Page 1 of 2 12 LastLast
Results 1 to 25 of 50
  1. #1
    Get Refuel! FromWayDowntown's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jul 2003
    Post Count
    19,921
    It's an interesting read -- I'm curious to see how much Hollinger adjusted backwards to meet his own beliefs about which players were the best and then created a system to prove that and how much his numbers just happened to support his observations. In any event, it should be something of a conversation starter. Claiming that Durant is significantly better than Carmelo Anthony was (in relative terms) is a pretty bold statement. Here's the piece (it's long):

    A new system to evaluate the pro potential of college players
    By John Hollinger

    As an NBA analyst, few things have confounded me more than trying to evaluate college players for the NBA draft.

    I have a pretty good system to deal with Europeans, at least the ones who get major exposure in the Euroleague (more on that later this week). And at the other end of the spectrum, the high schoolers and random foreign teenage 7-footers have played so few compe ive games that there's nothing much to evaluate.

    But in the middle lie college hoops -- close enough to the NBA game to be ripe for evaluation, yet just different enough to have remained elusive.

    Until now, that is. Today is the unveiling of a system I've developed for evaluating college players' pro potential, one that works very well when back-tested against previous drafts.

    This is version 1.0 of a system that is likely to undergo several iterations in future years, and it's by no means perfect -- there simply is no way to look at a group of 20-year-olds and tell with 100 percent accuracy which ones will be the best at age 25.

    But I feel comfortable releasing it now because the results are a big improvement on the actual drafts, and that's the important thing. While I can't make a system that's omniscient, at least this one doesn't result in consecutive lottery picks being spent on Jared Jeffries, Melvin Ely and Marcus Haislip, which is what actually happened in 2002.

    And on that point, this system clearly holds up. Take that 2002 draft, for example. The best collegian from that draft is Utah Jazz forward Carlos Boozer, who was taken with the 35th overall pick by Cleveland -- after Dajuan Wagner, Jeffries, Ely, Haislip, Ryan Humphrey, Kareem Rush, Casey Jacobsen, Frank Williams, John Salmons, Chris Jefferies, Dan au, Roger Mason, Robert Archibald and Vincent Yarbrough were off the board.

    My formula has Boozer 34 spots higher -- at the top of the draft. In fact, he's one of the five best college prospects to come out in the last half decade, according to the system.

    If you look below at the other charts, you'll notice the top-rated college prospect in each draft (Boozer in '02, Carmelo Anthony in '03, Luol Deng in '04, Chris Paul in '05, and Tyrus Thomas in '06) ended up among the few elite players in that year's college class.


    THE KEY FACTORS

    So how does it work?

    The difficulties of working with college basketball stats are myriad, but it starts with getting an apples-to-apples comparison between players. The raw stats are almost worthless on this front, because differences in both pace and schedule strength can vary much more widely than they do in the NBA.

    So obviously, I needed to adjust for this before I could go any further. The pace adjustment was done using pace factor, as you might expect; for strength of schedule, I used Jeff Sagarin's schedule strength ratings, which are archived at USAToday.com for previous seasons.

    From that point, there's still one big problem -- pro success doesn't always depend on how good you were in college. Obviously it helps, and I use PER (Player Efficiency Rating) as one of the factors in my rating formula, but it doesn't come close to telling the whole story.

    The other six factors that are indicative of pro success are:

    1. Age. Everything about the draft has to be seen through the prism of age. This is hugely important, yet teams underestimate it almost every year.

    That's why "veteran" rookies like Ely, au, Rafael Araujo and Francisco Garcia have underwhelmed at the NBA level, while the freshman stats of a player like Chris Bosh take on new meaning when you understand his youth.

    This year's draft is especially alluring on the youth front, since the NBA's new age restrictions resulted in several coveted players going the one-and-done route -- most notably Kevin Durant and Greg Oden.

    On the other hand, Florida State's Al Thornton is walking around with a giant neon warning sign, as he will turn 24 a little over a month into his rookie season. He's more than a year older than LeBron James, and nearly two years older than Darko Milicic.

    Teams need to view guys like this the same way they look at a grade-school kid who's been left back twice and dominates the lunchtime kickball game, but for some reason a lot of clubs don't.

    2. Steals. Though perhaps the most worthless stat for NBA analysis, there's no denying that college players who get a ton of steals tend to fare much better in the NBA than their less sticky-fingered brethren. This is the one item that gets the most weight, actually -- it's even more important than PER!

    For this year's draft, that's a big positive for Mike Conley, Jr., who picked off more than two balls a game, and a big negative for players like Arron Afflalo (22 all season), Nick Young (27) and Ramon Sessions (29).

    3. Blocks. This is the big man counterpart to steals, basically, although it's not quite as important.

    What's more, players who generate numbers in both categories are hugely successful as a group. Of the 13 NCAA players in the past five years to have a "50-50" season in blocks and steals, nine are in the NBA, five were lottery picks, and one was the MVP of the 2006 NBA Finals -- Dwyane Wade.

    In this year's draft, there is one player who was a "50-50" in college. His name? Kevin Durant.

    4. Rebounds. Boards, especially offensive boards, are a good indicator of future pro success as well.

    Note that it isn't necessarily the absolute rebounds as much as the rebounds given a player's height. Wade, for instance, has the best rebound rate in the past five years of any player 6-4 or shorter -- a whopping 13.0 his sop re year. Rajon Rondo was the best under 6-2 (11.5), and Nate Robinson was the best under six foot (8.6 his sop re year).

    The correlation isn't quite as strong with big men, oddly enough, because you get one-dimensional types muscling in on the action (a lot of marginal players like Reggie Evans). But big men who can rebound and show some skill in other areas tend to fare very well in the pros.

    The most notable rebound rate in this year's draft belongs to an Ohio State player -- just not the one you think. Daequan Cook, a 6-5 freshman, put up a whopping 12.1 mark, which is pretty incredible for a player that size.

    5. 3-pointers. Those previous three items are markers for athleticism, while these next two are markers of skill.

    Despite the longer line, college 3-point shooting translates very well to the NBA level, albeit sometimes with a year or two of adjustment needed. The key here is to look for players who both make a ton of 3s and shoot a high percentage.

    From recent drafts, the standout performers here are Kyle Korver and Salim Stoudamire. Korver nailed 129 at a 48-percent clip his final year at Creighton, while Stoudamire -- though undraftable according to pretty much every other metric -- hit a scalding 50 percent while making 120 his last year at Arizona.

    Nobody stands out quite to that extent in this year's draft.

    6. Pure Point Rating.[/B] Pure point rating is something I created to replace assist-turnover ratio, which is a fairly useless creation because it sticks turnovers in the denominator and thus tends to reward point guards who never penetrate. Pure point rating produces much more relevant rankings of a player's ability as a distributor; the formula is:

    Pure Point = (100* ((Assists * 2/3) - Turnovers))/Minutes

    Steve Nash led the NBA in 2006-07 at 11.3; the next four players were Jason Kidd, Chris Paul, Jose Calderon and Deron Williams.

    I thought this might just separate the wheat from the chaff among point guards, but it actually helps at every position.

    Obviously, guards such as Deron Williams, Marcus Williams, T.J. Ford and Steve Blake differentiate themselves by having college pure point ratings well over 2.0, but wingmen like Andre Iguodala and Luke Walton also helped themselves with extremely strong ratings in this category.

    On the other hand, Alexander Johnson and Rafael Araujo both put up -3.4 marks -- perhaps that should have been a warning sign.

    In this year's draft, Conley's 2.45 mark stands out with an exclamation point -- it's the fourth-best of any college player in the past six years with at least 500 minutes played, and easily No. 1 among this year's players.

    At the other end, Nick Young (-1.8) and Morris Almond (-2.9) have disastrously bad ratings for backcourt players.


    THE RED FLAGS

    I use the six factors above to produce a "rating" of a player's pro potential, but that's not the end of the story. It turns out just using the rating only gets you about halfway there, and still leads to a lot of mistakes.

    The rest of it is taken care of by what I call the four "red flags":

    Short guys. We're all familiar with the great hordes of 5-11 guards who have put up spectacular college hoops numbers only to implode upon reaching the pros. I had to put in a system of deductions to account for this.

    At the guard spot, a player got a minor deduction for being "somewhat short" if he was a 6-3 or 6-4 shooting guard, or a 6-1 or 6-2 point guard. He got a much larger deduction for being "very short" if he was a 6-2 or smaller shooting guard, or a 6-0 or smaller point guard. This seemed to even out a lot of the small-guard issues, as the best small players were able to overcome their size, but the others weren't.

    Up front, I had to put in a fairly severe deduction for short power forwards. Simply put, if you're shorter than 6-9 and plan to play inside in the pros, you'd better have some fairly overwhelming college stats to back you up. Those who do sometimes succeed -- such as Paul Millsap or the pre-Orca-phase Michael Sweetney. Those who don't pretty much always fail.

    7-footers. Hey, I don't make the rules, I just play by 'em. Seven-footers in the drafts I studied tended to greatly underperform their college stats in the pros.

    Perhaps there's an easy explanation for this: If they were that big and hadn't already bolted for the NBA, that was perhaps the first sign something was wrong. Additionally, they sure as heck weren't facing off against many players their own size.

    This affects Greg Oden, Spencer Hawes and Aaron Gray in this year's draft.

    Perimeter players who don't make 3s. Making fewer than 25 3-pointers in the year before being drafted is a pretty strong negative indicator for outside players.

    Some are good enough in other areas to overcome it -- most notably Wade. Many others fail because of it, however, as the athleticism they relied on to dominate in college isn't nearly as singular at the pro level.

    Really bad rebounders. This is a huge red flag -- if a player has an extremely poor rebound rate for his size, it's a strong indication that his athleticism is taxed to the limit even at the college level and he's going to be completely overwhelmed in the pros.

    This has several subsets by position, and as you can see it's kind of a gallery of busts:

    Guards with a rate below 5.0: Stoudamire and Jannero Pargo overcame this enough to become quasi-useful; perhaps J.J. Re will too. Others include au, Daniel Ewing, Reece Gaines, Dajuan Wagner and Roger Mason. This year's draftees who get red-flagged on this metric are JamesOn Curry, Taurean Green and Gabe Pruitt.

    Players in the 6-5 to 6-7 range with rates below 7.5: The most notable past failure here is Casey Jacobsen; two players from this year's draft -- Afflalo and Javaris Crittenton -- raise a red flag here.

    Players in the 6-8 to 6-9 range with rates below 8.0: Your past winners here include Adam Morrison and Jarvis Hayes. Nobody this year qualifies.

    Any inside player with a rate below 12: This is a big one -- Matt Freije, Matt Bonner, Marcus Haislip and Rick Rickert all get tagged from past drafts on this metric. In this year's drafts, there are two big kahunas with decidedly poor rebound rates -- Brandan Wright (11.1) and Spencer Hawes (11.3).

    Any center below 13: If Ely's age hadn't been a reason not to draft him, this should have been. Hawes falls under this category too (I didn't double-ding him).


    THE ONE ANTI-RED FLAG

    One positive, surprisingly, was if a player had a previous season that was better than the one just before the draft.

    You might think that this meant a player was "on the downswing," but actually counting one-third the previous season and two-thirds the current one improved the quality of the draft ratings significantly. Interestingly enough, the opposite didn't work -- counting a previous season where a player was worse didn't help at all.

    What this tells us, apparently, is that with players this young we should take most one-year improvements at face value.


    SUMMING IT UP

    After all that, we finally end up with a numerical rating for each player. While assigning each player a single number can't possibly address all the complexities involved in a draft (present versus future, team needs versus best player, etc.), it does allow us to do a few neat things.

    For starters, we can compare drafts between years, which allows us to see almost immediately that this year's draft is, indeed, absolutely freaking loaded. It has the highest-rated player of the past six years, and seven of the top 23 collegians in that stretch.

    Second, we can denote differences between players much more finely than we can by just using a ranking system. For example, the difference in the 2006 draft between the top-rated player, Tyrus Thomas, and the second-ranked collegian, Shelden Williams, was greater than that between Williams and the No. 35 player -- which would clue you in to just draft Thomas, regardless of need. On the other hand, the difference between Andrew Bogut and Channing Frye in the 2005 draft was only three hundredths of a point.

    Best of all, the system works. Obviously, you want more than just my word, so below are charts showing the top 12 collegians in each draft using my formula, compared to where they actually were picked among collegians and what players went in their place.

    Yes, there are a few stinking dogs thrown in -- Vincent Yarbrough didn't quite pan out, for instance, and the Paul Davis Era in Los Angeles is off to a slow start.

    That's OK. As I said, this is just the initial version of the system, and for it to work, we just have to have fewer busts than in the real drafts. And as you'll see, there are substantial fewer in the lists below.

    Let's look at how the past five drafts play out:


    COMPARISON TO RECENT DRAFTS

    2002 Draft: Top 12 rated players
    NO. PLAYER SCHOOL SCORE PICKED* ACTUAL ORDER*
    1. Carlos Boozer Duke 711.0 26 Jay Williams
    2. Drew Gooden Kansas 678.6 3 Mike Dunleavy Jr.
    3. Chris Wilcox Maryland 608.6 5 Drew Gooden
    4. Curtis Borchardt Stanford 598.1 12 Dajuan Wagner
    5. Mike Dunleavy Jr. Duke 561.0 2 Chris Wilcox
    6. Jay Williams Duke 515.0 1 Caron Butler
    7. Jared Jeffries Indiana 503.3 7 Jared Jeffries
    8. Udonis Haslem Florida 498.9 Undrafted Melvin Ely
    9. Vincent Yarbrough Tennessee 498.5 24 Marcus Haislip
    10. Caron Butler UConn 495.0 6 Fred Jones
    11. Tayshaun Prince Kentucky 480.1 16 Juan Dixon
    12. Casey Jacobsen Stanford 471.5 15 Curtis Borchardt

    * among collegians only

    As I mentioned before, Boozer went 35th in the real draft, but he's at the top of the list in this one. Haslem went completely undrafted but is the No. 8 collegian here, while Prince -- considered something of a reach at the time -- also moves up. Borchardt is a good example of how traditional scouting can complement this kind of system -- everyone avoided him because of a history of foot problems, and sure enough he kept having foot problems.

    Notables: Fred Jones (428.9, 19th); Juan Dixon (390.6, 25th); Jannero Pargo (377.6, 30th); John Salmons (360.1); Matt Barnes (352.0); Dan Gadzuric (285.5); Darius Songaila (249.3); Dan au (215.0)



    --------------------------------------------------------------------------------


    2003 Draft: Top 12 rated players
    NO. PLAYER SCHOOL SCORE PICKED* ACTUAL ORDER*
    1. Carmelo Anthony Syracuse 781.3 1 Carmelo Anthony
    2. Mike Sweetney Georgetown 702.8 7 Chris Bosh
    3. Chris Bosh Georgia Tech 688.4 2 Dwyane Wade
    4. Dwyane Wade Marquette 600.4 3 Chris Kaman
    5. Nick Collison Kansas 553.4 9 Kirk Hinrich
    6. T.J. Ford Texas 549.5 6 T.J. Ford
    7. Kirk Hinrich Kansas 504.0 5 Michael Sweetney
    8. Josh Howard Wake Forest 501.4 17 Jarvis Hayes
    9. Kyle Korver Creighton 499.7 31 Nick Collison
    10. David West Xavier 494.7 14 Marcus Banks
    11. Troy Bell Boston College 481.5 13 Luke Ridnour
    12. Jarvis Hayes Georgia 478.9 8 Reece Gaines

    * among collegians only

    It's tough to improve on the 2003 draft, but the system didn't screw it up either. Sweetney is the one brow-raiser -- he was an absolute monster in college (hard to believe the guy we see now averaged nearly two steals a game at Georgetown) and had a rookie PER on par with that of Wade and Bosh before he started losing the battle of the bulge. Korver, West and Howard were later picks (Korver going in Round 2) who moved way up on this list, ahead of busts Troy Bell and Jarvis Hayes.

    Notables: Marquis Daniels (474.3, 13th); Marcus Banks (472.3, 15th); Chris Kaman (462.7, 16th) Matt Carroll (447.5, 18th); Quinton Ross (443.5, 19th); Luke Ridnour (442.3, 20th); Matt Bonner (440.4, 21st); Maurice Williams (428.5, 23rd); Luke Walton (420.2, 25th); Keith Bogans (362.8); Steve Blake (361.8); Dahntay Jones (358.7); Brian Cook (349.9); Jason Kapono (337.2); James Jones (333.1)



    --------------------------------------------------------------------------------


    2004 Draft: Top 12 rated players
    NO. PLAYER SCHOOL SCORE PICKED* ACTUAL ORDER*
    1. Luol Deng Duke 650.7 5 Emeka Okafor
    2. Delonte West Saint Joseph's 626.9 12 Ben Gordon
    3. Devin Harris Wisconsin 614.6 3 Devin Harris
    4. Emeka Okafor Connecticut 579.4 1 Josh Childress
    5. Luke Jackson Oregon 558.5 8 Luol Deng
    6. Josh Childress Stanford 530.0 4 Rafael Araujo
    7. Ben Gordon Connecticut 529.1 2 Andre Iguodala
    8. Kris Humphries Minnesota 527.6 9 Luke Jackson
    9. Jameer Nelson Saint Joseph's 522.2 11 Kris Humphries
    10. Kevin Martin Western Carolina 517.7 14 Kirk Snyder
    11. Andre Iguodala Arizona 509.5 7 Jameer Nelson
    12. Andre Emmett Texas Tech 472.3 18 Delonte West

    * among collegians only

    Okafor was the top-rated collegian in the eyes of most, but the numbers here said to go with Deng, who has paid dividends as the fifth-drafted collegian for Chicago. Kevin Martin and Delonte West move way up, while you can see the large difference in score between the solid 11th-rated collegian (Andre Iguodala) and the woofer at No. 12 (Andre Emmett). Generally things start getting dicey once the scores dip under 500.

    Notables: Kirk Snyder (464.5, 14th); Chris Duhon (454.3, 15th); David Harrison (432.0, 21st) Tony Allen (377.0, 27th); Royal Ivey (341.8)


    --------------------------------------------------------------------------------


    2005 Draft: Top 12 rated players
    NO. PLAYER SCHOOL SCORE PICKED* ACTUAL ORDER*
    1. Chris Paul Wake Forest 705.9 4 Andrew Bogut
    2. Marvin Williams North Carolina 697.6 2 Marvin Williams
    3. Sean May North Carolina 690.4 9 Deron Williams
    4. Rashad McCants North Carolina 639.4 10 Chris Paul
    5. Andrew Bogut Utah 579.7 1 Raymond Felton
    6. Channing Frye Arizona 579.7 7 Charlie Villanueva
    7. Raymond Felton North Carolina 562.1 5 Channing Frye
    8. Chris Taft Pittsburgh 559.7 30 Ike Diogu
    9. Danny Granger New Mexico 554.4 13 Sean May
    10. Nate Robinson Washington 538.0 16 Rashad McCants
    11. Deron Williams Illinois 523.7 3 Antoine Wright
    12. Jarrett Jack Georgia Tech 523.3 17 Joey Graham

    * among collegians only

    We're still evaluating the 2005 draft, but it's pretty obvious that Paul was the top collegian from that group and should have gone ahead of the likes of Bogut and both Deron and Marvin Williams. If you're looking for Charlie Villanueva, he was 13th. You can also tell that this was a pretty strong group, as scores were still well over 500 even at the 12th position.

    Notables: Charlie Villanueva (521.9, 13th); David Lee (482.7, 19th); Kelenna Azubuike (454.2, 25th) Salim Stoudamire (449.4, 26th); Francisco Garcia (448.8, 27th); Daniel Ewing (446.3, 28th); Chuck Hayes (443.9, 30th); Ronny Turiaf (442.7); Ryan Gomes (430.9); Hakim Warrick (427.8); Luther Head (419.7) Ike Diogu (402.5); Antoine Wright (387.8); Joey Graham (353.8); Jason Maxiell (342.5); Linas Kleiza (308.4)



    --------------------------------------------------------------------------------


    2006 Draft: Top 12 rated players
    NO. PLAYER SCHOOL SCORE PICKED* ACTUAL ORDER*
    1. Tyrus Thomas LSU 756.8 3 LaMarcus Aldridge
    2. Shelden Williams Duke 583.1 4 Adam Morrison
    3. Brandon Roy Washington 557.6 5 Tyrus Thomas
    4. Ronnie Brewer Arkansas 555.8 11 Shelden Williams
    5. Rudy Gay Connecticut 552.1 7 Brandon Roy
    6. Patrick O'Bryant Bradley 551.9 8 Randy Foye
    7. Paul Davis Michigan State 546.6 28 Rudy Gay
    8. Kyle Lowry Villanova 538.6 20 Patrick O'Bryant
    9. Rajon Rondo Kentucky 534.5 17 J.J. Re
    10. LaMarcus Aldridge Texas 524.3 1 Hilton Armstrong
    11. Quincy Douby Rutgers 516.8 15 Ronnie Brewer
    12. Marcus Williams Connecticut 512.1 18 Cedric Simmons

    * among collegians only

    Thomas was head-and-shoulders above the crowd by my method, while second-rounder Davis is a lottery pick by the same reckoning; we'll see if it works out that way in real life. College stars Adam Morrison and J.J. Re are nowhere to be found, both red-flagged by horrid rebound rates, among other woes.

    Notables: Randy Foye (502.3, 13th); J.J. Re (479.6, 18th); Rodney Carney (469.4, 19th); Adam Morrison (466.3, 20th); Renaldo Balkman (458.3, 23rd); Daniel Gibson (456.3, 25th); Jordan Farmar (450.0, 27th); Paul Millsap (440.0); Josh Boone (414.8); Craig Smith (377.8); Hilton Armstrong (304.6)



    --------------------------------------------------------------------------------


    Overall, the system picked out several players who were drafted much, much later and turned into NBA stars. Better yet, it didn't miss out on a single important player. Looking at the past five drafts, the most prominent players taken among the top 12 collegians in their draft who weren't in the top 12 in my system were Randy Foye, who ranked 13th in 2006; Villanueva, who ranked 13th in 2005; and Chris Kaman, who ranked 16th in 2003. Compared to real life -- a system that passed on Carlos Boozer, Udonis Haslem, Tayshaun Prince, David West, Josh Howard, Kevin Martin, Delonte West, and Danny Granger, among others -- I'll take these results in a heartbeat.

    And it's not just the players who were picked that matters, but the ones who weren't. By my count, six of the biggest dogs of the past five drafts -- Dajuan Wagner, Melvin Ely, Marcus Haislip, Reece Gaines, Rafael Araujo, Antoine Wright -- would have landed well outside the top 12 collegians using this method. You can raise that number to 10 if Adam Morrison, J.J. Re , Hilton Armstrong and Cedric Simmons don't turn the corner.

    LOOKING AT THE 2007 DRAFT

    My apologies for the lengthy prelude, since at least half of you probably just scrolled straight down to this part, but I had to point out how well this thing works when used on past drafts if I'm going to have you believe it has some relevance for this one.

    But without further ado, it's time to introduce the top prospects from this year's draft -- as well as the hot prospects who surprisingly didn't make the cut.

    The raw info is in the chart, but here's how it breaks down:

    2007 Draft: Top 30 rated collegians, plus other notables
    NO. PLAYER SCHOOL SCORE CHAD FORD'S RANKING
    1. Kevin Durant Texas 870.7 2
    2. Greg Oden Ohio State 667.9 1
    3. Mike Conley Jr. Ohio State 637.9 7
    4. Thaddeus Young Georgia Tech 604.2 14
    5. Brandan Wright North Carolina 601.4 8
    6. Al Horford Florida 601.0 3
    7. Nick Fazekas Nevada 594.3 35
    8. Josh McRoberts Duke 566.7 26
    9. Rodney Stuckey E. Washington 557.7 16
    10. Jared Dudley Boston College 542.6 31
    11. Joakim Noah Florida 528.6 9
    12. Glen Davis LSU 521.0 25
    13. Sean Williams Boston College 511.3 20
    14. Jeff Green Georgetown 505.5 6
    15. Kyle Visser Wake Forest 503.5 57
    16. Herbert Hill Providence 503.0 49
    17. Javaris Crittenton Georgia Tech 492.2 18
    18. Wilson Chandler DePaul 483.1 30
    19. Julian Wright Kansas 481.4 11
    20. Daequan Cook Ohio State 470.0 27
    21. D.J. Strawberry Maryland 465.5 52
    22. Jason Smith Colorado State 464.9 17
    23. Alando Tucker Wisconsin 464.3 41
    24. Corey Brewer Florida 462.4 5
    25. Al Thornton Florida State 447.8 10
    26. Marcus Williams Arizona 445.8 33
    27. Acie Law Texas A&M 445.2 15
    28. Aaron Gray Pittsburgh 440.5 38
    29. Zabian Dowdell Virginia Tech 438.2 34
    30. Spencer Hawes Washington 433.9 12
    Other notables


    Morris Almond Rice 425.6 22
    Derrick Byars Vanderbilt 421.9 28
    Gabe Pruitt USC 421.0 21
    Nick Young USC 383.8 13
    Taurean Green Florida 350.4 39
    Arron Afflalo UCLA 336.1 32
    Ramon Sessions Nevada 334.7 37


    And based on the chart, let's close with some important conclusions heading into Thursday's big shindig:

    Kevin Durant is the best talent to come out of the college ranks in the last half decade. As we've learned, this doesn't necessarily mean he'll become the best player. But his 870.7 score blows the previous best -- Carmelo Anthony's 781.3 in 2003 -- right out of the water. If there's one thing that makes me reconsider the Oden versus Durant question after I thought it had been settled, this is it. I mean, how can you pass on this guy when his numbers are this overwhelming?

    Oden isn't the only no-brainer on his college team. It might surprise some to see Conley rated so close to Oden on this list. No doubt that's partly because Oden played all year with a bad wrist (again, an example of where scouting can augment what's being done with numbers). But among point guards in the last six drafts, only Chris Paul rated higher than Conley. Suddenly it doesn't seem so silly for the Hawks to take him at No. 3, especially given their glaring need at that spot.

    Four players are in a virtual dead heat for the No. 4 spot: Despite the red flag on his rebounding, Brandan Wright is in the mix here, along with Al Horford, Nick Fazekas and Thaddeus Young. Those last two names will probably surprise a few folks, since Wright and Horford have been the names we've been hearing the most. But Fazekas' numbers have been consistently outstanding -- particularly on the glass -- which makes him seem like Nick Collison with a jump shot. Meanwhile Young fared very well for a player his age, though one wishes his athletic markers were better.

    Jeff Green and Julian Wright don't rate as lottery picks. Though players with a score in the area of 500 usually are first-round quality and sometimes become very good, there are far better options to be had in this draft. Wright is a very poor shooter for his size, while Green's athletic markers were surprisingly average.

    Rodney Stuckey looks like the real deal. The unheralded guard from Eastern Washington actually had even better numbers in 2005-06 and rates as one of the 10 best players in this draft. Fans also might be surprised to see Duke punching bag Josh McRoberts so high, but he had the assist ratio of a point guard and fairly high rates of blocks and steals. Boston College's Jared Dudley finishes out the top 10 -- it's a bit troubling that he blocked only 10 shots all year, but he shot well from distance and has strong previous seasons helping him out.

    Corey Brewer rates way lower than most people would imagine. Brewer's rating of 462.4 makes him a marginal first-rounder, and that's only because of the paucity of Euros in this year's draft. Since this pretty much flies in the face of conventional wisdom, which has Brewer rated as a top-five or at worst top-10 pick, I have to point out that his numbers don't match the general consensus on his athleticism.

    Brewer's 8.5 rebound rate teetered on red-flag territory, and he blocked only 15 shots all season, which is quite low for a 6-9 NBA hopeful. Additionally, his ballhandling is a concern -- his -0.74 pure point ratio was pretty poor for a perimeter player. Brewer brings his share of positives too, but I'm not sure teams will get what they think they're getting if they take him high in the lottery.

    Acie Law and Spencer Hawes are best to be avoided. Both are viewed as late-lottery picks, but they look like solid second-rounders from here. Hawes has an unimpressive rebound rate, which is a huge red flag considering he was bigger than everyone he played against. And for all the talk of his great post skills, he had a run-of-the-mill 55.0 true shooting percentage and didn't even have the best PER on his mediocre team (that belonged to Jon Brockman). A lot of folks think he can become a quality pro post player; based on his numbers, I just don't see it.

    As for Law, his athleticism is a major question mark. He had 39 steals and one block, a pittance for a 6-3 guard. His 5.3 rebound rate is another poor indicator, while his age (22) only adds to the list of negatives. Additionally, Law's pure point ratio was only average, and he doesn't seem to be nearly dangerous enough a scorer to make up for it.

    The two USC guys look like total busts. There have been players to make the NBA with scores lower than those of Gabe Pruitt and Nick Young, and a couple of them even turned out to be halfway decent. But many more of the players who scored that low had disappointing pro careers, so clearly this is not the way to bet.

    Truth be told, Pruitt and Young both have major red flags to overcome. Pruitt's rebound rate of 4.9 is shockingly bad for a 6-4 guard alleged to have above-par athleticism, and he just didn't do much of anything else to set himself apart from the countless other guards trying to play their way into the league.

    Young is an even deeper mystery. If this guy's such a good athlete, how'd he have only 27 steals and 10 blocks this past season? As I mentioned above, his pure point ratio is hold-your-nose awful for a guard, his rebound rate is only OK, and it's not like he put up these numbers against stellar compe ion. Why exactly are we supposed to be excited about this guy?

    Actually, this formula doesn't seem to be kind to players from L.A. Arron Afflalo rated even worse than the two Trojans; Afflalo, Nevada guard Ramon Sessions, and Florida guard Taurean Green are fringe first-rounders who are best to be avoided.

  2. #2
    Still Hates Small Ball Spurminator's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jun 2003
    Post Count
    37,537
    Acie Law and Spencer Hawes are best to be avoided. Both are viewed as late-lottery picks, but they look like solid second-rounders from here.
    Let's hope so.

  3. #3
    Damn You Commies T Park's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Mar 2003
    Post Count
    54,855
    acie law will go into the second round the same day Luis Scola dons a Spurs jersey.

  4. #4
    5. timvp's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Feb 2003
    Post Count
    59,888
    I'm curious to see how much Hollinger adjusted backwards to meet his own beliefs about which players were the best and then created a system to prove that
    That's exactly what I thought after I read it. It's pretty easy to come up with a formula when you know what you want your results to be. I also give him a thumbs down for being vague when his system had an odd result, while he made sure to hype up the better results.

    Decent read but not as impressive of findings as Hollinger suggets.

  5. #5
    Jesus Loves UT IcemanCometh's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Apr 2004
    Post Count
    4,411
    if he was adjusting his system to provide the best results, how do you explain drew gooden, delonte west and luke jackson? wouldn't he have tried to get deron williams higher results?

  6. #6
    5. timvp's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Feb 2003
    Post Count
    59,888
    if he was adjusting his system to provide the best results, how do you explain drew gooden, delonte west and luke jackson? wouldn't he have tried to get deron williams higher results?
    There is only so much you can tweak a formula. Plus, Deron Williams had underwhelming stats in college. Unless he gave a bonus for going to Eisenhower, I don't think there is a way to make Williams college stats shine.

  7. #7
    Real Warrior Warlord23's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Dec 2004
    Post Count
    6,024
    Sweetney over Wade, Deron Williams at 11, Delonte West at 2? That doesn't really help his argument ..

    His system has Corey Brewer at 24, Almond and Young in the 2nd round, and that too without any foreign players considered. Interesting, but not happening

  8. #8
    Get Refuel! FromWayDowntown's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jul 2003
    Post Count
    19,921
    That's exactly what I thought after I read it. It's pretty easy to come up with a formula when you know what you want your results to be. I also give him a thumbs down for being vague when his system had an odd result, while he made sure to hype up the better results.

    Decent read but not as impressive of findings as Hollinger suggets.
    I think the test for Hollinger's system will, obviously, be what happens going foward from this point. If we look back in 5 years and Durant is pretty good, but not outshining Carmelo; and if Green, Brewer, Law, Thornton, or Hawes has turned into a pretty solid player (or if Fazekas, McRoberts, Stuckey, etc. have proven to be busts) Hollinger's system will be complete bunk.

    I'm not sold on the idea that one can meaningfully project collegiate players based on the statistics ac ulated by other players who've progressed through each level. College basketball is such a different beast than NBA basketball. As it is, Hollinger's system over-rewards for relatively meaningless numbers (steals) and seems to give too much credit to big guys.

    I just thought it was an interesting read for a variety of reasons. Honestly, I'm frequently intrigued by Hollinger's means of assessing players and teams, but I'm rarely sold on his methodologies or his conclusions.

  9. #9
    5. timvp's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Feb 2003
    Post Count
    59,888
    Honestly, I'm frequently intrigued by Hollinger's means of assessing players and teams, but I'm rarely sold on his methodologies or his conclusions.


    I think his PER stat is quickly becoming the most overrated stat in basketball. It's decent but has a lot of holes.

    I do appreciate that Hollinger has begun to help NBA fans transition to more of a use and understanding of numbers.

  10. #10
    Manu + SJAX = #5 50 cent's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jan 2004
    Post Count
    3,513
    I don't buy everything that he comes up with, but this type of stuff is very intriguing. I think he makes a great point about Afflalo and Acie Law. I'm an Aggie, and I just don't see him being that great in the NBA.

    It will be interesting to look at this 3 years from now.

  11. #11
    Hedo Layup Drill ShoogarBear's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jul 2003
    Post Count
    39,519
    This is a well-known problem is any type of modeling in the health and social sciences.

    You develop a model based on fitting it to past data, but just because your curve does a good job of fitting retrospective data in no way guarantees that it is a good predictor of future outcomes.

    There is a gigantic graveyard of curve-fit models that looked good when they were developed that proved to be useless as predictors.

  12. #12
    Hedo Layup Drill ShoogarBear's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jul 2003
    Post Count
    39,519
    I don't know the exact methods Hollinger used to develop his system, but what he SHOULD have done was develop the model using only a random selection of 1/3 or 1/2 of the draft picks, then tried to see how well it did at predicting the rest of the draft picks.

  13. #13
    Believe.
    My Team
    San Antonio Spurs
    Join Date
    Jun 2007
    Post Count
    5
    I don't know the exact methods Hollinger used to develop his system, but what he SHOULD have done was develop the model using only a random selection of 1/3 or 1/2 of the draft picks, then tried to see how well it did at predicting the rest of the draft picks.

    Here's a simple way to test it: seeing as how he only used the last six years to create the system, all he has to do is go back to previous years, plug everything in, and see if it still comes out.

  14. #14
    Believe.
    My Team
    San Antonio Spurs
    Join Date
    Jun 2007
    Post Count
    5
    I think people who look at things like Sweetney's relationship to Wade in the rankings are missing the point. The point is simply that Hollinger didn't miss anyone important. All the players who turned out good ended up giong in the first 15 picks or so.

    That's miles better than league executives pick.

    I think he's really onto something.

    My only problem is that Hollinger's rankings (and I'm mainly talking about PER here) don't do enough to take dfense into account. And I don't mean steals, I mean Bowen's kind of defense.

  15. #15
    Hedo Layup Drill ShoogarBear's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jul 2003
    Post Count
    39,519
    That's an option, but you want something that is robust now and for the future, not for the past.

  16. #16
    Veteran
    My Team
    San Antonio Spurs
    Join Date
    Oct 2006
    Post Count
    3,582
    I think that it is interesting. Considering the criteria, it is clear his model selects for good hand eye coordination, fast feet, and good reflexes, as well as aggressiveness and anticipation skills - this is where the weight on blocks, steals, and rebound rates categorized by height comes from. Age is obvious, in any sport (hockey, baseball, football, soccer), players tend to transition from being a boy playing against men to a man playing against men, and it shows in their increased productivity; the younger you are, the more likely you are to be before that transition. 3 pointers is a measure of jumpshooting skills, which is of course THE key scoring skill for guards - look at how much more effective Tony has gotten with something that resembles an effective jumpshot. Finally, his pure point rating is a good way of looking at court awareness, another key asset given the speed at which the NBA plays.

    I also find it flattering that of the people I think the Spur should be considering for their draft picks (ie players who will fall deep but I think will pan out), it agrees with me on Fazekas, McRoberts, Dudley, and it rejects, just like I would, Pruitt, Affaldo, Aaron Gray, Taurean Green.

  17. #17
    Veteran sprrs's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Aug 2005
    Post Count
    1,381
    I'm curious as to why steals are weighted so high. Is there a specific reason, or is he just using the logic that players with a high number of steals have succeeded in the past?

  18. #18
    Suppose there never was an Aaron? aaronstampler's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jan 2006
    Post Count
    2,821


    I think his PER stat is quickly becoming the most overrated stat in basketball. It's decent but has a lot of holes.

    I do appreciate that Hollinger has begun to help NBA fans transition to more of a use and understanding of numbers.
    Must be... Manu rates very well with it. I guess +/- is overrated too. Maybe we should just rate the players by Finals MVPs and get like a 400 way tie for 9th.

  19. #19
    5. timvp's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Feb 2003
    Post Count
    59,888
    Must be... Manu rates very well with it. I guess +/- is overrated too. Maybe we should just rate the players by Finals MVPs and get like a 400 way tie for 9th.
    Ah ... the best part of the offseason is only having to deal with aaronstampler once every couple weeks instead of every other day.

  20. #20
    Veteran
    My Team
    San Antonio Spurs
    Join Date
    Jun 2005
    Post Count
    2,156
    The Block:

    This should be split into two categories:

    1. The guy that blocks it into the cheap seats and the offensive retains the ball.

    2. The guy that has a softer type block, but usually results in his team getting possession of the ball.

    Give me the guy at nbr 2 everytime.

  21. #21
    You My Nikka Nikos's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Apr 2003
    Post Count
    3,314
    Although PER does have some holes, it is a much better stat then anything the NBA puts on their website. Efficiency? Efficiency per minutes? Eh, these are horrible.

    There is never going to be a model that even comes close to accurately predicting how players will fare in the NBA from college. Too many variables to consider, and a lot of it is chance anyway. But I do like Hollinger because at least he tries to put a value for player effectiveness and uses concepts like per minute and team context to level things out. Yeah its impossible to gauge defensive value -- but I think PER does a pretty good job on the offensive end for the most part. Much better than simple PPG, MPG, Team Wins etc... as most NBA fans seem to look at anyway.

  22. #22
    Drive for Five! ambchang's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Feb 2005
    Post Count
    17,577
    While going through his criteria, it appears to me like tall players that have a perimeter game (read Durant), young, and athletic would out rank a traditional lowpost centre in the Duncan mode. The reason I say this is because steals and 3pters seem to play a big role in the weightings, and rebounding is normalized with height, while blocks remain the only "big man" statistics where a perimeter based big man (like Nowitzki or Garnett) would be able to get a few blocks based strictly on their height.
    Maybe I have misinterpreted the criteria, but I would be very interested in seeing how Duncan, who is a traditional inside player who stayed all four years in college, would match up with a guy like Durant.

  23. #23
    Body Of Work Mr. Body's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Nov 2004
    Post Count
    27,701
    Like any system, it has its flaws, but can be highly illuminating. I'm not sure you should discount a guy like Pruitt, a PG, who has bad rebounding numbers, but it does shine a light on deficiencies.

    As has been mentioned upthread, it does show the Spurs in a positive light that they're looking at players who rate well in this system (whether they use it or anything similar). It also means well for us that several of the good players will be available around the changeover of the rounds.

  24. #24
    Suppose there never was an Aaron? aaronstampler's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Jan 2006
    Post Count
    2,821
    Ah ... the best part of the offseason is only having to deal with aaronstampler once every couple weeks instead of every other day.

    Whatever. You lost all credibility in my book the second you said that LeBron James and Michael Finley are better defenders than Manu. I don't know how anyone can take you seriously after that.

  25. #25
    The GodFather Vito Corleone's Avatar
    My Team
    San Antonio Spurs
    Join Date
    Dec 2004
    Post Count
    3,715
    By that Formula, Shawn Bradley would have been the #1 pick when he came out he averaged 6 blocks and about 14 rebounds to go with his 12 points as a freshmen at BYU.

    If the 2k6 draft were to happen today I doubt anyone is taken above LaMarcus Aldridge.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •