Lessons from the Portland Trail Blazers' "Improvement"

Last week I campaigned for Russell Westbrook as the "Most Improved Player" in the NBA. I was unsuccessful as the voters decided C.J. McCollum was most improved. According to our numbers, last season McCollum was worth -1.1 Points over Par per 48 minutes. This season he was worth -1.1 Points over Par per 48 minutes. You may notice that not only is this not an improvement, but it's also below average. Compare that to our pick, Russell Westbrook, who improved to MVP status, which you may recognize as at least "a little" above average.

I can't say I'm shocked by McCollum as the consensus pick for Most Improved. That said, his award lets us talk some fun stories that are, unfortunately, routine at this point.

Shot Creation, Missing Shots

Last season the Portland Trail Blazers were a 50 win team. However, free agency was harsh to them as they were picked clean. A recap:

  • LaMarcus Aldridge, 19.2 True Shots per Team Game (52.8 % True Shooting), lost to the Spurs.
  • Wesley Matthews, 9.9 True Shots per Team Game (56% True Shooting), lost to the Mavericks.
  • Nicolas Batum, 7.8 True Shots per Team Game (51.6% True Shooting), lost to the Bobcats.
  • Robin Lopez, 6.0 True Shots per Team Game (57.4% True Shooting), lost to the Knicks.

What's more Chris Kaman, who took 7.2 True Shots per Team Game last season was out most of the season due to injury. Outside of Damian Lillard, the Blazers lost five of their top six volume shooters. We've been told for years that Aldridge is a key part of the Blazers offense as he spaces the floor. So, let's take a look at the Blazers scoring losing tons of "shot creators" and a floor spacer.

Stat 2015 Blazers (Rank) 2016 Blazers (Rank)
Offensive Rating 108.2 (8th) 108.8 (6th)
True Shooting % 54.4% (9th) 54.8 (10th)
Field Goals per Game 86.0 (9th) 85.9 (8th)
Free Throws per Game 19.4 (28th) 23.0 (14th)

Holy cow! The Blazers, on offense, were close to if not the same team they were last season. What's more, they didn't seem to lose any shots! Here was what happened:

Returning Blazers Change in Shots

Player 2015  TS/TG 2016 TS/TG TS% Delta
C.J. McCollum 4.8 18.7 +1.0% (53.4% to 54.4%)  
Allen Crabbe 2.0 8.9 +5.0% (52.4% to 57.1%)
Meyers Leonard 3.2 5.5 -7.0% (63.1% to 56.1%)
Damian Lillard 18.7 20.5 0.0% (56.0% to 56.0%)
Tim Frazier 0.3 0.8 -17.0% (55.7% to 38.3%)

The easiest explanation of what happened is McCollumn and Crabbe were able to pick up over half of the "lost" shots from the leaving Blazers. The Blazers also picked up several new players that were able to pick up most of the remaining shots with varying levels of success:

  • Al-Farouq Aminu, 9.6 True Shots per Team Game (53.2% True Shooting)
  • Mason Plumlee, 8.1 True Shots per Team Game (56.4% True Shooting)
  • Gerald Henderson, 7.2 True Shots per Team Game (52.6% True Shooting)
  • Maurice Harkless, 5.6 True Shots per Team Game (54.1% True Shooting)
  • Ed Davis, 5.2 True Shots per Team Game (61.6% True Shooting)

The key point I'll make is that the Blazers were essentially able to recreate their offense from last season by letting decent players shoot more (McCollum and Crabbe) and picking up several other good scorers in the offseason. It's remarkable that without Aldridge's shot creation, which we've often questioned, that the Blazers were not only able to take as many shots as last season, but do it slightly better.

As a slight tangent, we often bash "shot creators" like LaMarcus Aldridge, Carmelo Anthony, and Kobe Bryant. The reality is that shooting in the 54-56% True Shooting range is a highly replaceable skill in the NBA. Case in point, no one on the Blazers this season costs more than $7 million! Now, of course if a starter, even an overrated one, goes down midseason it can hurt your team if all of the backups are subpar (see Memphis.) But if the team gets decent players in a trade for a "star" (see Denver after Carmelo Anthony left) or they have an offseason to pick up reasonable replacements (see the Blazers) then they'll be okay. And this year's Blazers aren't some brand new story; we've seen this tale over and over with players like Allen Iverson, Carmelo Anthony, and now, LaMarcus Aldridge.

What Improvement?

Another odd note is as follows, McCollum's True Shooting improved by roughly one percent while his shots went up significantly. The NBA has an issue in that your performance is all relative. It doesn't matter if you score a hundred points a game if the average NBA team scores a hundred and ten. Last year the average NBA Wing shot 53.2% True Shooting. This season? The average Wing shoots 53.8% True Shooting! That means that when we look at Lillard and McCollum, who have improved their scoring output, that their "improvement" is a little laggy when we adjust for inflation! And as mentioned, even after his improvement, McCollum is shooting at league average (when we look at all positions.) It's not a surprise that McCollum got the vote as his points went up. However, the trick here is that he just took more shots. And despite the odd obsession with the "Usage Curve", the history of the NBA doesn't seem to indicate a hard fast rule between the shot amount and shooting percentage

Getting a twenty-four-year-old shooting guard to take twenty shots a game at a slightly above average clip is not impressive! When the player was already shooting close to that efficiency already, it's hard to call it an improvement. In the NBA whenever a team takes multiple "underrated players" and "outperforms" expectations, it's evident that voters have a hard time figuring out what happened. When this happens the old stand by is to point to the player with the most points, or the biggest jump in points (see Lillard feeling neglected out of the All-Star game, and McCollum's MIP) But the real story behind the Blazers is much more fun. A decent offense lost a lot of their shots, and they were able to replace all of them in less than an offseason. It's a lesson I think more GMs might benefit from.



True Shots (TS) This is the total number of shots factoring in free throws as 0.44 shots.

Team Game (TG) This is when I average a total by 82 instead of games played for the player. The goal being to show how the player is performing "per game" at the team level for players that miss games.