Follow up to: "Want to Make it in the NBA?" Article. Re: Duke vs. UNC

Follow up to: "Want to Make it in the NBA?" Article. Re: Duke vs. UNC
Nov 12, 2006, 02:54 pm
Original Article


There have been many great college basketball programs over the years; UCLA, Kansas, Kentucky, Duke and Indiana have all had their day. Some have been great for years, even decades at a time. When it comes to impact on the NBA, however, one program clearly stands apart from the rest. Two schools, The University of North Carolina at Chapel Hill and UCLA, have developed the most quality NBA talent over the span of the 60 years of league history, and over the last 25 years the Tar Heels have left the Bruins far behind. Carolina has had more first round draft picks than any other program (beating out archrival Duke for the honor), and the second highest number of draft picks overall (behind UCLA). Those draft picks have gone on to consistently meet or beat expectations for their performance in the NBA. As icing on the cake, Carolina has also had more former players develop into winning NBA coaches than any other college program, and in this regard there is not even a close second.

Email Response:

Dear Sir,

I’m not sure if you envision your site as a legitimate media outlet or not. If not, then I guess that pseudo-scientific opinion piece on UNC-Duke NBA players is acceptable as simply a rambling bit of unscientific rhetoric. But if you actually see yourself as a legitimate source of unbiased information, how could you possibly print that article?

How can you print an article in 2006 that relies on statistics that only go through 2001? Did it occur to anyone how easy it would be to apply that ranking system to the years since 2001, or did the authors/editors intentionally ignore those years so that they could hide the bias of the article? Clearly, Carlos Boozer and Chris Duhon meet the criteria for “hidden gem” status now, don’t they? How can you make mention of UNC’s 2005 draft class, but then fail to mention anything about the NINE players that Duke has put into the league since the end of your “statistics” in 2001… players who have ALL met or exceeded expectations by your own criteria? A responsible, respectable media outlet might at least provide some kind of qualification for the fact that Duke’s greatest period of NBA success has come after the endpoint of your pseudo-statistical analysis, but I guess DraftExpress doesn’t fit that category.

I would also ask how it’s possible that someone could manage to rank Michael Jordan and Vince Carter as “hidden gems” despite being the #3 and #5 picks in the draft, while not ranking Grant Hill as a “hidden gem” despite being named to the All-Star team in every single full season he’s played since coming into the league. I would ask that, but I already know the answer. By your own criteria, Grant Hill was expected to be only a “solid starter” by his draft position (#3, same as Michael Jordan), but he was a superstar “top 10-15 player in the league” prior to his injury. That’s your criteria, not mine. Clearly you only apply that criteria if it suits your purposes.

Finally, how can you have the audacity to print that Bobby Hurley was a bust? That’s not just poor journalism. That’s poor taste, even with the qualification pasted in afterwards. The kid almost died.

I’m shocked by the lack of professionalism displayed in that article.


-Steve McSwain

Paul Gearan’s response:

Mr. McSwain’s main critique is about the 5 year outcome data (Table 2), and as with previous articles I stand by the fact that to get a decent measure of outcome you need 5 years, especially in an era where so many players come out early, so in essence it is 2006 NBA outcome data with the last draft class being 2001. We applied that criteria evenly across the board. But that was only one of three elements of the article, and the other two were clearly all the way up through 2006, so to claim this article “only” relies on data through 2001 is not accurate. Also, in this Table 2 analysis, these were all first round picks, but, yes, Boozer and Duhon probably deserve special mention post-2001 draft as 2 of the best second round picks, and in retrospect would have been good mentions as an aside like the Carolina 2005 draftees, but because they fell out of the analytic scope they did not immediately come to mind. Mr. McSwain makes a valid observation there.

However, Duke's "greatest period of NBA success," according to Mr. McSwain, since the 2001 draft also yielded Mike Dunleavy who at #3 in 2002 would have to be considered a mild disappointment at best now entering his fifth season, Jay Williams, another accident victim whose career has sadly been altered, so we'll likely never know what he could have been, and Dahntay Jones who at #20 in 2003 is possibly meeting expectations, but certainly not in the plus category. Luol Deng though looks like he may end up in the "gem" category. But the careers of Williams, Felton, May, McCants also need some time to unfold before Duke is declared the definitive winner over the past 5 years.

As far as the "gem" analysis itself, we're not the only analysts who are examining the success rate in comparison with the expected outcomes of that draft slot. It's one method, and we certainly think there are others, and that there can be endless debate on exactly how to classify a player's outcomes, but again we tried to apply this fairly across the board. The non-UNC author had done the subjective ratings months ago for 2 previous DraftExpress articles ( first and second) whose scopes was well beyond the UNC-Duke rivalry. The UNC-Duke comparison was considered when we found UNC’s coaching success in a separate analysis, which led us to want to use multiple criteria to assess the impact of all the major programs. The UNC-Duke story became, we think, an interesting theme.

I think we tried to be sensitive to the Hurley situation, by specifically mentioning it, and reassessing, in the text, what the numbers would be with him removed. But if you ever analyze data from hundreds of individuals, there are many situations like career ending or altering injuries, and to go case-by-case and have an asterisks after every college whose player outcome numbers are altered would seem to be somewhat overkill and open to accusations of manipulating the data even more than Mr. McSwain believes we did. If we tried to hide the Hurley data point, we would never have mentioned him directly, but we did and tried to be fair. Because he did come back and play, it’s also hard to just throw his data out completely. How much was injury, how much would he have struggled anyway? Hard questions to answer. Hurley himself has said that he cannot blame the accident for his NBA struggles, but he just may be a stand-up guy who doesn’t want to find excuses. We have no ax to grind with Bobby Hurley.

We certainly tried to be factually accurate and fair with the data we presented and the conclusions we reached, and we stick by those. Even if one wants to debate the data in Table 2, by inserting the drafts of 2002-2005, the data in Tables 1 and 3 is hard to refute regarding the Carolina success in 60 years of NBA history. The complete body evidence in arriving at this conclusion, would be unchanged by the slight tweaks that outcomes of 4 drafts may influence. However, we also tried to add a little sense of levity about the rivalry, just as Heather (whose affiliation with UNC we were completely up front with) and her husband engage in throughout the basketball season. We used hard data, applied criteria as fairly across the board as possibly, and, yeah, maybe rambled a bit as Mr. McSwain stated, but there was a lot of ground to cover.

Anyway, we respect Mr. McSwain’s views and knowledge about these topics, he clearly gave this a thoughtful read and offers a couple of worthy retorts. We do, however, stand behind the data and conclusions.



Recent articles

Twitter @DraftExpress

DraftExpress Shop