Here are two factual statements about Marcus Smart:
- He is a terrible shooter.
- His terrible shooting significantly improves the Celtics’ offense.
Through more than a quarter of the season, Smart is shooting 29.6 percent from downtown, 34 percent inside the arc, and 36 percent on shots inside of 5 feet. Among players who attempt more than six field goals a contest, only Lonzo Ball and (barely) Frank Ntilikina have a lower effective field goal percentage.
This is an especially bad year for Smart, but he’s never been an expert marksman. He’s a career 35 percent shooter from the field, 29 percent from downtown, and 41 percent on two-pointers.
And yet, that terrible shooting is somehow essential to the Celtics’ offense.
When Smart is in the game, the Celtics score at a rate of 109.3 points per 100 possessions. When he is not in the game, Boston scores an average of 101.1 points per 100 possessions. Only Al Horford — and crucially not Kyrie Irving — has a more significant effect on Boston’s offensive efficiency than Smart.
Most of that difference is made up in the Celtics’ shooting. Boston attempts better shots when Smart is in the game and makes those shots more often. Sixty-three percent of its buckets are assisted with Smart in compared to less than 57 percent with him out. The Celtics’ effective field goal percentage as a team rises to 53 percent with Smart in and drops to 50.3 percent with him out. They shoot a higher percentage of their shots from three-point range and convert restricted area attempts at a clip 8 percent higher than with Smart out of the game.
How can these two factual statements both be true? Because opponents treat Marcus Smart like the player he pretends to be, not the one he actually is.
Fake it till you make it (even if you never do)
Marcus Smart may be a sub-35 percent shooter, but you’d never know it watching him play. When he’s open — and often when he’s not — he’s firing away. He’s attempting nearly five three-pointers a contest, and only three Celtics (Irving, Marcus Morris, and Jaylen Brown) have a higher usage rate.
He doesn’t just stand there and take the shots given to him, either. More than 55 percent of his shot attempts come off at least one dribble, a percentage higher than Brown or Horford. And it’s not like Smart only attacks to score; he’s just behind Horford in third place on the team in assist percentage and nearly 13 percentage points ahead of the next player in the regular rotation.
This is not the offensive profile for a 32-percent shooter. Consider a comparison between these two players:
Shot distribution comparison
|USAGE||CATCH/SHOOT||PULL UP||INSIDE 10 FEET||TS%|
|USAGE||CATCH/SHOOT||PULL UP||INSIDE 10 FEET||TS%|
Player 1 is Smart. Player 2 is Washington forward Otto Porter. Their shot portfolios are about the same. Their roles are about the same. The only difference is that Porter actually makes shots.
Smart ... doesn’t. And that should crush the Celtics, because that’s a ton of possessions ending in the hands of a player that converts his attempts to put the ball in the hoop less frequently than Jose Altuve makes successful contact with a baseball.
And yet, it has the exact opposite effect. Why?
You could go deep on this, as CBS Sports’ Matt Moore did last month. But the answer lies in a single sentence from that article:
With Smart, no matter how he performs, they play him the same way: as an offensive threat.
In other words, he’s so good at pretending to be a legitimate offensive player that he cons defenses into thinking he actually is. He has so overwhelmingly thumbed his nose the obvious reality that he cannot shoot that the opponent can’t help but react to his overconfidence in real time.
Because of that, he is defended as if he is a legitimate shooter. And the beauty of legitimate shooters is that the threat of them making shots sucks in defensive attention and opens opportunities for others.
Just look at these clips
Here are a series of Bucks players running Smart off the line and swarming like he’s a superstar.
Here’s Serge Ibaka flying out for no reason.
Here’s Devin Booker giving up an easy drive instead of letting Smart take a 28-footer.
Here’s Wesley Matthews reacting as if Smart shooting a corner three is as devastating as Steph Curry doing the same.
Here’s Luke Babbitt leaving 50-percent three-point shooter in Jayson Tatum to supply unneeded rim protection on a Smart drive.
But Marcus Smart is a 32 percent shooter. Don’t these players know this?
Maybe they don’t. Maybe they’re actually watching Netflix instead of the curated playlist their video coordinator has supplied. Maybe the coaches don’t know where to find basic statistical information that took me a couple minutes to look up.
Or maybe Smart’s success should remind us that basketball players must make snap decisions on the court incredibly quickly. They don’t get time to hit pause, carefully identify the shooter, and double-check his catch-and-shoot percentage on the NBA’s stats site to determine the precise speed that rotation requires. They have to immediately plot a course of action, usually split seconds after immediately plotting six other courses of action on a single possession.
This is exceedingly difficult, especially as NBA teams push the limits of three-point shooting. With big men stepping out and guards pulling up from well behind the line, defenders have to cover a significantly bigger slice of the court than ever before, using the same number of people as they did 20, 30, 40, and 50 years ago. This is especially true when playing the Celtics, who zip the ball around the court and make defenders dizzy with off-ball cuts.
That means more snap judgments, all while working harder.
When that happens, the best story takes over, not the most accurate
In his landmark book Thinking Fast And Slow, Nobel Prize-winning psychologist Daniel Kahneman introduces two ways in which our brain processes information.
System 1, as he calls it, is how we make instinctual decisions. It is independent, automatic, and often subconscious, but it also can be prone to systemic biases.
System 2, on the other hand, is how we think more deeply. It is more rational and calculating, but also moves slower. It can check the negative effects of System 1, but is lazy and won’t invest the massive effort unless it’s absolutely necessary. (Think of it like a smartphone with an unlimited data plan, but terrible battery life).
In the book, Kahneman speaks of a process he calls substitution. When we are presented with a complicated question, our System 1 replaces that question with an easier one that can be answered quickly. System 2, lazy as it usually is, then rubber stamps the easier question that System 1 has answered, thinking it has actually answered the more complex question. We are especially prone to this process when we are tired or when we’re focused on another task.
Why does this have anything to do with Marcus Smart?
Because NBA players defending him incorrectly are victims of substitution. As they rotate to cover Smart, they are faced with a complex question: To what degree should they close out based on this player’s catch-and-shoot tendencies and three-point proficiency, as measured by a host of statistics?
That is an impossible question to answer accurately in a split second. So instead of answering that question, they answer a simpler one: Does this dude look and act like a shooter?
Marcus Smart looks and acts like a shooter. By constantly firing away and taking the same shots as a great spot-up shooter, Smart has created an easy narrative for his opponent. They’ve closed out on so many players that act this exact way, but actually make the shots. When they see Smart, their System 1 identifies him as one of those players, not a 32-percent bricklayer that could build a mansion with all his misses.
Because of that, Smart is able to take advantage of all the benefits a great shooter has on an offense without actually being a great shooter. He can attack his off-balanced defender, draw help, and kick to open shooters.
He can draw two to him in pick-and-roll and find the open man, even without being a scoring threat.
He creates enough doubt to cause a help defender to react slowly to someone else’s cut.
He is even treated like a real transition three-point threat, even at the expense of allowing a dunk.
All that activity creates a vision in defenders’ minds, particularly when they are tired from chasing everyone else around and grounded down by the NBA’s unforgiving regular-season schedule.
That’s how the Celtics’ offense benefits from Smart’s hilarious overconfidence.
Perception isn’t always reality, but actually it kinda is
Smart’s unique situation leads to an obvious question: Do facts actually matter on the basketball court? (This question could be applied to any number of world concerns, but alas).
We have so many ways to quantify a player’s production these days, down to the most specific game situations. We know what a player shoots on average when he dribbles six times as opposed to four. We can isolate rebounding production by different zones of the floor. We can determine when players create assist situations and even when they create hockey assist situations.
But in the heat of the moment, players’ brains cannot tap into all that information. They have to make too many decisions while covering too much ground on either end of the floor. In those moments, they fall for those systemic biases and clean narratives that seem to make sense, even if they are factually incorrect. They fall for con schemes like Marcus Smart launching jumpers as if he’s actually a good shooter.
Smart is an edge case and it’s likely his positive effect on Boston’s offense will fade in the playoffs when opponents have more time to focus on him. But in the interim, his situation should teach analysts a valuable lesson. The game doesn’t happen in a vacuum. It happens in real time.
And in those situations, what matters isn’t necessarily what a player does. It’s what an opponent believes a player does.