't lead us to think worse of an avg #30 pick. Half of those taken at #30 have <1 WS.

After a lot of smoothing, here are suggested 'expected' career Win Shares (regular season) for various draft picks since 1985.
It's just multiplying the median at each pick by the factor 1.15 to adjust for players whose careers are not completed.
I'd also suggest that less than half of each median is arguably a bust.

Code: Select all

pk  exp        pk   exp        pk   exp
1    63        11    18        21   10
2    54        12    15        22   10
3    48        13    13        23    9
4    42        14    11        24    8
5    35        15     9        25    7
6    30        16     9        26    6
7    27        17     9        27    4
8    24        18     9        28    3
9    22        19     9        29    2
10   20        20    10        30    2

No amount of smoothing seems to take out the bulge around pick #21 -- relative to the slump at 16-18. Maybe smarter teams are drafting then?

[bbstats]

I think this list is very funny because adjusted +/- *also* looks extra-favorably (though moreso) upon Collison and Bradley, and you're using WP (!).

I will not use my posting power to hijack this thread into a discussion about WP...I will not use my posting power to hijack this thread into a discussion about WP...I will not use my posting power to hijack this thread into a discussion about WP...I will not use my posting power to hijack this thread into a discussion about WP...

[@docoolstuff]

All great points and I appreciate all the feedback.

I wanted to be more clear about quantifying what a "bust" was, but defining a quantifiable measure turned into a fool's errand. Additionally, I agree with the ROI sentiment, but I think the point I was making (and should have done a better job explaining) was that even those players who didn't match the expected ROI still generated value somewhere within 1-2 SD of that expected production.

Also, apologies for not varying up the metric used more - I obviously looked at more than WP when compiling the list and will show those next time.

[Mike G]

Also, a return on investment is relative to what's available in a given draft. Win Share totals for rookie classes may vary by a factor of 3. After 5 years, we see these totals for entry years since 1980:

Code: Select all

5yWS  class   PER   WS/48     5yWS  class   PER   WS/48     5yWS  class   PER  WS/48
628   2009   16.5   .121      572   1981   14.8   .101      466   1983   16.3   .117
626   1982   16.1   .103      571   1999   16.5   .116      463   1988   16.1   .120
610   1990   16.2   .119      555   1986   17.1   .124      453   2008   16.3   .127
610   2004   15.2   .099      517   2000   16.9   .123      441   1991   15.3   .106
603   2002   16.1   .114      497   1989   15.2   .107      440   1997   17.2   .126
601   2005   15.9   .116      496   2010   16.6   .120      438   1992   16.7   .119
595   1980   16.4   .123      489   1993   16.0   .099      404   1995   16.2   .116
591   1985   17.9   .139      481   2003   16.1   .117      392   1998   15.8   .114
582   1994   15.6   .113      478   1984   15.4   .112      385   2007   15.3   .103
576   2006   16.4   .114      474   1996   15.9   .112      355   2001   14.1   .080
.                             472   1987   14.7   .098               

The average is about 512 WS in 5 years, and the range is from 1.23 to .69 of this.
Only player-seasons of 900+ minutes are in the sample. That may mostly explain the predominant above-avg performance here.