Analytical Mailbag: Trouble in Cleveland?
"I mean, I never heard anybody say, 'I'm going to Cleveland on vacation.' What's so good about Cleveland?"
The long break is over - thank you to all who didn’t forget about my little blog after my extended vacation. Today, I'll be posting the first question from the mailbag Q&A - the next few weeks will be trying to answer as many of them as I can.
Hey TZM,
Thoughts on the Cavs? Should they trade someone?
Thanks for the question! It’s something I’ve thought about before in the past - but I wavered on writing a full article. Thinking about who the Cavs would be most likely to trade at this point - either Darius Garland or Jarrett Allen seem likely. But what do the numbers say?
I used on-off metrics from 2022-23 and 2023-24 on all four of the Cavs’ stars, with a filter of at least 300 minutes (150 minutes per season seems fair to adjust for…adjustment after injury plagues). The results had some surprises.
When Garland and Mobley share the floor together without overlap from their respective backcourt and front court partners, they succeed greatly. However, having watched the Cavs play basketball - I knew this was a crude measure of success. I resolved to find a better (but still crude) way to calculate contribution to the net rating of a team. Using the on-off data, I calculated Shapley Values for each player.
Shapley Values are a way to fairly distribute credit among participants based on their contribution to a group’s outcome. In basketball, it can be adapted to understand a player's contribution to the team across various lineups. The basic idea is to look at all possible combinations of players on the court and calculate how much each player individually adds or subtracts from the team’s performance.
By considering every possible subset of players, Shapley Values give us a more nuanced view of each player’s influence, even when they share the floor with others. It’s a more refined way to gauge a player’s impact compared to simple plus-minus stats, which can be noisy or misleading depending on who they’re paired with.
For the Cavs, I used this method to measure the contributions of key players like Garland and Mobley. While traditional stats show success when they share the floor, I knew this was too simplistic. Shapley Values let me break down how much each player actually drives the team’s success, offering a more direct, albeit still crude, measure of their true impact.
The Shapley Values showed what I thought - the young guys are still lagging behind their slightly older counterparts.
But that’s what youth entails in the NBA - nor is anyone expecting Darius Garland to be at the level of Donovan Mitchell anytime soon. Mobley has shown some very slight statistical improvement, over his three seasons - but in the playoffs, he really exploded after being played at the five.
Darius Garland himself had an under the radar jaw injury, and had to be fed through a tube, which undoubtedly effected his conditioning. His DARKO curve shows his sudden drop-off, which continued into the playoffs.
Despite all the valid excuses for their underwhelming play, I still find myself not as optimistic about the Cavs as I used to be. The core of my doubt remains in their offense, which ranks in the bottom third of spacing.
This shouldn’t be a surprising result to anyone - if you play two non-shooting bigs in your front court, you probably will not have great spacing. Trying to make up for that with Georges Niang and Sam Merrill might help a little, but the core issue remains. For what it’s worth, I’m still on the neutral side - I don’t anticipate any massive packages for Jarrett Allen given his injury history, and I think the Cavs would do good by giving Mobley one more season to really prove himself.
The Cavs are your prototypical above-average but not good enough NBA team, and a lot of it hinges on Mobley’s so far stagnant development. Perhaps the addition of Kenny Atkinson will be enough to juice him and the rest of the team this season.
What am I listening to?
I thought I’d throw a new segment of these mailbags where I just highlight a song I like. This past summer, I’ve been listening to a lot of J-pop. Hikaru Utada has been at the top of the genre for 25 years now - her debut Japanese language album, First Love, remains a massive success. Here’s my favorite track from that work.