What exactly does Sam Hinkie look for in a draft prospect?
A while back I took a look at this question when I examined some of the moves that Hinkie made as general manager or otherwise had input in (while in Houston), focusing on the attributes, statistics, and predictive models that I thought the Sixers head honcho could value.
At the time, I had no definitive conclusion: the question clearly needed more data to be answered. Well, not answered, because we could be 5 drafts into Sam Hinkie's career and still not come close to knowing what value he places on certain things. Still, looking at what Sam Hinkie did during his one year with the Sixers seemed to show a pattern with the guys he drafted, even if his time with the Rockets raised questions on whether this was just a small sample size or a philosophical difference between Hinkie and Rockets GM Daryl Morey.
Now that Hinkie has drafted 7 (!) more players, we have some more data to look at. So I'll once again take a look at the players the Sixers have drafted, looking at some common metrics that I generally assume analytically minded GM's value: size, athleticism, ability to generate steals, and projected WARP, Kevin Pelton's predictive model. Though instead of focusing on the statistics as I did last time, I'm focusing on the positives and negatives of the players' statistical and athletic profiles.
|Player||Year||Pick||Pos||Height||Wingspan||Reach||Max Vert||Stl%||Stl/40||WARPp||WARPp Rank|
For reference, here are averages for the athletic measurements of first round draft picks in the DraftExpress database:
Joel Embiid doesn't have the near-historic athletic results and steals combination that Nerlens Noel does, but he still grades out well across the board. His 1.5 steals per 40 minutes was 7th in the DraftExpress database among centers who played 20+ minutes in the NCAA last year, and his 7'5" wingspan is above-average. Unfortunately, due to his back injury, we don't have his full athletic testing results from the NBA combine, but it's clear he has an extremely good athletic base to build off of.
What is interesting is his projected WARP, which at 2.9 only ranked 7th in last years draft. This year's draft was a little unique in the number of foreign prospects that ranked highly in projected WARP, with Clint Capela, Jusuf Nurkic, and Dante Exum all coming out in the top 4. As Pelton looks to continually modify his formula to fix any inefficiencies in projected WARP, this will be an interesting year to see how well formula accounts for the ability for stats to translate from the various European leagues.
Still, taking the 7th ranking out and just looking at the projected WARP figure, while Embiid rates out well, he doesn't rate out like a superstar. His 2.9 projected WARP is well below darlings of previous classes, which included a 5.5 projection for Anthony Davis, 5.1 for Kenneth Faried, and 4.2 for Kyrie Irving, James Harden, and Blake Griffin. Still, unless Hinkie went with Marcus Smart (3.6 projected WARP) or Exum (3.3), he really would have had to have bucked the consensus to take somebody with more projected WARP than Embiid at #3.
If there's one draft pick from this class that really bucks the trend and makes you re-evaluate what you think you know about one Samuel Blake Hinkie, it's the Dario Saric selection. While some other selections in the second round don't hit all the check boxes, Saric was a high selection, a player Hinkie used a significant asset to acquire. And he fits not one of the presumed check boxes.
While Saric would be tall for a small forward, his wingspan (6'10" height, 6'10" wingspan) is below average for either forward position. 1.4 steals per 40 minutes ranks about average for a power forward and is nowhere near exceptional for a small forward.
Things Are Changing
Things Are Changing
Then there comes his projected WARP. While the top of Pelton's WARPp rankings were littered with European prospects, Saric was not one of them. Saric came in with a relatively pedestrian 1.4 projected WARP, ranking 21st in the draft class. Everyone from Doug McDermott to Gary Harris to Kyle Anderson and Jordan Adams were available and rated out higher on Pelton's predictive model.
This goes without saying, but whatever predictive model(s) Hinkie uses is/are most definitely different than the one publicly available from Kevin Pelton, and young players dominating European basketball are one of the scenarios that it's extremely easy to see different models significantly deviating from each other. If you think you can predict Hinkie's moves based on putting WARPp into a spreadsheet and sorting, this is as good of an indicator as any that he is more nuanced than that.
K.J. McDaniels is a player WARPp liked quite a bit, placing him at 19th in the draft class, well above the 32nd draft slot he was selected at. He has only average height on the perimeter, but his long arms and athleticism more than make up for it. Still, it's not so much the steals that make McDaniels intriguing - 2.3% steal rate and 1.4 steals/40 are just average numbers - it's the blocks. His 3.3 blocks per 40 minutes rates as the best figure in the DraftExpress database for small forwards in the top 100 last season. And that was not by a small margin, either: Josh Huestis came in second at 2.2 per 40.
Grant, on the other hand, was hated by just about every non-athletic measurement. He measured well, at 6'7.75" in shoes, and a 7'2.75" wingspan and 8'11" standing reach that would be incredible for a small forward. The problem will be making him into a productive basketball player.
While Grant obviously needs to develop his perimeter skill, it's his inability to turn his athleticism into defensive contributions that is perhaps his most troubling future indicator. Despite his excellent athleticism, Grant posted average to below-average marks in steal percentage (1.7%), block percentage (2.5%), and defensive rebounding percentage (17.2%). All this considered, it's no surprise that Grant ranked 69th in Pelton's WARPp, well below where the Sixers took him at 39.
The (re-)acquisition of Pierre Jackson allows us to take a look at how he measures out. Despite small size (5'10.5" in shoes, 5'10" wingspan, 7'6" standing reach are all well below average for his position), Jackson has the athleticism (42.5" max vert) and solid steal numbers I assume Hinkie looks at, and he ranked 11th in Pelton's projected WARP last year, posting a 2.1 WARPp.
Micic is another prospect who may or may not fit the Hinkie mold, depending on which mold it is you presume he's looking for. He's a tall point guard, something that the Sixers seem to value (although Pierre Jackson would not fit that mold). He also excelled at forcing turnovers, generating 2.3 steals per 40 minutes last season in the Adriatic league last season and came in slightly ahead of his draft slot at projected WARP, with a 0.6 WARPp placing him 39th in the draft class. On the opposite end of the spectrum, despite his height his length is relatively pedestrian (6'7" wingspan), and, while we don't have official athletic measurement results, he's not an explosive athlete.
Finally, there's Jordan McRae. The length is there (7'0.5" wingspan), but his athleticism is average (34.5" max vert), his ability to generate steals below average (1.4% and 0.9/40), and his projected WARP of -1.4 rated 89th in the draft class.
The data still suggests that while Hinkie likely does value a predictive model, he's far from beholden to it. Even beyond that, many of the declarative statements we make (i.e. "Hinkie would never draft somebody with a terrible steal rate) need to be re-evaluated. He's shown a willingness to draft guys rated far lower than their projected WARP would indicate, willing to reach on both skilled guys who are not necessarily athletic marvels (Saric) as well as incredible athletes who, as of now, don't have the skill level to really capitalize on them (Grant).
This serves as a word of caution not to make too many assumptions about what you think you know about Hinkie. My guess, as always, is that fans tend to oversimplify his views, sometimes painting the advanced statistics community with one brush. I think he uses advanced statistics and predictive models as part of his method, but is not someone who is necessarily defined by it.