Magic number: Terry Rozier’s passing shows room for improvement

As we settle into the dog days of the NBA summer, with the Celtics’ offseason largely concluded, we here at CelticsBlog are looking back to last season. With the Celtics retaining largely the same roster from last season — but healthier, we hope — we’re looking to identify some statistical areas for potential improvement from last year’s squad. Our subject today, as I’m sure you’ve guessed: Terry Rozier.

If it weren’t for the spectacular arrival of Jayson Tatum or the dramatic emergence of Jaylen Brown, the 2017-18 season in Boston might have been remembered as the Year of Scary Terry. After a couple years spent buried on the bench becoming something of a punchline for critics of Danny Ainge’s then-inactivity on the trade market, Rozier broke out in a big way last season, playing tough defense, providing energy and scoring punch off the bench, and shooting an impressive 38% from behind the arc. Come playoff time, Rozier was a fully emerged starting point guard and posting 16-5-5 splits while helping the undermanned come a game away from the NBA Finals. As far as breakout seasons go, it would be hard to draw up a better one for Scary Terry.

This is far from Rozier’s final form, of course, and with the Celtics looking to contend for a title in 2018-19, further improvement from the 24-year-old would be welcome. With that in mind, here’s the benchmark we’re setting for Rozier to improve upon in 2018-19: 6 potential assists per game.

Put simply, potential assists are passes leading to a shot attempt that, if made, would count as an actual assist. We’re using them as our frame of reference because, while still a little flawed, they illustrate passing in a bit more illuminating fashion than raw assists themselves. There’s a little too much out of the passer’s control when it comes to assist numbers — even a perfectly thrown kick-out for a corner three won’t count if the shooter doesn’t make the shot.

So here we have Rozier, who, as I’m sure you’ve surmised, left a little something on the table in this regard. His six potential assists created per game ranked only 74th in the NBA among players who appeared in at least 50 games, in the neighborhood of players like lackluster Wizards bench guard Tim Frazier and Magic center Nikola Vucevic, and a distant fourth on the Celtics behind Kyrie Irving, Marcus Smart, and Al Horford. This was less than optimal for a player who saw significant time as the team’s only healthy point guard down the stretch after two of those three were sidelined due to injury.

Common NBA analysis generally discusses Rozier among the higher upside point guard prospects in the league. The upside part of that designation is certainly undeniable, but the point guard part, I’m not so sure about just yet. Rozier reminds me more of the role Avery Bradley filled for the Celtics in years past — an undersized two-guard who plays tough defense and contributes offensively through a heavy dose of three-point shooting. I’d argue Rozier is even a more fully realized version of Bradley, far more capable of creating off the dribble and impacting the game with his athleticism. For the first three-quarters of the season, Rozier’s season looks much the same as Bradley’s final season in Boston. Bradley generated 4.9 potential assists per game on an assist-to-pass percentage of 7.0%, while Rozier generated 5.6 on 6.4%.

At the time of the injuries, the Celtics were fresh off their best offensive month of the season, posting an offensive rating of 108.8 despite brief slumps from Jaylen Brown and Jayson Tatum. In March, things changed. With Smart’s injury and Irving’s rest-of-season shut-down, that mark dropped to 105.9 in March and 101.4 in April. Rozier began throwing over 30 more passes per game, but his assists per pass improved only a negligible amount to 7.2%. In other words, he was throwing more passes, but not passing much more efficiently. This, combined with the ball spending more time in the hands of weaker facilitators in Brown and Tatum, caused the Celtics’ turnovers to balloon from 12 per game in February to 15 and 16 in March and April.

Rozier’s struggles as a passer were most apparent in pick-and-roll situations, where he developed something of a reputation among Celtics fans for not seeing the roll man or misfiring on lob passes. Under normal circumstances as a two-guard off the bench, this wouldn’t really be an issue, but once he was starting at the point, it became impossible to overlook. Rozier is just not very natural as a pick-and-roll passer at this time. His natural inclination towards looking for his own shot tends to clash with his responsibility for getting other players involved, and as the season went on, it led to a number of awkward possessions.

To Rozier’s credit, his performance as a passer improved markedly as the Celtics’ starting point in the playoffs. He dished out 3.8 more total assists per game, including 3.6 more potential assists, and these weren’t simply a product of playing 10 more minutes per game — his assist-to-pass percentage rose to 8.6% as well, showing some notable improvement in his pass selection. While the Celtics’ offense certainly struggled at times in the postseason, it wasn’t the garbage fire many anticipated it might be without Irving on the court, and that is in no small part due to Rozier developing on the fly.

The playoffs illustrated that, in the grand scheme of things, Rozier’s passing struggles are more of a small weakness than a massive fundamental flaw. The NBA has become increasingly friendly towards score-first point guards who can attack mismatches and find their own shot. Even the point guard he’ll be backing up in 2018-19, Irving, has himself never been the most prolific passer during his career. Furthermore, a healthy Marcus Smart stands to see most of the facilitation responsibility on Celtics’ bench units, allowing Rozier to focus on more of a scoring-first role. It’s only in the absence of Irving or Smart that he should be pressed into bona fide point guard work.

That said, we’re interested in seeing players become the best versions of themselves in this series, and Rozier’s passing offers the most obvious route for his improvement as a player. He may not see the sheer minutes that allowed him to record nearly 10 potential assists per game in the playoffs, but he still figures to be a very significant part of the rotation, and he’ll have even more weapons to work with on a Celtics team that finally looks to be 100% healthy. Pushing his potential assist numbers up from the 6.0 he posted per game in the regular season would be a significant improvement for him as an individual, and it would make the Celtics’ offense that much more dynamic as a result.