The best thing DeMar DeRozan could do this offseason is to improve his three-point shot. This is not groundbreaking, and it was also the refrain last off-season.
In 2012-13, he improved his three-point rate from 26.1% (92 attempts) to 28.3% (120 attempts). That’s an appreciated, albeit mostly insignificant, improvement. If nothing else, it shows he probably didn’t get worse. Over four years now, though, he’s at 23.9% (280 attempts) and really not leaving a lot of hope that he can develop a long-range game.
It’s paramount to his long-term success that he does, assuming you define success as being an above-average offensive weapon. DeRozan has a very good mid-range jumper and does a great job creating separation to get it off. Those shots, though, are more difficult to get than, say, a corner three. A player has to work for them, risk a turnover, and then are rewarded one less point than for a three. For DeRozan in particular, the threat of a three-point shot could improve his driving ability, too, by forcing the initial defender to stay closer to him.
With Rudy Gay in the fold, now, there are also less balls available for DeRozan, at least in theory (his usage rate actually went up after Jan. 30 when Gay was acquired). DeRozan capably adding a three-point shot will help space the floor for Gay to drive and vice versa.
There are plenty of reasons to hope DeRozan develops three-point range. But can he?
To find out, I went to my newly constructed “database” (read: giant Excel file) that contains all qualified player seasons in the three-point era. The first point of exploration is how the league as a whole moves their three-point percentage as they gain experience. The graph below plots all players with at least 50 field goal attempts in a given year against how many years into their career they were (note: for players who played before and after the introduction of the three-point line, “career year” is actually “year of three-point play”).
Surprisingly, experience doesn’t seem to have a large impact on three-point shooting. With the inclusion of survivor bias (whereby bad players retire early and poor shooters stop shooting threes), the expectation was that experience would factor in.
For DeRozan, this doesn’t mean he can’t or won’t improve, just that he can’t be expected to improve based on experience alone. For cases more similar to DeRozan, I spent an entire Saturday morning fiddling with my new toy (I’ve really gotta learn SQL) and found 62 instances of players posting three or more seasons with a three-point percentage below .300 (again, minimum 50 attempts). Of those 62, 34 went on to post multiple seasons above .330, so not all is lost. But that really just shows that three-point shooting can be fairly random year-over-year, since there was no consistent pattern of improvement.
To get a better handle on DeRozan’s comparables, I tried to filter it further to players who either shot below 30% or didn’t qualify in each of their first four seasons (DeRozan didn’t qualify as a rookie and then had three straight years of sub-.300 shooting). Since I was doing this by hand, I also allowed some leeway if a player had a low-30s rate in one of the seasons. I also removed players who didn’t qualify at all in their first few seasons.
This list is much smaller, trimmed down to 31 somewhat similar players. Have a look:
That’s a lot to take in. Reproducing the graph from earlier can help clear things up and answer the question of “do players who struggle shooting threes to start their career get better?”
It looks like experience for poor shooters helps. However, there are two factors at play here: poor three-point shooters stopping shooting, and regression to the mean, whereby extreme performances would be expected to move back towards a league mean (or a player-specific mean, though it’s unclear at present when three-point shooting percentage stabilizes at the individual level).
Of these 31 similar players, 10 would go on to have multiple seasons of 33% three-point shooting or better. However, the group combined to post 64 more seasons of sub-.300 shooting and just 46 seasons above .330.
That is to say, the odds are somewhat against DeRozan significantly improving; based on his start, he’s more likely to post a sub-.300 season than one above .330. There are examples, of course, but counter-examples, too. It’s impossible to quantify things like offseason work, mechanic changes and more – whether or not DeRozan improves has a lot to do with him and not a whole lot to do with comparable players.
I also have a feeling that if data went back far enough, mid-range shooting percentage would be a strong indicator of future three-point shooting. DeRozan is above-average in terms of mid-range percentage, though not elite, hitting on 16-to-23-footers at a 40 percent clip.
Even without that extra data, history shows that it’s at the very least possible for DeRozan to turn into a useful long-range threat, if not a sharp-shooter.