Analytic models see Wright, Looney, Wood as undervalued draft prospects
There's a great deal of information that goes into making a draft day decision. While most people think primarily of scouting and game tape, the combine, individual workouts, and interviews all play a large part, too.
And yes, as much as it may rub some of the more old school fans the wrong way, most teams will take a look at what data models have to say. Analytic departments don't exist around the league just to comb over SportVu data and perform post-facto efficiency analysis, after all.
Factoring in size, measurements, athleticism indicators, college performance, age, comparable players, and more, there are several predictive player models that are made public ahead of the draft. Leading draft site DraftExpress brought together six such models this week to get a feel for which players are uniformly appreciated, and while most of the results aren't surprising, there are a few players who stand out.
Without reproducing their entire work and rankings, here are the players deemed to be most under- and overvalued potential first-round picks based on DraftExpress and ESPN rankings.
| Prospect | Model Composite | DX Rank | ESPN Rank | Average Difference |
|---|---|---|---|---|
| Delon Wright | 9 | 25 | 26 | -16.5 |
| Kevon Looney | 5 | 17 | 19 | -13 |
| Christian Wood | 11 | 22 | 25 | -12.5 |
| Chris McCullough | 18 | 27 | 29 | -10 |
| Robert Upshaw | 23 | 26 | 38 | -9 |
| Tyus Jones | 7 | 11 | 20 | -8.5 |
| -- | -- | -- | -- | -- |
| Willie Cauley-Stein | 16 | 4 | 8 | 10 |
| Trey Lyles | 24 | 16 | 12 | 10 |
| Michael Frazier | 45 | 31 | 36 | 11.5 |
| Devin Booker | 33 | 9 | 13 | 22 |
| Rakeem Christmas | 60 | 29 | 39 | 26 |
| Jonathan Holmes | 71 | 21 | 56 | 32.5 |
These results aren't all that surprising. Delon Wright is a senior, and seniors notoriously take a hit due to their age, even if they may be able to contribute right away. Kevon Looney, Christian Wood, and Robert Upshaw are all players seen as projects, but there is obviously something in the profile of each worth waiting on. Chris McCullough is still recovering from a torn ACL. Tyus Jones has long been heralded as an analytics-friendly player thanks to his low-mistake game and despite his diminutive stature.
The players that the models dislike are explainable, too. Willie Cauley-Stein dominates on the defensive end, which is more difficult to quantify. Devin Booker and Michael Frazier may turn out to be one-dimensional shooters without further development. Trey Lyles played his freshman year largely out of position and is somewhat unknown as a power forward. There's also a chance all three Kentucky players are being undervalued by the models thanks to decreased roles on such a loaded team.
None of this is damning for any one player, and it's not a deal-maker, either. Models like these are one more tool teams have to evaluate players ahead of these very important, potentially franchise-altering decisions. Leaving any available information unexplored would be reckless.
Note: The six models compared by DraftExpress are explained in greater detail here and come from Layne Vashro, Steve Shea, Nick Restifo, Jesse Fischer, Masseffectlenk, and Daniel Myers.