Jamal Murray and Gary Harris have had stretches of incredible 3-point accuracy and stretches where they can’t hit anything. Are they the shooters Denver needs?
In 2017-18, the Denver Nuggets took an undeniable leap. They finished over .500 for the first time in four years, missing the playoffs by the slimmest of margins — a loss to the Timberwolves on the final night of the regular season put Minnesota in and pushed Denver out. That season laid the groundwork for what they are now, a rising contender, currently third in the Western Conference standings, and appeared to establish their foundational stars — Nikola Jokic, Jamal Murray and Gary Harris.
Jokic was and is an MVP-caliber offensive fulcrum. Murray complemented him with shooting and creation. Harris rounded out the trio with excellent perimeter defense and more outside shooting, creating as much space as possible for Jokic to work in the middle. Three years later, it’s all working, but the formula is not at all what it appeared to be, particularly the shooting numbers for Murray and Harris.
In that 2017-18 season, Murray and Harris made 37.8 and 39.6 percent of their 3-pointers, respectively. Both players saw their percentages decline last year and both have bottomed out this season — Murray at 32.2 and Harris at 31.0 percent.
The Nuggets have thrived because Jokic is sensational, the defense has improved, Murray and Harris are still solid, and players like Jerami Grant, Monte Morris and Will Barton have filled the gap as floor spacers in the rotation. But the Nuggets’ ceiling and long-term championship hopes are pinned, at least somewhat, to the ability of Murray and Harris to be an elite backcourt combination. That just can’t happen if they are both below-average shooters from beyond the arc.
The interesting thing is that, season-by-season, both players have had percentages all over the map. Harris shot 35.4 percent on 297 attempts in 2015-16, 42.0 percent on 255 attempts in 2016-17 and 33.9 percent on 242 attempts two years later in 2018-19. Murray shot 33.4 percent as a rookie and then had two strong seasons before crashing back to earth again this year.
There are a few external variables that can help explain some of the variations. About 47 percent of Murray’s attempts were pull-ups this season, with the rest as catch-and-shoot attempts. That’s a fairly big shift from 2017-18, when 40 percent of his attempts were pull-ups. We see the same trend for Harris, whose 3-points splits have gone from about 24 percent pull-ups to about 32 percent this season.
While the mix of 3-point attempts they take will always affect their final numbers, the huge swings that can’t be completely explained by shot selection alone. Some of this variation is noise, but given these enormous shifts from season-to-season how can we estimate the “true” ability level of each player as an outside shooter?
Enter, the DARKO projection system which, in the words of creator Kostya Medvedovsky, addresses some of the more micro questions that aren’t answered by all-in-one player metrics.
"These metrics don’t answer questions such as “how good a three-point shooter is Jaylen Brown?” or “how many two-point attempts can we expect Marcus Smart to take?”DARKO (Daily Adjusted and Regressed Kalman Optimized projections) is an attempt to fill that gap. As will be familiar to baseball fans, DARKO is a basketball projection system similar in concept to Steamer, PECOTA, and ZiPS. To my knowledge, it is one of the few public computer-driven NBA box-score projection systems (as opposed to the “hand-curated” systems offered by some fantasy basketball sites)."
Medvedovsky’s system incorporates new box score data daily to create updated estimates of a player’s “true talent” level in several statistics. The mechanisms of the system, explained here, account for a variety of factors in their estimates, including recency, opponents, aging, seasonality, rest and home-court advantages. They have their noise and are, by nature, continuously evolving estimates, but they can be helpful in understanding the evolution and present state of a player’s skill.
The graph below shows the “true talent” progression for the 3-point shooting of Murray and Harris.
The DARKO system doesn’t use NCAA or international data to make projections for rookies so the model is trained on each player by their early results. The early struggles of Murray and Harris — 33.4 percent on 344 attempts as a rookie for Murray, 31.5 percent on 400 attempts across his first two seasons for Harris — helped shape the model’s understanding of them.
Even as Harris peaked at over 40 percent on 3-pointers in 2016-17, the model estimated that his true ability level was closer to 37.0 percent. As he’s continued to struggle since then, the estimate has dropped off. The same thing is true for Murray, to a lesser degree, and you can see it reflected even more strongly in the graph below which shows the rate of change in their “true talent” 3-point percentage over time.
This is just one model and can be taken with a grain of salt, but the current estimates here make sense intuitively — Harris and Murray struggling this year is an outlier, but they’re not elite shooters either. The estimates for both player’s “true talent” 3-point percentage is around 36 percent, below their career peaks but much higher than what they’re shooting right now.
Some progression to the mean and some positive tweaks to shot selection could have them bouncing back towards that number before the end of the season, something that could pay huge dividends for Denver in their hunt to make the Finals. All else being equal, if both players had been shooting 36 percent on 3s all season long, the Nuggets point differential would jump from its current plus-3.4 points per 100 possessions to plus-4.5, leaping past both the Rockets and Jazz.
The Nuggets will need more than just better shooting from Murray and Harris for a deep playoff run, but it would certainly help and it’s probably something they should expect to see eventually.