Possession Stats and the Future of Phoenix’s Three Point Guard System


Sep 29, 2014; Phoenix, AZ, USA; Phoenix Suns guard Goran Dragic (left), guard Isaiah Thomas (center) and guard Eric Bledsoe pose for a portrait during media day at the US Airways Center. Mandatory Credit: Mark J. Rebilas-USA TODAY Sports

There have been many words typed out this week on the Phoenix Suns’ short-lived three-point guard experiment. Hindsight being 20/20, it is easy to point out the differences in statistics between this season and last for the players involved. But one line in particular, from ESPN’s Tom Haberstroh, got me trapped in the SportVU treasure chest of data:

"Dragic has every reason to feel slighted. He was one of the NBA’s best point guards last season, and they made him split time with two other point guards after inking both Isaiah Thomas and Eric Bledsoe to long-term deals. How much has Dragic sacrificed in the process? Last season, he possessed the ball 6.4 minutes per game, according to SportVU data. This season? Just 3.9 minutes – a 40 percent drop."

So I was curious: Which players have had the largest changes in their time of possession this season from last? One factor that had to be accounted for was playing time. For Dragic, his playing time has hardly changed – 35.1 minutes per game last season compared to 33.4 this season – but for some players, that could cause a massive fluctuation.

Unfortunately, SportVU data is only presented in per game stats. But by looking at  Time of Possession%.[1. Simply time of possession divided by minutes played.], holding playing time constant for all players, I could compare changes from last season. I also only looked at the 218 players with at least 750 minutes played in both seasons. Here are the top-10 decreases and increases, as of the All-Star break:

Dragic, Isaiah Thomas and Eric Bledsoe are all in the top eight for largest decreases in Time of Possession%. This isn’t just about the difference that Dragic faced from 2012-13 to 2013-14 when Bledsoe joined in the fold in the desert — SportVU data only was available in all arenas beginning last season — but spectacularly, all three saw less time on the ball this season. It’s intriguing to theorize reasons behind some of the other large increases and decreases in the game.

One other item I wanted to look up was how the Suns did in combinations of the three point guards. ESPN’s Kevin Pelton, among others, pointed out that Dragic was far more efficient individually — even while in a higher usage role without his backcourt teammates. But how did the team do in lineups with all three guards, with two of them or with them all separated?

Obviously, there are two key caveats with that data. First, the usual disclaimer with regards to plus-minus data. When two or more of these point guards are playing, it usually was with fellow starters and other good players. When they were alone or none or playing, it was the Suns bench. The second disclaimer is that Dragic had the fewest individual minutes. Only 7 percent of Dragic’s minutes were completely separate from the other guards, compared to 8 percent for Bledsoe and 29 percent for Thomas.

Now, all three point guards have been dispersed completely. Dragic is off to Miami where he replaces Mario Chalmers as the starting point guard alongside lead guard Dwyane Wade. Thomas is off to Boston to work alongside the rookie Marcus Smart, guard Avery Bradley and recent lead distributor Evan Turner. And Bledsoe now has a new backcourt mate in Milwaukee’s Brandon Knight, who nearly was named an All-Star just recently. How do these guards conceivably fit, in terms of the earlier possession stats?

This doesn’t begin to include all of the other components of the Point Guard Personality Test, but it’s a neat little start and perspective on every situation involved. For every team, there are only a certain number of possessions per game. Players, especially the best ones, will overlap. Who will be possessing the ball the majority of time during their minutes? This tool can begin to sort things out.

In Miami, Chalmers and Cole were both more involved in handling the ball than I expected them to be. LeBron James had a 13.4 ToP% last year in Miami and is up to 16.4% this year in Cleveland. So Dwyane Wade will continue to be very involved from his slashing role, but there certainly seems to be a large opportunity for Dragic to step right in.

In Boston, Rajon Rondo was heavily involved before his mid-December trade to Dallas. Since then, it’s been mostly Turner. Marcus Smart has been used in more spot-up possessions than one might have initially guessed before the season, as Seth Partnow mentioned yesterday. Thomas can now conceivably replace Turner as the lead ballhandler here too. The Smart-IT combo could be lots of fun in Brad Stevens’ system.

In Milwaukee, the replacement is Michael Carter-Williams (22.7 ToP%), who was handling the ball quite a lot this season in Philadelphia’s system. But previously, Brandon Knight had a very large role in the operations. Perhaps the Bucks are comfortable with Carter-Williams in that role? Or perhaps Bayless will take over more frequently? It’s a pretty intriguing situation to watch.

And that brings Knight over to Phoenix to recreate a new two-guard system with the lone holdover, Bledsoe. It’s not as dynamic as Dragic-Bledsoe was last year, perhaps, but there are still reasons to be optimistic. The Suns have had quite a bit of team success implementing multiple-guard looks of any kind over the past year and a half.

More from Nylon Calculus on the Phoenix Suns experiment:

Why The Phoenix Suns Want All The Point Guards – Blake Murphy, July 2014
Guard-ians of the Phoenix Suns – Matt D’Anna, January 2015