Has Marcus Smart improved this season?

Marcus Smart is the most polarizing player on the Celtics. Many see him as an essential contributor to the Celtics’ high effort identity. Others are driven mad by his shooting and mental lapses. Brad Stevens praises him and touts his playmaking abilities. But rumors swirled around the deadline indicating that he was available for a first round pick.

Old friend Kevin O’Connor often reminds us that NBA free agents are paid for projected value, not what their talent has already contributed. That’s why budding talents, such as Anthony Davis, receive mega-deals, and 2018 Vince Carter plays on league minimum deals as they race into the sunset of their careers.

Free agents coming off their rookie deals can be tricky. Beyond the obvious stars, players that show incremental progress sometimes return on sizable deals that anticipate growth. In particular restricted free agency sometimes produces significant offer sheets that force the drafting team to decide between matching and losing a player they invested in developing. Others get squeezed by a lack of market and an unwillingness to tie up free agent money on offer sheets. It will be interesting to see where Smart falls in that spectrum.

The deal Avery Bradley signed in Boston comes to mind. He was an unrestricted free agent and Danny Ainge offered him perhaps a little more money than the initial market might have dictated, anticipating the improvement that followed. He soon became a bargain at his price point.

Smart might be the hardest player the Celtics have had to evaluate coming off his rookie deal. He’s not a star by any means. His game is defined by unpredictability, for better or worse. He’s emerged as one of the best defenders and worst volume shooters in the NBA.

Through it all, his touted “winning plays” (credit Stevens) have illuminated as brightly as his infuriating tendencies produce criticism against him. In the Dec. 28 game against the Rockets Smart was 6-of-14, 1-of-5 from three and a -1 (+/-), before single-handily flipping a loss to a win.

As a draft selection, Smart checks off as an initial success. He rose from handyman to sixth man in a system that developed hustling overachievers to reach the East Finals within three seasons. But with 27 games and playoffs left until his rookie contract expires, it’s evaluation time.

How much more can Smart grow as a player? Let’s break down his evolution and improvements through his first four NBA campaigns.

Shooting

Smart takes the sixth most shots per game on a roster despite ranking 11th in field goal percentage. This year started with an optimistic video of Smart’s improved mechanics, hitting three-point bombs in a gym and he proclaimed himself a “different me.”

His preseason results teased a potential improvement. Through four exhibitions Smart drilled 7-of-12 threes, exhibiting his compact form with a smooth (not drooping) follow-through motion on quality shot selection.

That progress faded as quickly as Boston’s projected Gordon Hayward starting lineup. On opening night, Smart missed all four of his three-point attempts and finished the night 31 percent on 16 shots.

Into his fourth season, whether or not Smart will ever be an average NBA shooter lingers as a major concern toward his renewed contract. There’s statistical evidence that his defensive impact outweighs his offensive misfires. Still, Smart continued to fire away, even as he posts percentages in the 20s to start the season (that’s field goal percentage, not 3pt.).

Smart’s volume increased once again this year, attempting 0.2 more field goals per game (9.7). Only 26 NBA players sit between 9-10 FGA. Smart’s 35.7 is the worst FG% among them. Second is Terry Rozier at 39.6 percent.

Smart’s FG% is 0.2 points below his 16-17 mark. His three-point percentage, up and down around his 29 percent career average on a steady four attempts per game, has risen from 28.3 to 29.9 percent.

Poor shot selection contributes to his inconsistency. Stevens has never publicly stated any reservation against the number of Smart shots, but with more shots per game, more chucks like these completely waste possessions.

Smart’s threes are not all nonsensical heaves, but there’s a solid split between the good and bad. 152 of his 211 (72%) attempts came from 25-29 feet, which are shots one step or more behind the line above the break. He shoots 29.6 percent there.

From the corner or right behind the line, he’s taken 52 shots at a 32.7 percent clip.

Those splits are typical of the guards in Boston’s offense, but Kyrie Irving and Rozier convert around 40 percent on above-the-break threes.

When Smart shoots in the corner, preferably the right side (46.7% on 15 attempts), he’s 31 percent on a low sample size (29). In 2016-17, he hit from both sides, at 43.8 (left) and 35.7 (right) efficiency. Plays like these are good looks for him, but don’t occur often.

Last year was a different story. Smart’s attempts between close/far threes divided on a nearly 50/50 line (149/165). 28 percent of his shot attempts came at the rim, compared to 23 this season. His post-up attempts produced 0.99 points per possession last year at 9.6 percent usage, but are down to 0.63 PPP and 6.3 percent thanks to Boston’s increased interior presence.

Smart is playing above the break more than ever, finishing less at the rim and the miscellaneous play types appear far less often for him now. It’s resulted in his worst career offensive box plus minus (-2.3), offensive rating (94), and player efficiency rating (9.4).

Smart, as a shooter, is definitively worse since last year. Does that mean his potential is gone?

Passing

As Smart’s passing game rapidly improved over the first 3 seasons, it qualified him to assume sixth man role. Regardless of his shooting, he could be pound-for-pound the best facilitator on the court in many lineups.

Smart is pick-and-roll qualified and can pick apart holes that the defense leaves. Marcus Morris’s efficiency benefits immensely from this. It’s easier for Morris to roll than force up shots he’s creating off his own dribble with the bench unit.

Watch Smart and rookie center Daniel Theis pulverize one of the best defensive bigs in the game.

It’s amazing how defenders seriously stick with Smart in the pick-and-roll despite his poor efficiency in-between (36.2%). But his unpredictability as a shooter and keen eye as a passer startles defenses. He’s not afraid to use the bounce pass either.

It’s no surprise that Smart accounts for 30 percent of the Celtics’ PnR ball-handling and produces 0.79 PPP. That’s a good mark, up from 0.67 last year with more usage.

Smart is creating additional points while still shooting under 40 FG% in P&R sets that end with his shot. His in-between game remains awkward, which is strange considering he was effective in that area in college.

Smart’s assist percentage jumped from 22 to 22.9 percent last year to this. He’s added 0.1 assist per game to his average. So why does he look so bad in offensive analytics?

The turnover frequency in his PnR sets is down, but overall Smart’s turning the ball over at a substantial rate. In the same number of minutes per game, Smart is only averaging 0.3 more minutes on the ball. His role is essentially the same, but he’s causing more turnovers than Irving in far less minutes handling.

Plays like these feel common and inexplicable for Smart.

Among players appearing less than 31 minutes per game, Smart ranks 11th worst in turnovers; he and Eric Bledsoe feature the most giveaways among players in the 30 minutes/game range. Rozier, with a nearly identical usage (19%), boasts a 2.78 assist-to-turnover ratio to Smart’s 1.89.

Rozier’s presented a balance of good offensive (107) and defensive rating (101). For the first time ever, he’s producing a whole point per 100 possessions over a replacement player (1.8 box plus-minus).

Smart’s creating almost less than one point below an average player (-0.8). His turnovers have spiked at a higher rate than his assists, depreciating his passing game.

Defense

Smart’s defensive skills stand as his most valuable asset. He rips through screens routinely, decimating pick-and-roll sets. But can we anticipate more improvement?

His defensive rating (103) and defensive box plus-minus (1.4) sit in line with his career averages. So do his steals (1.3). He’s needed less in the rebounding game this season, with more size on the Celtics roster.

This year Smart’s been a good defender, but with length and size everywhere on the roster, does he remain irreplaceable? On the surface, it would seem no. But the on/off numbers tell a different story.

Though the Celtics spike from a 107.1 to 108 offensive rating when Smart exits, opponents increase their offensive output from 101.4 to 105 points per 100 possessions. That’s a significant 2.7 point difference in the positive direction when Smart is out there.

Opponents drop their turnover percentage by 1.4 points with Smart off the floor, because not only can Smart play between different positions on defense, he transitions between his spots quickly, anticipates, helps and attacks the ball with intensity. This play ended in a foul, but a quality one.

With 2.5 deflections per game, Smart doesn’t always make the steal or block, but contributes heavily to assisting the teammate who does. He’s 19th in deflections, among many players who average well over 30 minutes per game.

Boston’s steal and block percentage drop by nearly 1 percent with Smart off the court.

While the overall on/off defensive stats settle in the same range for Smart as last year, the Celtics do seem better suited to counteract Smart’s offensive issues with their effective field goal percentage only dropping about one point with him playing.

It is hard to quantify Smart’s defensive improvement year over year, but his impact within the system is still substantial. It is hard to imagine replacing with anyone else and getting similar results.

Conclusion

The statistical evidence indicates that Smart is roughly the same player he’s always been (with some offensive regression). He’s valuable to the Celtics’ system, but it is difficult to foresee improvements coming due to his deficiencies. The turnovers and three-point shot selection indicate that he’s pressing on the offensive end. That could be blamed on integrating into a new roster or it could speak to a fit issue with this new group.

Either way, this is the sample size that Boston will have heading into this summer. If some team makes an unexpectedly bets big on Smart, the growth of Rozier may be a better value investing in long-term. That said, Boston is still better this season with Smart than without, despite his shot selection and turnovers diluting the impact of his effort.

The Celtics didn’t trade Marcus Smart at the trade deadline. Either that means that other teams weren’t willing to part with a first rounder for him or the Celtics valued him more than other teams (or some combination of both). Regardless, he should be a valuable player on the court for the rest of this season.

The more interesting valuation is coming up this summer.