Skip to content

Some Past Results

March 21, 2013

I’ve had a lot of people asking me about previous years, etc. of my draft model, so I thought I’d post a couple of things. Here’s the top twenty prospects from last year  and the top ten at each position overall. Remember that only 2010, 2011, and 2012 are out of sample. Also, please keep in mind that this model is ever-evolving and these numbers are subject to change in the future, especially as more data becomes available.

Top 20 from 2012:

Only Davis, MKG, and Drummond were better than +1, and only Davis was better than +2.

1. Anthony Davis
2. Michael Kidd-Gilchrist
3. Andre Drummond
4. Jared Sullinger
5. Bradley Beal
6. Dion Waiters
7. Maurice Harkless
8. Damian Lillard
9. Terrence Jones
10. Harrison Barnes
11. Tyler Zeller
12. Meyers Leonard
13. Quincy Miller
14. Jae Crowder
15. Royce White
16. Kendall Marshall
17. Thomas Robinson
18. Jeremy Lamb
19. Draymond Green
20. John Henson

And here are the top tens by position:

Top Ten Point Guards Year Projected Impact
Kyrie Irving 2011 4.9
Chris Paul 2005 3.5
Derrick Rose 2008 3.2
Jrue Holiday 2009 3.0
Russell Westbrook 2008 2.8
Mike Conley 2007 2.7
Stephen Curry 2009 2.0
Kyle Lowry 2006 1.9
John Wall 2010 1.8
Raymond Felton 2005 1.8


Top Ten Wings Year Projected Impact
Kevin Durant 2007 5.6
Carmelo Anthony 2003 2.9
Marvin Williams 2005 2.7
Luol Deng 2004 2.7
James Harden 2009 2.4
Dwyane Wade 2003 2.3
Tyreke Evans 2009 2.2
Danny Granger 2005 2.1
Xavier Henry 2010 2.0
Rudy Gay 2006 1.8


Top Ten Bigs Year Projected Impact
Anthony Davis 2012 3.4
Kevin Love 2008 3.2
Greg Oden 2007 3.0
DeMarcus Cousins 2010 2.7
Blake Griffin 2009 2.5
Andrew Bogut 2005 2.4
Chris Bosh 2003 1.9
Hasheem Thabeet 2009 1.8
Michael Beasley 2008 1.8
Derrick Williams 2011 1.6

Otto Porter looks like the second best wing prospect in the last ten years. And Marcus Smart, Nerlens Noel, and Trey Burke each make the top ten for their position. It’ll be very interesting to see how these guys pan out.



Shabazz Muhammad: The Second Coming …of O.J. Mayo

March 18, 2013


The NCAA stirred controversy earlier this season when it initially ruled Shabazz Muhammad, the best prospect in the nation, ineligible due to slave amateurism violations. This all added to the Shabazz hype – at the time, he sat alone atop most mock drafts for 2013. O.J. Mayo, too, was a hype machine – one of those rare household names before he was even in college (at least, if your household is full of huge basketball fans). I found the first minute of this video particularly amusing : “is this guy as good a prospect as LeBron James?” “Yes.” But despite being (allegedly) illegally paid to go to prominent Los Angeles schools, neither O.J. nor Shabazz seemed to live up to the hype in college, especially when you venture beyond their scoring totals.

With a quick glance at the box scores, you might think Shabazz is everything he was supposed to be: he’s averaging over 18-per-game and his team finished the regular season on top of the Pac-12 standings. Just the other week, in fact, the beat writer for my beloved Washington State Cougars observed the following:

But when you look past the scoring, it becomes apparent that Shabazz is essentially one-dimensional. Specifically, of the wing players who have a shot at being drafted this year, Shabazz falls in the bottom 20% in steals, and the bottom 10% in defensive rebounds, assists, and blocks.

Ignoring the other numbers for now, Shabazz’s assist numbers are particularly alarming. We can sit and argue the value of the subjective assist stat all day, but I’m confident we can all agree that it has value. And it becomes especially important for guys, like Shabazz, who use a lot of possessions because good scorers will get double teamed in the NBA, and when they do, they need to be able to find the open man. Shabazz’s assist rate is 5.8%. That means less than six percent of his teammates’ field goals are assisted by him when he’s on the floor. That’s astronomically low, especially for someone who is supposed to be a great offensive player. To illustrate this point, here is a list of all guards who have had at least ten win shares in an NBA season while maintaining an assist rate of six or less. (there’s not an error, there are zero guys on the list.) Ha! Ok, let’s give some leeway: here’s the same list, but with the minimum assist rate moved to ten.

The list is nine seasons long. Three of them belong to Peja Stojakovic, who was basically an elite role player – one of the best spot up three shooters the league has ever seen. One belongs to Dale Ellis, who was probably a bit more offensively versatile than Peja, but was still primarily a spot up shooter. Both were good scorers, but were always surrounded by other good offensive players – Peja had C-Webb and was on Kings teams where everyone was basically a scoring threat; Ellis had Xavier McDaniel and Tom Chambers. In contrast, Shabazz Muhammad is supposed to be a primary offensive threat – someone who you can give the ball to and let him go to work. Plus, while he’s a good shooter, he’s not near the level of Peja or Ellis, and would likely have trouble filling roles like theirs in an offense.

Two of the nine seasons on the list belong to Chet Walker and one belongs to Doug Collins. These guys played (at least with respect to the seasons in question) before we measured steals, blocks, or turnovers; the “10” win shares are much more an estimation than they are for the other players on the list. Then we have Marques Johnson and Adrian Dantley.  Whether or not these guys actually even played “guard,” their style of play is not even remotely similar to how we see guards play today. And so it’s difficult if not impossible to compare Shabazz to any of these guys. As a result, it’s apparent that Shabazz’s current inability to assist baskets ain’t gonna fly in the big league. It almost necessarily puts him in a role-player box, which is fine, but isn’t what you want from a super-high draft pick. And it’s certainly not what you want to see from your primary option on offense, especially with today’s sophisticated defenses that will undoubtedly force Muhammad to make tough passes in certain situations.

Ok, now let’s talk about Muhammad’s D. While defensive numbers are just a small part of measuring a player’s defensive contributions, they nevertheless matter. I have already noted that, out of the wings drafted in the last ten years that stole the ball at 1.5 times per pace-adjusted 40 minutes or less, none have been all-stars. Shabazz is below 1.0. Things begin to look even worse when we look at his poor shot blocking numbers. Sure, shot blocking isn’t particularly important when we’re evaluating shooting guard/small forward types, but it can be indicative of defensive effort and ability. And just like steals, history is not very kind to wings who can’t block shots in college. Specifically, if we look at wings who blocked 0.4 or less shots per pace adjusted 40 from 2002 to 2012 (Shabazz blocked less than 0.2), none  have been all-stars, and the best of the bunch have only been marginally successful (Rodney Stuckey, Arron Afflalo, and Kevin Martin, for example), and certainly unable to be the best player on a good team.

And believe me, I realize that talking in abstract concepts like this can be silly. Despite history, which by the way only goes back ten years, a shooting guard’s blocks or steals per minute shouldn’t sway a team whether to take the guy or not. But having poor defensive stats all around – and I can do the same exercise with his poor defensive rebounding – does raise serious questions about Muhammad’s defense. Maybe someone with access to Synergy can shed more light here, but since Shabazz’s quickness and athleticism are questionable – and if you haven’t watched him play, believe me his quickness and athleticism are questionable (julienrodger from A Substitute for War has written a couple of good articles on the subject – look here and here), I’m not sure he wouldn’t be a defensive liability in the NBA.

So when you put it together you have a skilled scorer who is crafty though not particularly athletic, who couldn’t find the open man if he had five guys guarding him, and who has provided no evidence to suggest he’s even an average defender. This just doesn’t sound like a top five prospect or a game-changing star/primary option.

Of course, like with Ben McLemore, I’m not suggesting that you shouldn’t take Shabazz in the first round. I’d probably even take him in the lottery. He’s a good and versatile scorer. He hustles, and he steals a lot of boards on the offensive end. He obviously has a lot going for him, and I think he could be a decent to solid NBA player. Hell, O.J. Mayo is a decent to solid NBA player. But he never lived up to his superstar hype. And I’m not so sure Shabazz will either.


Underrated Draft Prospects: Trey Burke

March 13, 2013

trey burke

Burke certainly is an interesting prospect. Everyone and their mothers recognize him as the best point guard in college, he’s a sophomore by the way, and yet all of the big time draft sites peg him as a mid-first-rounder. So where’s the disconnect?

For one, Burke is pretty short. Now I’m not going to try to sell you the size doesn’t matter bullshit like I did last year, because size does matter, especially for wings and especially on defense. But size is less of a limitation for point guards than it is for other positions. And this is demonstrated quite convincingly by the size of some of the league’s best point guards. Chris Paul, of course, is one of the three best players in the league at six feet even with shoes on. And it doesn’t end there: Rajon Rondo, Mike Conley, Ty Lawson, and Kyle Lowry are elite at the position – and they’re all under 6’2″. Even Jameer Nelson, Raymond Felton, T.J. Ford, and Nate Robinson have had reasonable success in the NBA. And Burke has a case for being a better offensive college player than any of these guys. He has almost certainly been the best scorer of the bunch – shooting a 59% true shooting percentage while averaging more points per pace-adjusted 40 than all but Nelson (his senior year), who played against weaker competition. Plus Burke is on pace to average more assists per pace-adjusted 40 than every player on that list except T.J. Ford. Factor in turnover rate – Burke’s is the best of the bunch – and what we have is an efficient, finely tuned offensive weapon who is still only 20.

Though he is not a superb vertical athlete, Burke is extremely quick and fantastic at handling the ball. This allows him to create space and penetrate at a very high level, and because of his court vision and elite passing, he can find the open man when defenses collapse on him. It is difficult to find a comparison to Burke because of his offensive versatility. Chris Paul was not the scoring threat in college that Burke is, though he developed into a great scorer with time. Paul separates himself from Burke on the defensive end though, where his instincts were much better than Burke’s are. And this is where Burke’s critics are the harshest.

Burke steals the ball at a respectable rate, but his on-ball defense has been criticized as sub-par for an NBA prospect. However, while defense is unquestionably important, it is less of a factor for perimeter defenders than for interior defenders. I’ll expand more on this in a future post, but a bad defensive center is generally much more detrimental to a team than a bad defensive point guard. And I don’t think anyone is calling Burke a bad defender, only a sub-par one. Besides, quantifying defense by observation is tough and there is plenty of room for error. Either way, it is very doubtful that Burke’s shortcomings on the defensive end even approach canceling out his sensational offense.

My draft model, of course, projects Burke as a +2 in the league, which puts him in elite company, and suggests that he’s the fourth best prospect in the draft – and the second best point guard. As much as I love the top point guard prospect, Marcus Smart, his offense is just not on par with Burke’s at this point. As I’ve already noted, Burke’s quickness and ballhandling allows him to go just about anywhere on the court any time he wants. And this ability is amplified by his extraordinary jump shooting. Just a few weeks ago, Jonathan Givony tweeted that Burke was statistically the best off the dribble jump shooter in college basketball. Burke’s deadly shooting is the final piece in a combination that makes him a constant threat with the ball as soon as he crosses half court. And though his defense is pretty far behind his offense, it’s not far enough behind where I’d wait until the mid-first-round to take him. There are a very limited number of players in any given draft that will pan out in the NBA. Burke looks to be one of them – he crosses my +2 threshold and he’s one of the best players in college basketball as a sophomore. It would be a shame if he slipped past the top ten in this year’s draft.


I Wouldn’t Take Ben McLemore Number One If I Were You

March 8, 2013

Marcus Smart, Ben McLemore

Don’t get me wrong. I think McLemore is a solid wing prospect. He’s a fantastic athlete and his shooting stroke is great. But I’m just not sure he has the characteristics of a player that a team wants to spend its coveted first overall pick on. And I’m gonna tell you why.

This Draft Isn’t As Weak As the Pundits Would Have You Believe

The more time passes, the more I hear how “weak” this draft is. I hear people make this assertion every year, but this year the notion has become particularly popular. Chad Ford has been saying it for a while, and I think the more people say it, the more others buy in. Just the other day I heard Jay Williams claim that it’s the weakest draft in the last twenty years. Yet, somehow I can’t imagine that Williams would be picked second in this draft. Regardless, the “weak” draft claim is the reason why McLemore has a good chance of going number one. “Oh, he’s not as good as your typical number one pick, but it’s a weak draft.” I simply can’t get behind this.

According to my projections, this may be the best draft since 2009. It’s still early, and some lottery projected players could still certainly take their names out and decide to stay in school, but consider this: my model projects six players this year as being +1 or better in the NBA. Compare this to the three players from last year’s draft, just two from the previous year, and four from 2010. You have to go back to 2009 (7) to find a draft that had more guys projected at +1 or better than this year.

And it’s even more impressive that I’m projecting four guys at +2 or better. This has only happened in three drafts since 2002. Sure, maybe this changes by the end of June, and maybe this isn’t one of the strongest drafts in history, but the point is this: there just isn’t any evidence that this draft is as “weak” as people are claiming. As a result, picking McLemore – who is a solid shooter and may develop into a good offensive player – first overall just because it’s a “weak” draft is not just silly, it’s flat out uninformed.

McLemore’s Size-Skillset Combination Poses Legitimate Issues

McLemore, who is 6-5 on platform shoes, can really only play one position in the league – shooting guard. He is simply not big enough to guard opposing small forwards today’s league. And he can’t play point guard because he is below average at ballhandling and penetrating and his court vision is poor. In general, one-position-only guys are only particularly valuable if they’re point guards who run offenses and centers who anchor defenses. Think about it, how many great NBA players can you think of that could only play shooting guard and weren’t great with the ball? Jordan and Kobe were both ball dominant players with an ability to create good shots in practically any given possession. Jordan proved he could play point guard in 1989, and both he and Kobe could swing to the three position if their team needed them to. Wade could play point – and did early in his career. But Wade isn’t a good comparison either anyway – unlike McLemore, Wade’s strengths have always been slashing and either finishing or finding the open man. Same goes for James Harden – a ball dominant guard whose value primarily lies in his ability to get to the basket. McLemore has not demonstrated this kind of ability. To compare, Harden and Wade at McLemore’s age both averaged well over twice as many assists per pace-adjusted 36 as McLemore while maintaining a significantly higher usage rate.

So who does that leave? The only comparison that’s even partly legitimate is the guy I guess everyone is comparing Ben to: Ray Allen. But is McLemore Ray Allen? I’m not so sure. For one, Allen was a much more prolific scorer than McLemore, even at the same age. Allen’s ability to create good shots for himself seems to have been much more advanced than McLemore’s. But even if McLemore does become Ray Allen (and I don’t think he will), is Ray Allen the guy you want with a #1 pick? Yeah, he led some good Milwaukee and Seattle teams, but he is a player better suited to be a second or third option.

McLemore’s position limitations are even more problematic when we look at some of the bottom teams – who are most likely to wind up with the top pick. New Orleans just gave up the best point guard of the last decade for Eric Gordon and used its lottery pick last year on a 19-year-old shooting guard. Neither of these guys can play the 3. Sacramento has 15 shooting guards. Washington and Cleveland used their #3 and #4 overall picks, respectively, last year on small shooting guards. So that leaves Orlando and Charlotte. Both these teams could use a shooting guard, but would both be smarter to take a different position. Orlando starts Arron Afflalo (also small) at the 2, and while he’s certainly not great, he might be their strongest starter at this point. Charlotte would be better off going after a big or a true point guard – two areas where the team lacks big time. A guy who can only play shooting guard just doesn’t meet any of these teams’ needs. And yeah, team needs go out the window if we’re talking about a big time player, but remember, “he’s not as good as your typical number one pick, but it’s a weak draft.”

He Isn’t As Young As You Think

Ok, you might think I’m just being nitpicky, but this kind of stuff matters. I’ve said before age is one of the most (if not THE most) important factor in predicting a college player’s future success. Hell, I even revised my model to make sure my values for age were exact to the day. So when you hear that McLemore is the highest scoring “freshman” in college basketball, take it with a grain of salt. McLemore is older than, among others, Anthony Davis, Michael Kidd-Gilchrist, and Bradley Beal (the top three picks of last year’s draft).

He Doesn’t Steal the Basketball

Now this just seems silly. But I’m not so sure it is. Whether the number in the steals column measures hustle, athleticism, a general feel for the game, a combination of the three, or something else, it is highly indicative of future success for wings. And when we look at the results, it’s pretty telling. McLemore averages 1.4 steals per pace-adjusted 40 minutes. When we look at all the players who were drafted between 2002 and now who averaged 1.5 or fewer steals, very few have been successful in the NBA by any player measure, and none have been all-stars (or, if you don’t like the all-star measure, none have been better than above average players). That’s not to say that a player can’t be good if he doesn’t rack up the steals in college, I’m just saying that history isn’t on Ben’s side here.

Who I’d Take Instead

It wouldn’t be fair to McLemore or the reader if I didn’t at least offer some alternatives. In my mind, there are three reasonable choices a team could use its number one pick on: Nerlens Noel, Marcus Smart, and Otto Porter. Each of these three guys plays a different position, and each of them boasts a very unique skillset. Oh yeah, and they all project as +2 or better in my model (if you recall what I’ve said before, +2 is basically the threshold where a player is all but guaranteed to be successful in the NBA if he crosses it).

Nerlens Noel

Yes, Mr. Noel had a knee injury. But I’m still confident he will have a solid NBA career and I think he’ll be able to help a team’s interior defense immediately once he recovers. Noel leads the entire NCAA in blocks per game (as a freshman!) and he’s second in the SEC in steals per game (as a center!!). He’s a solid rebounder as well. The one nitpick, besides his ACL injury I guess, is his offensive game, which is miles behind his defense. He is sloppy with the basketball and very unpolished in the post. His jumper is basically nonexistant: he shot 37% on his jump shots in college. But despite offensive struggles, his defensive prowess – and particularly his ability to protect the rim – is very intriguing. At the risk of sounding cliche, offensive skills can be taught. Defensive instincts can’t. Having an elite interior defensive presence can change a team’s fortunes quickly. And when we consider what Noel is – a giant teenage super-athlete – I think we should be willing to look past his hiccup of a knee injury, assuming of course  his recovery continues to go as planned.

Marcus Smart

Smart has been on top of or very close to the top of my draft rankings since I first put them out last month. The reasons are simple: he’s 18, he’s built like a brick shithouse, and he’s really good at basketball, specifically with respect to the aspects of the game that reflect hustle and athleticism. Of the point guards in the draft discussion this year, Smart is the best at rebounding (and is particularly good on the offensive glass), the best at stealing the ball, and the best at shot-blocking.  My favorite way-too-early comparisons are Westbrook and Wade – where Smart loses ground in athleticism he makes up for it in strength. Smart is very good at getting to the rim and he’s especially good at posting up – with his size there are few if any point guards who can guard him in the post. Smart’s primary weakness appears to be his jump shooting. But again, he’s 18 and there’s lots of time to work on this – he’s not that bad at shooting. Smart could also become a better passer, which I think will happen. His instincts and court vision are there, he just needs to polish his decision-making. Perhaps the most intriguing thing about Smart  is his reputation for being a leader and willingness to fill that role. With everything he brings to the table, I’m taking Smart first if I need a point guard for the future. If I need interior defense I’m taking Noel. And if I need a solid all-around wing, I’m taking…

Otto Porter

Porter was able to fly under my radar for a few weeks until I fixed an error I had in my data. Then he immediately shot up to the #4 prospect in my top 100. Then he started playing out of his mind, including a game where he scored 58% of his team’s points in a win over Syracuse. Now he’s my #1 prospect. He has a fantastic size-skill combination, which will cause nightmare matchups for teams with small small forwards or slow power forwards. He’s an underrated shooter – he shoots 45% from three (!) and he’s basically a scoring threat from anywhere on the court. But his skills don’t end with his shooting: he is a very good rebounder, he is quite adept at forcing turnovers (2.4 pace-adjusted stl/40) and blocking shots, and he’s exceptionally good at taking care of the ball. I think an old exercise that I used to use quite frequently would be particularly enlightening here. The following table compares Porter’s numbers with McLemore’s:

Player Age Height points TS% ORB% DRB% TRB% AST% STL% BLK% TOV% USG% impact
Porter 20.1 6’8″ 20.2 0.610 7.1 18.9 13.4 18.0 3.6 3.3 10.3 24.5 12.4
McLemore 20.4 6’5″ 20.2 0.636 5.6 12.5 9.4 13.5 2.2 2.5 13.0 23.6 9.3

Compared to McLemore, Porter is younger, bigger, and better at just about everything. McLemore is a bit more efficient from the field, but his usage is a bit lower. And just about all of Porter’s other numbers are substantially stronger. In other words, the two are comparable scorers, but Porter is better at every other facet of the game. Plus he’s younger and his size allows him to be much more versatile both as an offensive threat and as a defender who can check multiple positions. In fact, Porter’s production and versatility make him the best wing prospect in this draft – my draft model rates him as the best player. So if I were a team in need of a wing, I’d take Porter first overall.

Again, I don’t mean to get too down on McLemore. I think he’s a very good prospect. I also think he could be a very good NBA player. I just wouldn’t use a number one overall pick on him. I’d go for Noel, Smart, or Porter instead.


Draft Rankings!!

February 4, 2013


After a lot of messing around with the numbers and regressions, I finally have something that I feel comfortable releasing. Of course, I’m obsessive by nature, so this certainly won’t be the final iteration, but I’m pretty happy with it at this point. I posted on my methodology a couple of weeks ago, and the changes have been relatively minor, with the biggest ones being a) using all college players drafted regardless of whether they actually played minutes (bigger sample size – I gave the guys who never or barely played a -5), and b) separating point guards from wings from bigs. But this has helped tremendously.

The results are far from perfect as they probably always will be – remember what we’re doing here, we’re taking stats from college kids playing against varying levels of competition in a very limited sample size and trying to project their careers. But I feel pretty confident that we can make educated guesses with this data – and do a much better job than what we’ve actually seen in the past.

When I observe the results (again, I have the data for all drafted players who played in college from 2002 to present), the thing that stands out the most is that players who project as +2s are pretty much can’t miss. 23 players over the eleven year time span have projected as +2 or better. Of the 23, thirteen have been all-stars (Dwyane Wade, Chris Paul, Carmelo Anthony, Kevin Durant, Chris Bosh, Derrick Rose, Russell Westbrook, Kevin Love, Luol Deng, Danny Granger, James Harden, Jrue Holiday, and Kyrie Irving), five have been at least solid NBA starters (Mike Conley, Kyle Lowry, Andrew Bogut, Ronnie Brewer, and Raymond Felton), and three are still under 23 years old and all have a good shot at being all-stars before their careers end (Anthony Davis, DeMarcus Cousins, and Derrick Favors).

That leaves one. Easily the most enigmatic of the bunch, this player is either the one lemon on a pretty solid list or could have been a great NBA guard. Ever heard of Nick Calathes? He played two years at Florida (averaging 17-5-6 his sophomore year), then opted to go pro in 2009. But not the NBA. He committed to play in the Greek league before the NBA draft and signed a three year-2.4 million euro contract. He was still drafted, but hasn’t made it to the league yet. It seems that the Mavs have his rights. He’s still young so maybe we’ll hear something from him. Or maybe he was just a lemon. It’s hard to know at this point.

When you look at the projections, I think there are a few general rules that you should follow in decision making (always keeping in mind that nothing is for sure here):

  1. if a player is worse than -1, you probably shouldn’t waste your pick on him. Very few of these players have had NBA success – it’s not impossible, but the risk in taking these guys rarely pays off.
  2. if a player projects between -1 and 1.9, take him, but with caution. You want to make sure this guy fits your system and what you want. These guys have the highest variance – most are decent, but some will fall short.
  3. if a player projects at 2.0 or better, take him. He’s as close to a sure thing as you’ll see.

So that brings us to this year. I have posted the rankings here. You’ll notice the permanent link at the top of the site. Some early observations. If I need a point guard, and even if I don’t, I’m looking hard at Marcus Smart right now. The guy does everything, he’s a brick shithouse, and he’s 18 years old. He could be scary good. If I need a big, I’m taking Nerlens Noel first chance I get. Most analysts agree that Trey Burke is the best point guard in the NCAA right not, but few have him as a lottery pick – I disagree, maybe his draft stock will rise as time passes, but this guy has got to be taken in the lottery when it’s all said and done. I’m really scared of Shabazz. He doesn’t project well and I’m beginning to think he’s just overhyped. I could be wrong, but I don’t know if I would take this guy top 5 at this point. Other than that, I’m sure there are a million stories to be told in there. Keep coming back, I’ll update these rankings weekly. Enjoy!


NBA Draft Projection Model and More

January 14, 2013


Edit: I’ve substantially changed the model since this post. I have also settled on using just height for the time being because any added accuracy by measurements is minimal. For more information, see this post.

The yearly draft is one of the most captivating things about NBA basketball. Part of its lure is its unpredictability. Throw in the stakes – pretty high for many struggling franchises – and we can’t help but watch with wide eyes and watering mouths. And while we’re still a long way from being able to predict players’ performance with great accuracy (30 college games is a small sample by anyone’s measure, and the inconsistent level of competition makes it harder), it’s fun to try. So I’ve been working on developing a model that attempts to project NBA performance of college players. Basically I ran multiple linear regressions on pretty much all the data I could collect (like pace-adjusted box score numbers, team sos, measurements, etc.) for college players from 2002-2008 who have played at least 2,500 minutes in the NBA (Edit: I have now included all drafted players) against their career NBA RAPM. I of course cut out unhelpful factors, and eventually came up with a fairly reasonable predictor. It’s far from perfect though, and I’ll continue to work toward accuracy. It’s better at projecting who will be successful and who won’t than it is at predicting exactly how good a player will be – and it’s really good at predicting who will perform poorly in the league. Edit: I want to point out that the model in its current form is much more accurate – and is probably best at predicting who will be an all-star caliber player in the NBA.

It’s worth noting that the model is more accurate when we add in players’ measurements – particularly standing reach and wingspan – that are available at Draft Express. But these measurements usually aren’t available, at least for many players, until right before the draft. So for now, I’ll rate some players based on a regression without those measurements – which still does a pretty reasonable job. For those interested, I have found (and I think this confirms something John Hollinger has said in the past) that the single most important predictor for NBA success is steals. For a little sneak peak, here are the players from Draft Express’s top 25:


It’s early, but it looks like Nerlens Noel is THE guy this draft class – something I have suspected for some time now – and I wouldn’t be remotely surprised if he went #1. Kentucky’s other freshmen aren’t nearly as intriguing, which is also reflected in my NCAA Estimated Impact. Needless to say, this is the reason the team is struggling to meet its lofty expectations. Elsewhere, this could be a good draft for teams that need point guards – Carter-Williams, Smart, and Burke all project pretty well at this point.

In other, probably less exciting news, I’ve updated NBA Impact by introducing pure RAPM as part of the equation, so this stuff (and check it out, I update the current NBA Impact pretty much daily) is much, much  more accurate now. While we’re on the topic, by the way, just as I felt obligated to point readers to xRAPM when J.E. first put it out, I feel obligated to point you to IPV (Individual Player Value ratings). This is very similar to xRAPM – the primary difference is the SPM prior: IPV places much more emphasis on usage (particularly usg%*ast%), which the SPM prior in xRAPM ignores. The creators seem to have certainly done their due diligence, and the result is a very accurate predictor of player performance.

Anyway, stay tuned, I’ll have much more on the draft in upcoming months.


Is Miami’s Defense Struggling Because of Small Ball?

January 2, 2013


Last year the Heat were a +4 on defense. This means per 100 possessions Miami held its opponent to four points less than the average NBA team. This year they’re a -1. That’s a five point per 100 possession relative to league average swing. Now I know that may not sound like a lot, but it really is. Consider this: the best teams in the league right now (the Spurs, Clippers, and Thunder) beat their opponents by about 9 points per game. The Heat beat their opponents by about 5. That means if Miami’s defense was as good as last year, the Heat would be the best team in the league (by point margin at least). If Miami’s defensive efficiency dropped by 5 more, you’d expect the team to have a losing record. This all isn’t to say that the Heat won’t win the championship or isn’t the team to beat in the NBA, but it certainly explains the team’s early season struggles.

But why is the defense so much worse? The easy answer is because the Heat, like many great teams that won the previous season’s title, are coasting – merely waiting for the playoffs to strike. The most salient example of this kind of team is the ’01 Lakers. The 2000 Lakers had the best defense in the NBA (+5), and with a pretty good offense, they were a +9. By contrast, the 2001 squad, while remaining one of the league’s elite offenses, dropped to nearly a -2 on D! Glen Rice wasn’t an elite defender. No, in the playoffs, the Lakers turned it on and went back to being the league’s best defensive team, this time a +6 (in the playoffs!!), and went on to win the title. But is this the case with the Heat? If it is, who do we blame? It was easy with the Lakers, while it was probably a bit of a team lack of effort, most of the blame can be pinned on Shaq, who admitted to coasting, and it reflected in his numbers. But the Heat’s superstar, LeBron James isn’t lacking effort – whether you watch him or observe his numbers, you can see LeBron is playing as hard as ever. D-Wade has been struggling, but he’s not a defensive anchor, so it’s hard to pin this kind of a swing on him.

The better answer is small ball. The Heat have canned traditional 2-big lineups in favor of a lineup where Chris Bosh plays center. And this has worked – they won the championship last year doing this (as a side note, one of the big reasons the small lineup worked in the finals was because of its strategic advantage – having a 4 who could play outside drew OKC’s defensive anchor to the perimeter and minimized his impact). But it might be taking its toll this regular season, at least on the defensive end of the floor. When we look at Miami’s minute distributions, we can see that most of its key players are playing the same percentage of the team’s minutes as last year. But with the additions of Ray Allen, who plays a big portion of the team’s minutes, and Rashard Lewis, whose minutes are at least substantial, someone had to have taken a hit. Well, two Heats took notable cuts in minutes compared to last year: Udonis Haslem and Joel Anthony. When we look at their long term RAPM, we see that both Haslem and Anthony are negatives overall because of poor offense. But both players are positives on defense. Now while RAPM certainly helps, basketball isn’t plug and play (as much as we’d like it to be). Lineups are important, and mixing in a couple of defensive minded guys with an already stacked offensive squad could go a long way, especially considering neither player uses many possessions on offense. When we look at this year’s lineups, we see that replacing Bosh with Haslem or Anthony (even if the sample size is small) has made a considerable difference on defense. But Haslem is smaller than Bosh you say? Well, maybe it’s not small ball, maybe its more having-no-bigs-who-focus-mainly-on-defense ball. This year, lineups featuring Haslem or Anthony as center have performed better on defense than lineups featuring Bosh at the 5. But even this is far from conclusive: Anthony barely plays at all and Haslem usually is on the floor with Norris Cole, who is a whole other story altogether. Hell, if the Heat didn’t play Cole at all, we probably wouldn’t have to worry about any of this. Regardless, it would be interesting to see if Miami’s defense would improve by playing either Haslem or Anthony (not both, we don’t want the offense to suffer too much) more alongside Bosh/Battier/James in the frontcourt.