Thursday, June 27, 2013

The End of Assists & The Best Facilitators in the Draft


According to most public draft boards, Phil Pressey is a late 2nd-round pick. There are legitimate concerns with his game, including his scoring and turnovers (however, his 4.5% tunovers per touch is still lower than plenty of starting guards in the NBA - including Chris Paul at 5.0%). Despite these concerns, our data indicates that Pressey may be the first or second best facilitator in this year's draft.

Ignore Assist Numbers

Traditionally, the only number used to judge a player's ability to facilitate the scoring of teammates has been assists. Here are the assist numbers for Michael Carter-Williams, Trey Burke, and Phil Pressey
Michael Carter-Williams: 7.3
Phil Pressey: 7.1
Trey Burke: 6.7
These numbers tell us that Carter-Williams was the most productive facilitator, with the spread between them being only 0.6 and Pressey only producing 0.4 more assists per game than Burke.

However, looking at assists alone may mislead us in a number of ways:
(a) it does not take into account the ability of teammates to make shots.
(b) it does not take into account the level of shot defense applied to made shots.
(c) it does not take into account most passes to shooting fouls, as assists are not given when the shooter is fouled and misses the shot as a result.
(d) it does not take into account passes before the assist that were just as crucial as the assist.
The drawbacks of the assist stat can lead to faulty analysis which cannot be easily or accurately accounted for with mathematical adjustments and assumptions.  Vantage Data and the following Vantage Stats attempt to correct these distortions and provide insight into this year's crop of facilitators.

Assist+ Per 100 Chances: Pressey Impresses

Assist+ is a number that tracks not only traditional assists, but also gives credit to a passer when:
(a) a pass results in a shooting foul
(b) a pass results in a missed open shot (attempts to de-penalize facilitators for teammates' poor shooting on open shot attempts)
(c) a pass is deemed crucial to an assist or a subsequent pass resulting in a shooting foul. Crucial passes can be thought of as "hockey assists," or assists-to-assists.
This number provides a more holistic view of a facilitator's ability and is a good place to start the analysis.  The top 3 in Assist+ in this year's draft class are as follows:
Phil Pressey: 14.2
Michael Carter-Williams: 11.5
Trey Burke: 11.1
Pressey's rate of production is #1 and 23% higher than Carter-Williams. Further, as stated above, one of the big problems with assists in general is the lack of adjustment for the shot defense faced by teammates. This means that a player can rack up high assist numbers just by passing to covered teammates that make contested shots, and our numbers show this is exactly the case with Michael Carter-Williams.

True Facilitation: A Shot Defense-Adjusted Standard

True Facilitation is the number of passes to uncontested shots per 100 offensive chances, regardless of whether the shot is made or missed. In other words, it excludes those assists where the shooter made a contested shot, but includes passes that resulted in missed uncontested shots. Therefore, True Facilitation is the best measure of a passer's ability to "find the open man."

Watching a few of Rajon Rondo's passes to open shots makes it clear that - until now - we've been undervaluing the facilitation ability of players that can do this consistently.


Here are the True Facilitation numbers for the top 3 Assist+ guys:
Trey Burke: 3.05 (#1 in TF)
Phil Pressey: 2.2 (#2 in TF)
Michael Carter-Williams: .738 (#23 in TF)
Pressey trails Burke, but is still #2 in the draft, but we see a major drop-off for Carter-Williams. This means that Carter-Williams's teammates were good at making contested shots, not necessarily that he was good at getting his teammates good looks.  This is a red-flag for those interested in Carter-Williams, and a confirmation for those interested in Pressey.

How They Facilitate 

The Vantage data set is capable of providing even deeper analysis than these new metrics on facilitation. And we hope you'll forgive us for diving a bit deeper here, but we think it is important to understand how these guys are facilitating opportunities for teammates.  Are they simply standing on the perimeter and reversing the ball?  Are they playing off a screen or driving?  Understanding how guys produce is extremely important in assessing fit for specific teams.

Again, for the three players we've been highlighting, here are the %'s of their Assist+ generated through the drive, screen, into the post, and in transition:


The outlier is, again, Carter-Williams.  Whereas both Pressey and Burke generated about the same number of facilitations via screens and drives, Carter-Williams thrived in transition and drives, not utilizing the pick and roll as much.

Here's a look at a few of Pressey facilitations off screens:




and off the drive:




Pressey and Burke are a close 1-2 in terms of best facilitators in this years draft.  While Pressey has the higher assist+, Burke has the higher True Facilitation, and both produce in similar ways.    

More New Metrics to Come

We are using these posts as a friendlier introduction of Vantage's capabilities than just providing data dumps or full scouting reports of individual players. If you missed any of our prior posts about new statistics, please check them out below and check back regularly (or follow on twitter) for the rest of the introductions.

And for those following along at home (admit it - you're at work), enjoy the draft!

1) Scoring
2) Facilitation
3) Rebounding
4) Screening
5) Turnovers and Fouling
6) Shot Defense

  • Contest+
  • Points Allowed Per Shot
  • Shots Defended Per Chance
  • Overall FG% Against

7) Disruptions

  • Turnovers Forced Per Chance
  • Deflections Per 100 Chances
  • Passes Denied Per 100 Chances
  • Pressure Rate

8) On-Ball/Screen Defense

  • Keep in Front %
  • Close Out Points Allowed
  • Points Allowed Per Screen
  • Effective Screen Defense Rate

9) Help/Double Team Defense

  • Double Teams/Helps Per 100 Chances
  • Points Allowed Per Double Team/Help
  • Effective Double Team/Help Rate

10) Movement and Involvement

2 comments:

  1. Interesting...

    ReplyDelete
  2. I've been saying this about assists for years!!!

    ReplyDelete