A deep dive into 48 years of Milwaukee Bucks data reveals when teams typically show their true identity and when we can trust what we're seeing in the NBA season.
I tried to (quickly) run through the chunks of the season that showed the defining version of the eventual participants. Your deep dive on one team's history is a great way to explore variations and results.
I wonder how basketball's nature as a strong link sport effects the 10-game mark's significance. How many teams alter their best players or highest usage rates over the course of the last 7/8 of the season? How often does that coincide with a change in end result? (I know you didn't go into transactions here.)
Fascinating observation about strong link sports. The "when do we know" question gets even more layered when you start digging into player usage and configuration changes.
I like your point about tracking how teams alter their highest usage rates over the last 7/8 of a season. Exactly the kind of rabbit hole I love. The current "True Bucks" metric is admittedly simple – no adjustments for coaching changes, player shifts, or schedule strength. But your suggestion opens up a whole new avenue of investigation.
How often do teams fundamentally reconstruct themselves mid-season? And more importantly, how does that reconstruction correlate with end results? That's the kind of question that makes me want to dive back into 48 years of box scores, lol.
The meta-analysis here is what fascinates me: not just "when do we know", but "how do we know", and "how much can a team's identity actually shift?"
Would love to hear more about what patterns you might be seeing or hypothesizing.
The idea of knowing the "true" version of a team fascinates me. I tinkered with a look at a different angle on this idea before the 2024 MLB wildcard round: https://thrillshot.substack.com/p/mlb-narrative-bracket-wildcard-edition?r=10qtpx
I tried to (quickly) run through the chunks of the season that showed the defining version of the eventual participants. Your deep dive on one team's history is a great way to explore variations and results.
I wonder how basketball's nature as a strong link sport effects the 10-game mark's significance. How many teams alter their best players or highest usage rates over the course of the last 7/8 of the season? How often does that coincide with a change in end result? (I know you didn't go into transactions here.)
Hey! Thanks for reading!
Fascinating observation about strong link sports. The "when do we know" question gets even more layered when you start digging into player usage and configuration changes.
I like your point about tracking how teams alter their highest usage rates over the last 7/8 of a season. Exactly the kind of rabbit hole I love. The current "True Bucks" metric is admittedly simple – no adjustments for coaching changes, player shifts, or schedule strength. But your suggestion opens up a whole new avenue of investigation.
How often do teams fundamentally reconstruct themselves mid-season? And more importantly, how does that reconstruction correlate with end results? That's the kind of question that makes me want to dive back into 48 years of box scores, lol.
The meta-analysis here is what fascinates me: not just "when do we know", but "how do we know", and "how much can a team's identity actually shift?"
Would love to hear more about what patterns you might be seeing or hypothesizing.