The lastest draft number-crunch from ESPN's John Hollinger has some harsh words for Westbrook, but Augustin gets the nod...
Once again, we boldly venture into the inexact science of predicting what 21-year-olds will do when they're 25.
For those who missed Part I on Friday, we're taking an analytical approach to the draft by looking at college stats from the past six years and seeing which players swam and which ones sunk once they reached the NBA (that's if they reached the NBA).
Friday's piece looked at the frontcourt players. Today we're going to look at the perimeter.
As noted Friday, I've scrapped the formula I used a year ago and instead applied regression analysis, with the goal being to predict a player's third-year Player Efficiency Rating (PER). The idea is that we can use the results to identify current college players who have the attributes of successful NBA players. Obviously, this analysis is only one piece of a scout's work -- factors like conditioning, work ethic, and other intangibles play an important role in a player's development, too.
Nonetheless, a player's college stats have a surprising predictive power when it comes to determining his success. For both frontcourt and perimeter players, the college stats explain, on average, about 82 percent of their pro performance. Another way of saying this is that the average error is about 18 percent. As I said, it's an inexact science.
One important test is how this system fares vs. the actual draft. And as you can see from the side-by-side comparison at the bottom of the page, this method produced a more accurate draft board than the actual draft. In other words, this system has done a better job of sorting out and selecting prospects than NBA teams have done in the draft the past few years. This is true for both frontcourt and perimeter players.
"Perimeter players" actually consists of two groups -- I had to split them into point guards and wings, as the historical projections were much more accurate this way. With all that said, let's dive in.
I'll start with the point guards, since they're the big story here. Dating back to 2002, all 12 point guards with a rating above 14.0 have been first-round picks, and the three above 16.0 were all lottery picks; this mirrors the trend we saw with the frontcourt players.
And as we saw with that group, the players rated between 12.5 and 14.0 tend to be hit and miss -- a Jameer Nelson here, a Steve Logan there -- while those below 12.5 have trouble establishing a career.
Top-rated point guards since 2002 Player Projected Yr. 3 PER Draft Year Chris Paul 18.40 2005 T.J. Ford 16.67 2003 Raymond Felton 16.29 2005 Mike Conley 15.55 2007 Marcus Williams 15.52 2006 Deron Williams 15.52 2005 Rajon Rondo 15.23 2006 Jason Williams 14.73 2002 Javaris Crittenton 14.51 2007 Devin Harris 14.40 2004
However, the failure rate of players who rate this low isn't as bad as it is for frontcourt players. Of the 30 point guards rated between 10.0 and 12.5, five (Nate Robinson, Jannero Pargo, Chris Duhon, Steve Blake and Luke Ridnour) became regular rotation players, while two others (Aaron Brooks and Ramon Sessions) seem to be well on their way and another two (Travis Diener and Marcus Banks) can't be ruled out. That said, the odds of failure are still far greater than the odds of success for a player rated this low.
Below 10.0, the only two point guards to establish any kind of traction in their careers are Chris Quinn and Royal Ivey; most didn't even get a sniff of the NBA.
Knowing the history, let's take a look at our point guards for this year, where the top-rated point player is, well ... let's just say it's not who you think.
D.J. Augustin, Texas (14.88); Derrick Rose, Memphis (14.69); Mario Chalmers, Kansas (14.03); Jerryd Bayless, Arizona (14.03).
Surprisingly, nobody rates as a slam-dunk lottery pick, not even Derrick Rose. In fact, Rose came in only second here, partly because he had such a slow start to his freshman season before picking up steam at the end.
Is this enough reason to draft Augustin ahead of Rose? No, because the margin between them is miniscule, and even if their career PERs end up the same, Rose will have far more defensive value given his superior size.
The bigger question is whether a team can justify taking Rose ahead of Michael Beasley, whose 19.19 is the best mark by anyone in the six years for which I have data. Yes, Beasley appears to be a space cadet and that's troubling, but what these numbers say to me is that the talent disparity is simply too big. That is, unless Beasley is such a train wreck off the court that he sabotages his own career, he's probably going to have much better results than Rose. In fact, you can make a strong case that Kevin Love should rank ahead of Rose on draft boards as well.
Here's another interesting fact: Rose rated slightly higher as a wing (15.34) than as a point guard. That seems crazy, but he has the size to play the 2 if he has to. Just humor me and store that in the recesses of your brain in case you need to access it in a few years.
Finally, it's notable that Chalmers, though widely presumed to be a fringe first-rounder, rates even with Bayless and pretty close to Rose and Augustin.
Mike Green, Butler (12.75)
Who? The four-year player from Butler is on the fence to even get drafted, but the numbers say he has a good chance at becoming a decent pro.
Better update that passport
Drew Neitzel, Michigan State (11.14); George Hill, IUPUI (11.05); Sean Singletary, Virginia (10.88); Tyrone Brazelton, Western Kentucky (10.84); Brian Roberts, Dayton (10.73); Jamar Butler, Ohio St. (9.72); Mike Taylor, Iowa St. (8.32).
Despite my hopes that Singletary would double the NBA's contingent of Virginia Cavaliers, it appears he's facing long odds.
Taylor is the only one of these players who has been mentioned as a first-round possibility, but he has the worst projection of any of them. But note that Taylor's numbers are from his 2006-07 season at Iowa State.
The others look like second-rounders at best.
OK, on to the wing men. This isn't exactly a banner year for the 2 and 3 positions, as we're about to see, but a few players should be able to push legitimately into the first round.
First, a look at history. You'll notice the same story as with the point guards and centers. A rating above 16 pretty much guarantees a lottery spot; in fact every player rated above 15 in the last six years was a lottery pick, and every player above 14 was a first-rounder.
As with the other spots, things get dodgy between 12.5 and 14, where we've had a few very good players (Caron Butler and Ben Gordon) and a couple of very bad ones (Julius Hodge and Vincent Yarbrough). Most but not all were first-rounders.
The interesting part for wings is that players rated between 11 and 12.5 have been much more successful than those at other positions -- 15 of the 38 such players in the past six drafts eventually cracked a pro rotation, and a couple (Kevin Martin and Mo Williams) became really good.
For players rated below 11.0, the luck ran out -- we see a few rotation players but no long-term starters. But the odds of bucking the projections still appear better for wings than for players at any other position.
Top-rated wings since 2002 Player Projected Yr. 3 PER Draft Year Dwyane Wade 17.81 2003 Carmelo Anthony 17.38 2003 Luol Deng 16.71 2004 Rudy Gay 16.21 2006 Julian Wright 15.86 2007 Rashad McCants 15.41 2005 Andre Iguodala 15.16 2004 Danny Granger 14.55 2005 Shawne Williams 14.37 2006 Josh Howard 14.31 2003
Looking at our top candidates this year, they're going to need to buck those odds, because only a couple project as long-term rotation players in the NBA.
Joe Alexander, West Virginia (14.58).
That's it. Thanks for dropping by, everyone.
We'll see a couple other highly-touted wings further down the list, but Alexander is the only one who cracks the threshold of a typical lottery selection. And in this case, the fact that Alexander took up the game at a late age means we're probably understating his pro potential.
Donte Greene, Syracuse (13.17); O.J. Mayo, USC (13.00); Chris Douglas-Roberts, Memphis (12.78).
This is going to surprise some people, as Mayo has been projected as high as No. 3 in the draft. Truth be told, in this draft he probably deserves to be a lottery pick even going by these ratings -- he's rated 16th overall, and his one season of stats means his estimate is more prone to error than the projections of players who stayed in school longer. But he'll be 21 on November 5, and this system takes age into account when projecting potential.
Greene and Douglas-Roberts are where everyone expects, so I don't have much to add here.
Jamont Gordon, Mississippi St. (12.35); Reggie Williams, VMI (12.22); Pat Calathes, St. Joe's (11.95); Eric Gordon, Indiana (11.93); Russell Westbrook, UCLA (11.91); Malik Hairston, Oregon (11.85); Anthony Randolph, LSU (11.78).
OK, another controversial outcome. Both Gordon and Westbrook are projected lottery picks, but neither fares well here.
Though Westbrook is seen by many as a point guard, he actually rated even worse when I tried him there. It's possible he was just playing out of position at UCLA, but the projections say it's not worth using a top-10 pick to find out.
Eric Gordon's rating is less surprising to me -- subjectively, I've been suspicious of him for some time, and I'm a little unsure what has everyone so excited.
I also tried Randolph here, since some scouts see him as a 3 rather than a 4, and his rating at small forward was better than his awful one as a power forward. It's still hard to put him in the first round based on these numbers, but he's at least draftable.
Jamont Gordon shapes up as a bit of a sleeper -- he rated even higher as a point guard, and at 6-4 could be a decent third guard backing up both positions. At the moment most see him as a second-rounder.
Bill Walker, Kansas St. (11.74); Kyle Weaver, Washington St. (11.61); Brandon Rush, Kansas (11.34); Luc Richard Mbah a Moute, UCLA (11.27); Josh Duncan, Xavier (11.25); Courtney Lee, Western Kentucky (11.17).
Because of the general lack of depth in this draft, a number of these players are projected as potential first-rounders for whom the numbers say buyer beware.
Walker may be the most interesting of the bunch since the evaluation period covers a time when he was coming back from a serious knee injury, so it's possible he'll be a much better pro than this projection indicates. Rush may also have a bit more value than the numbers show because of his rep as a defensive ace.
Better update that passport
Davon Jefferson, USC (10.99); Chris Lofton, Tennessee (10.80); DeMarcus Nelson, Duke (10.78); Sonny Weems, Arkansas (10.76); Shan Foster, Vanderbilt (10.58); Richard Roby, Colorado (9.77); Gary Forbes, Massachusetts (9.49); J.R. Giddens, New Mexico (9.16); Bryce Taylor, Oregon (8.79); Marcelus Kemp, Nevada (8.16).
No big surprises here, as none of these players are likely to go before the middle of the second round.
For posterity's sake, here's how the top 15 picks rank in this year's draft, regardless of position. As you can see, my system rates it a strong draft for frontcourt players and a weak one on the perimeter.
2008 Draft: Top 15 Projected Yr-3 PER Player Projected Yr. 3 PER School Michael Beasley 19.19 Kansas St. Kevin Love 17.80 UCLA Darrell Arthur 15.82 Kansas Marreese Speights 15.02 Florida D.J. Augustin 14.88 Texas Derrick Rose 14.69 Memphis Joe Alexander 14.58 West Virginia Brook Lopez 14.21 Stanford Mario Chalmers 14.03 Kansas Jerryd Bayless 14.03 Arizona Roy Hibbert 14.02 Georgetown Kosta Koufos 13.32 Ohio State Donte Greene 13.17 Syracuse Darnell Jackson 13.17 Kansas DeAndre Jordan
13.17 Texas A&M
Finally, as promised above, here's how my system rated the top college perimeter players in the past six drafts, alongside who was actually chosen in the first round of each draft. (The same kind of info was provided on Friday for frontcourt players.) As we saw in the analysis of frontcourt prospects, while the system isn't perfect, it's a clear improvement on what actually took place.
Top-rated perimeter players of recent drafts Player Proj. Yr 3 Year Actual order Jay Williams 14.73 2002 Jay Williams Mike Dunleavy 14.28 2002 Dunleavy Dajuan Wagner 13.97 2002 Wagner Caron Butler 13.78 2002 Butler Steve Logan** 13.61 2002 Jones Vincent Yarbrough** 13.60 2002 Dixon Casey Jacobsen 12.98 2002 Rush Tayshaun Prince 12.57 2002 Jacobsen Kareem Rush 12.55 2002 Prince Tony Akins* 11.95 2002 Frank Williams Fred Jones 11.84 2002 Salmons John Salmons 11.76 2002 Chris Jefferies Juan Dixon 11.56 2002 Dan Dickau
Dwyane Wade 17.81 2003 Anthony Carmelo Anthony Wade 17.38 2003 Wade T.J. Ford Hinrich 16.67 2003 Hinrich Josh Howard 14.31 14.31 2003 Ford Kirk Hinrich 13.65 2003 Hayes Luke Walton** 13.07 2003 Marcus Banks Reece Gaines 12.57 2003 Ridnour Jarvis Hayes 12.35 2003 Gaines Kyle Korver** 12.16 2003 Troy Bell Luke Ridnour 12.08 2003 Dahntay Jones Marquis Daniels* 11.79 2003 Howard
Luol Deng 16.71 2004 Gordon Andre Iguodala 15.16 2004 Harris Devin Harris 14.40 2004 Childress Josh Childress 13.78 2004 Deng Ben Gordon 13.78 2004 Iguodala Jameer Nelson 13.74 2004 Jackson Delonte West 13.55 2004 Snyder Luke Jackson 13.50 2004 Nelson Kirk Snyder 13.24 2004 West Tony Allen 12.43 2004 Allen Blake Stepp** 12.25 2004 Kevin Martin
Chris Paul 18.40 2005 Williams Raymond Felton 16.29 2005 Paul Deron Williams 15.52 2005 Felton Rashad McCants 15.41 2005 McCants Danny Granger 14.55 2005 Antoine Wright Jarrett Jack 14.20 2005 Joey Graham Julius Hodge 13.89 2005 Granger Francisco Garcia 13.48 2005 Hodge Aaron Miles* 13.39 2005 Robinson Linas Kleiza 12.37 2005 Jack Travis Diener** 12.35 2005 Garcia Jawad Williams* 12.31 2005 Luther Head Travis Diener** 12.35 2005 Garcia Nate Robinson 12.25 2005 Kleiza
Rudy Gay 16.21 2006 Morrison Marcus Williams 15.52 2006 Roy Rajon Rondo 15.23 2006 Randy Foye Shawne Williams 14.37 2006 Gay Brandon Roy 14.17 2006 Redick Jordan Farmar 14.07 2006 Brewer Hassan Adams** 13.79 2006 Rodney Carney Renaldo Balkman 13.75 2006 Williams Ronnie Brewer 13.44 2006 Quincy Douby Daniel Gibson 13.30 2006 Balkman P.J. Tucker** 13.08 2006 Rondo Dee Brown** 12.95 2006 Williams Kyle Lowry 12.75 2006 Lowry Adam Morrison 12.71 2006 Shannon Brown J.J. Redick 12.63 2006 Farmar
Julian Wright 15.86 2007 Conley Thaddeus Young 15.31 2007 Brewer Javaris Crittenton 14.51 2007 Law Jeff Green 14.20 2007 Young Rodney Stuckey 14.02 2007 Wright Taurean Green** 13.64 2007 Thornton Corey Brewer 13.55 2007 Stuckey Marcus Williams 13.38 2007 Nick Young Wilson Chandler 13.01 2007 Crittenton Dominic McGuire** 12.90 2007 Cook Acie Law 12.83 2007 Dudley Reyshawn Terry** 12.76 2007 Morris Almond Jarrious Jackson** 12.08 2007 Aaron Brooks Jared Dudley 11.94 2007 Arron Afflalo Al Thornton 11.76 2007 Alando Tucker