Projecting Tobias Harris using Physically Similar NBA Players

In order to project the likely numbers Tobias Harris will produce next season for the Milwaukee Bucks, I looked to the past, and then applied the lessons learned to Toby’s resume.

I chose six players who I believed were “substantially similar” to Tobias Harris in terms of (1) height (6’7”.75) (2) weight (223) (3) reach (8’7”.5) and (4) strength (12 reps) and then examined how the college numbers produced by those players translated to the NBA.

The six players from the Draftexpress “Measurement History” files whom I believed were the most similar to Tobias Harris were: Carmelo Anthony, Damion James, Danny Granger, Caron Butler, Cartier Martin, and Jeff Green. I looked at all six in terms of their: (1) percentage decline in Win Score from college to NBA; (2) percentage decline in TS% from college to NBA (TS is “True Shooting Percentage”, a brilliant statistic created by John Hollinger that fully encapsulates a player’s ability to turn points into possessions. TS% = Points / (2x FGAs) + FTAs) ; (3) percentage decline in rebounding from college to NBA, and (4) positions played during their rookie season. I found a very surprising amount of consistency between the six.

Let me make a side comment here before I continue. I also used the Usage statistic in my similarity analysis. While I do not believe Usage is a good “value” statistic, I do believe it is a good “descriptive” statistic (how often is the player likely to shoot?). In that sense it was helpful. Tobias Harris is a “fair Usage” player, meaning his Usage rate in college was his fair share, 20%. Among the six similar players, Cartier Martin, Jeff Green, and Damion James were also “fair Usage” players in their last collegiate seasons. The other 3 took more than their fair share.

What position will he play?

First, I looked at the basketball positions each of the physically similar players were asked to play as rookies, to try to determine where Harris will play. I found that all of them played the majority of their minutes at small forward. Two of them (Green and Granger) played 58% SF and 42% PF.  Anthony played 83% SF and 17% PF. Butler and James were 100% SF. And Martin mixed time between SF (87%) and SG (13%).  So, based on that, and based on the available minutes on the Bucks, I would tend to place Harris closer to the Green/Granger mix.  (the Bucks have Jackson, Delfino, and Moute at the small already. They have Moute, Ilyasova, and Gooden at the power. So Skiles will probably just slip Harris in where he can, and Skiles is not afraid to play smaller players at the power forward.)

How did the six comparables produce?

When you analyze players based upon the difference between their college and NBA rookie numbers, it is unusual to find any consistency. I found remarkable consistency across the group of six, in terms of both Win Score decline, and general statistical declines.

The Group of Six each had a nearly identical decline in Win Score production, with the declines being almost perfectly distributed around the average of 58.2%.  The six were also remarkably similar in their TS% decline, with each being around 14.6% less efficient in their rookie season than they were in college (Anthony being the outlier at 7.5%). The rebounding decline was also remarkably similar, with the average decline being 26.7%. 

Based on straight WS decline, Harris would be expected to produce a way below average 5.09.  Based on statistical comparison, I would expect Harris’ box score to look like this:

Per 48 Stats FGA FTA Reb Ass Stls Blks Tos PFs Pts WS 
T Harris 13.9 4.4 8.7 2.1 1.1 1.1 2.1 3.8 14.8 6.11

What will his Defensive Win Score be? Harder to project. On average, the DWS for perimeter rookies last season was 107% of average.  That would be 9.03, if we assume Harris plays a 58/42 Small to Power Forward split. The averages among the Group of Six were 6.78 at Small Forward, and 12.78 at Power Forward. That would be 9.29.  The defensive averages at each position for the Bucks last season were 6.71 at Small Forward, and 10.46 at Power Forward.  That would be 8.28. If you average the 3 numbers, you come up with 8.86.  Let’s use that.

So we have the average range of Tobias Harris’ likely WS (5.57) and DWS (8.86). That’s a bad combination. That comes to a MWS of -1.64, which carries a player Winning Percentage of 0.223%. That sucks.

Lets be optimistic and use the better Win Score projection of 6.11, which came on a stat-by-stat analysis. Now we have a projected MWS of -1.37, and W% of 0.269% — still shitty. But we have to work with it. Now we need to figure out how many “player games” Harris is likely to consume. To do that, we have to speculate about how many playing minutes he will receive.

Speculating on rookie playing minutes is dicey. I looked for analagous situations to help me. I looked for players who were selected down draft, who did not really have a specialty, who were not that highly regarded, and who were unlikely to wrestle the starting position away from the team’s established veterans.  Those kind of players seem to fall somewhere in the 900 minute range.  Using the general per game Team Player Minute average of 241.2, we can project Harris’ “Player Game” number at 3.7 games. Thus, we could expect his line on next season’s Win Chart to look like this:

T Harris 6.11 8.86 -1.37 0.269 1.0__2.7 -0.9 0.1

Getting Real

Let’s throw all that away for a second and just talk.

What evidence is there to suggest that Tobias Harris will be a productive player for the Bucks? 

Lets look at the pros. He rebounds above average. He is 19 years old. He could improve. He is reputed to be a decent defender.

Now, let’s look at the cons.  He is not above average at anything in particular.  He is not good at turning possessions into points. His collegiate 2pt%, 3pt%, and FT% are all below average. You think those numbers will improve in the NBA?  Not likely.  His rebounding is above average, but his rebounding numbers are likely to decline in the NBA because he was a frontline player in college and will be a perimeter player in the NBA. As for the other box score statistics, no great value stands out.  Harris was below average in steals, and he never passes the ball, so it is unlikely he will produce assists. He may be a good defender, but we do not know.

In summary, Tobias Harris could be an above average player for the Milwaukee Bucks. I hope he is. But to do so, he must become a much more productive player than he was at the University of Tennessee. If that does not happen, all of the signs I am reading tell me he will be lucky to be an average NBA player.

Others more positive about Harris

I must note in closing that the Wages of Wins team project Harris’ production much more highly than I do. One of their models has his projected production very high.

I asked for an explanation, and they told me the model was most likely impressed that Harris could produce the numbers he produced at such a young age. That makes sense, and I hope it is correct.

But again, if you break the numbers down, you don’t have much reason to believe it will be correct. Will Harris improve his shotmaking? Maybe, but there is nothing to support the idea that he will. Will Harris do a better job of maintaining his rebounding numbers than other similar players were able to do? Again, maybe. But I am not counting on it.

Look at Damion James, the small forward drafted last summer by the Nets. He was a savage rebounder at Texas. He obviously has a rebounder’s mentality and will.  Yet even he had to take a big haircut when he got to the NBA (31% decline). Why would you expect a different result for Tobias Harris?


One Response to “Projecting Tobias Harris using Physically Similar NBA Players”

  1. MB Says:

    Great post – agreed that while productivity from Harris is a POSSIBILITY, it would take a serious triumph over the odds. Why not play WITH the odds, instead of hope for a miracle?
    I would imagine the answer is that they don’t view the odds in the same way.
    Incredibly frustrating that Faried was there, and the Bucks didn’t take him.
    What is going on in the Bucks front office? What statistics do they use as their measuring stick?
    I suppose what I’m asking is, what is the lens through which Tobias Harris (or nearly anyone else in this draft) looks like a better prospect than Ken Faried? and how is it that there are 21 such lenses being gazed through in NBA front offices? What am I missing?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: