clock menu more-arrow no yes

Filed under:

Composite NBA Draft Modelling and Undervalued Prospects

New, comments

A recent article put together by Draft Express brings together various statistical draft models. What do they tell us?

Troy Taormina-USA TODAY Sports

Here at Canis Hoopus we have a long history of trying to tease out what college statistics can tell us about potential draftees. After all, the draft has been our playoffs for the last decade, and the subject of a massive amount of debate ever since the site got started.

Stop-n-Pop used to figure something called the Hoopus Score to rate draftees, and since then Madison Dan came up with his own rating system, and of course our own now-sort-of-famous-in-nerd-basketball-circles-and-why-don't-the-Wolves-pay-attention-to-him Layne Vashro (vjl110) has a model that he's been working on for years.

Others have rating systems as well, and Draft Express recently brought together several of these models and model makers to take a look at how they work and where they differ in an article that includes vjl's system.

One of the great benefits of a model like Layne's (and others) is not necessarily that it tells us who should go first, second, or third in the draft; rather I think it's greatest strength is in identifying players who may be undervalued and available later, and also that it reminds us to take a second look at guys who are getting a lot of lottery attention they might not deserve.

SIDEBAR: Awaiting Flip's annual radio commentary in which he both shows that he's familiar with the models while at the same time mocking them by willfully missing the point.  Two years ago he went on the radio and said that Mike Muscala would be the first pick in the draft if you "went by the stats."  In fact, Muscala was seen as an undervalued prospect by vjl's model, and would have been a terrific mid-to-late first round pick; Atlanta is not unhappy to have him. Last year he pointed to Jordan Adams as the "stats top pick," and I think many of us would have seriously considered him at 13, again thanks to Layne's work.

Taking a look at multiple models together is interesting, because we can begin to see how they differ in analyzing prospects. For example, Layne's 42nd rated prospect (by EWP), Bobby Portis, still ranks 12th overall in the composite rankings that include Layne's rating. In fact several of the players that Layne "underrates" compared to other models are these sorts of non-stretch power forwards. Though as we'll see below, two of the more underrated players in the draft according to the models, including Layne's, are players loosely of this type.

They also serve as a check on each other, reminding us not to take any one approach to analyzing prospects as gospel.

At any rate, I wanted to take a look at the composite rankings from these models to try to identify both types of players in this year's draft--those who are mocked too high according to the numbers, and the values that might exist later in the draft.

There aren't many players who are getting mocked in the lottery that the models as a group dislike.  The one player that fits that description is Devin Booker, the freshman shooting guard from Kentucky. Booker shot the ball quite well from the perimeter during his one year in college but did very little else. He's very young (won't turn 19 until opening night of next NBA season), and appears to have one NBA level skill, so he isn't undraftable, but the models see him as an early 2nd round pick, while DX has him mocked in the lottery. Teams can probably find a value shooter later in the draft without spending that asset on Booker.

On the flip side, the models identify several players who are mocked in the mid-first round or later that seem to deserve more consideration. This is interesting for us Wolves fans, who can imagine the Wolves packaging their 2nd rounders and other "assets" to move up and grab one of these guys.

UCLA power forward Kevin Looney jumps out. The models composite ranking has him as the 5th or 6th best player in the draft, while DX has him going 19th. Looney was a freshman this past season at UCLA, and like several before him, got off to a strong start to the season but faded as the year went on. Conditioning appears to be an issue for many UCLA players, and whether that was the problem for Looney or not remains to be seen. Still, he's a strong offensive rebounder with some perimeter skills on both ends of the floor.

UNLV PF Christian Wood is similar, in that he's seen as the 11th best prospect by the models but mocked 26th by DX. A tall and skinny shotblocker who also showed some rebounding chops, Wood should get stronger as he fills out. He showed some good touch inside the arc, but though he took a fair number of 3s (3 attempts per game), he didn't make them.

Two point guards join the list of players the models like a lot more than the mock drafters: Tyus Jones of Duke and Delon Wright of Utah. The Models have them as the 7th and 9th best players respectively, while DX currently have them going 18th (Jones) and 30th (Wright).

Jones helped Duke to the National Championship this season as a freshman, and of course hails from Apple Valley, making him quite popular among the Wolves fan base. He's short and not particularly athletic, but he showed a knack for making the right play, appears to have a high basketball IQ, can make a pass, and shot it pretty well (38% from three, 89% from the line).

Wright is much older (23) and bigger (6'6") and not the shooter Jones appears to be, but is a tremendous floor leader and terrific defensive player who averaged over 2.5 steals per 40 this season. He looks capable of stepping in right away and being a quality back up point guard and possibly one of the handful of best players on a team like the Wolves from day one.

A few other guys who the models see as somewhat undervalued, and who might be available when the Wolves 2nd round picks come around include Chris McCullough, a Syracuse freshman power forward, Richaun Holmes, senior power forward from Bowling Green, Terry Rozier, sophomore point guard from Louisville, Vince Hunter, sophomore power forward from UTEP, and Seth Tuttle, senior power forward from Northern Iowa.

The Wolves need to get better at taking advantage of their opportunities to add talent. There are guys who can help at all stages of a draft, especially one like this year. Whatever they do, selling off those high 2nd round picks for cash will be a mistake given the amount of talent available and the value of those picks. Somehow they need to be used to add talent, whether it's by trading up, trading for a veteran, or drafting foreign players who could become Wolves sometime down the line.

What would you like to see the Wolves do with those 2nd rounders given the guys these objective models have identified? I'm inclined to try to go after Delon Wright if they can move up to nab him with their two second rounders. What about you?