Counting the Baskets

Point Guard Island: Draft Model Update

I ran my draft model against the latest Draft Express data, which includes the first weekend of the tournament, so only a few players will have more than more one or two games of data added (though most likely that will include all of the prospects on Kentucky).

Before I get to the inevitable list I wanted to both look at the relationship between the model out put and Draft Express Top 100 prospects as well as a couple of visualizations of the stats in the models.

Using the rank given by the model and the DX rank I plotted them and ran some correlations based on position and a split between top and bottom 50.

So basically the model and Draft Express ratings agree much more at the top of the rankings than at the bottom, where there is no agreement. There's some logic to that given that the history of the draft indicates that there is more consistent differentiation at the top of the draft order than the bottom, which gives us the famous logarithmic shape of draft value.

Point Guard Island

The relationship by position is interesting. It is tempting to see the higher correlation for centers and power forwards as something indicative of the PAWS model being more in tune with scouting for bigs than guards. But last year the model and the consensus had Marcus Smart, for example, in the same neighborhood, with the disagreement the greatest with a couple wings and power forward/center in Clint Capela.

In any case, right now there is basically no agreement between my model and the DX rating of point guards, but a decent relationship with centers, power forwards, and shooting guards.

Below we can see a picture of what no correlation looks like, along with my little (literally) point guard island on the left side of the graph.

The diminutive trio of Tyus Jones, Fred VanVleet and Tyler Ulis are ranked 2nd, 5th and 19th by the PAWS model, while they're 28th, 86th and 89th respectively by Draft Express. Any analytics draft model is at a disadvantage on the defensive end, and it's clear that is a concern with these three. The defensive measures we do have, blocks and steals are below average for both of the freshmen, Jones and Ulis, but VanVleet is competitive with the other higher ranked point guards.

Below the statistics used to inform the draft model are graphed in standardized ratings for scoring, rebounds, blocks plus steals, distribution (assists and turnovers) and a rating combining age, consensus high school recruiting rank, and competition level. The weights mimic the model, so offensive rebounds are more valuable than defensive, steals are more valuable than blocks and age is the dominant factor in the Age and Competition rating.

Jones, VanVleet and Delon Wright form the models' first tier. Jones gets there in significant part due to his age and high consensus recruiting rank and his distribution numbers. Wright essentially does everything well, but is on the older side. VanVleet is younger than Wright, but has been a below average scorer. We'll see how VanVleet's tournament performance affects his draft stock, a good game against Duke could make him a fair amount of money.

Here's the same stacked visualization for the top ranked shooting guards:

D'Angelo Russell is the clear leader, with something of a second tier in Jerian Grant and R.J. Hunter. Mario Hezonja does not grade out as well as his scouting, but he is playing on a high level professional team in Spain, where he has struggled to get playing time consistently.

Here are the small forwards:

The model has Stanley Johnson and Aleksandar Vezenkov as a first tier of sorts, with Justice Winslow, Kelly Oubre, Sam Dekker and Rondae Hollis-Jefferson bunched behind. The big IF with Vezenkov is his defense, his scoring in the Greek leage has been efficient at a high volume. The model does not rate Jake Layman as a top prospect.

Here are the power forwards:

Frank Kaminsky, Bobby Portis and Kevon Looney top the power forwards. Kristaps Porzingis rates less highly than he did last year, after failing to make much progress in a chaotic season for his club in Spain, though his scouting stock hasn't yet suffered.

Finally, the centers:

There is a clear first tier with Jahlil Okafor and Karl Towns, and then a very talented second tier. Willie Cauley-Stein is somewhat lower in the model estimate than his DX rank, in part due to the limitations valuing defense, but also Cauley-Stein's age and relative lack of scoring.

 

 

Defensive PT-PM: Work in Progress

I debuted my Player Tracking Plus Minus here last year as a Beta stat. Basically I have never stopped making changes and already put out one update on Nylon Calculus. 

In terms of writing I try to find a balance between writing up every little finding, for whomever is actually interested, and just holding on to the end. My plan is to wait for the end of this season, sadly less than twenty games away to do a full 2.0 version. That will give me a better target variable in terms of reliability from last year and not have to worry about synchronizing any data updates from the different source I am using. But the defensive side is the most interesting and I thought I'd put some of the prep work out there in order to clarify my own thoughts and get any feedback.

Play by Play Fouling

As I alluded to in my fouling rate model the other day. I have been adding the specific foul data via NBA Miner to the model, and it appears to be a big improvement. The two specific foul types that consistently show value in cross validation are offensive fouls drawn as a positive and shooting fouls as a negative. Happily these are nice common sense terms that any basketball coach would recognize; defend without fouling and beat the offensive player to their spots.

The general term for offensive fouls dawn, which includes charges as a subset, is +1.26 per offensive foul drawn per 40 minutes. That is a high figure, higher than steals for example. So it's important to keep in mind what a orthogonal RAPM derived model is trying to do, which is to determine an estimate of player value not a description directly of the value of each play. Sometimes a model will pick up an attribute that seems to be associated with good or bad players, and may, in fact, act partially as a proxy for other actions not seen in the model. This is especially true on defense, where our data is fairly limited.

Blocks, prior to SportVU data, have been a proxy for rim protection, a decent, but flawed one. Steals in draft model especially have acted as proxies for spacial awareness and defensive engagement. Here I think that offensive fouls drawn act as a proxy for physicality on defense and defensive anticipation. Another interesting note is that there is no other term in the model with a strong direct correlation to drawing offensive fouls, so there is definitely some independent information being conveyed by a player's ability to draw offensive fouls, however that ability is interpreted.

Contesting Shots

Contesting three point shooting is noisy business, with little year over year relationship in the results, as I found here. And showing a coin flip like pattern in the results. Defending shots within the three point arc is still noisy, but there also seems to be some signal. So, for that reason I have used the percent plus minus on two point shots contested by the player. Even though including all shots improves in sample correlation to this year's defensive RAPM more, it is less likely to help predict next year's performance.

Defensive Rating

Lastly, the NBA version of defensive rating (DRTG), which is a descriptive statistic of how the team performed with the player on the floor, is used in as a second stage variable on the residual of the individual defensive statistics model. I have used that technique for a while, the good thing about this method is that it allows me to pick up the information from individual performance without being overwhelmed by correlation between DRTG and defensive RAPM, then allows me gleen the additional information from team performance rather than throwing that information away. It is also an improvement over using over all team performance that includes defensive performance when the player isn't even on the court.

One good piece of news is that the DRTG effect is just over two thirds of what it was in the older version with the new information added, so the model is able to attribute a bit more value to the discreet actions of the player.

My player updates at Nylon this year will continue to use the PT-PM as developed earlier, since it is confusing to change both the metric and the input data at the same time. But next year I will unleash the power of this fully operational player metric.