I found the recent debates on counting stats vs. contribution to winning interesting (but admittedly also infuriating) so I took a quick look at historical data for Rookie of the Year voting from 2004 to 2023. TL;DR at the end.
Before I share my findings, some notes about the data I used:
1. All data is sourced from the [ROTY page on Basketball-Reference](https://www.basketball-reference.com/awards/roy.html).
2. I chose the 2003-2004 season as the starting point primarily because I was lazy but also because of the oft-cited race between Bron and Melo.
3. Only the top 3 rookies in terms of points received were included because I, again, was lazy.
Now onto my findings:
**The rookie leading in points per game became the ROTY in 13 out of 20 seasons.** The following are the seven who led in points per game but lost the ROTY race:
* Cade Cunnigham (17.4 ppg) behind Scottie Barnes (15.3 ppg) and Evan Mobley (15 ppg)
* Anthony Edwards (19.3 ppg) behind LaMelo Ball (15.7 ppg)
* *Zion Williamson (22.5 ppg)* behind Ja Morant (17.8 ppg) and Kendrick Nunn came (15.3 ppg)
* Donovan Mitchell (20.5 ppg) behind Ben Simmons (15.8 points with 8 rebounds and 8 assists per game)
* *Joel Embiid (20.2 ppg)* behind Malcolm Brogdon (10.2 ppg) and Dario Šarić (12.8 ppg)
* O.J. Mayo (18.5 ppg) behind Derrick Rose (16.8 points and 6.3 assists per game)
* Carmelo Anthony (21 ppg) behind LeBron James (20.9 ppg)
It is important to note, however, that both Zion and Joel played significantly fewer games than other rookies (24 and 31 respectively). Also interesting: two out of three redshirts (there might be more redshirts that I am forgetting) won the ROTY in their second season.
In examining the team records of these rookies, the candidates with the best team record won in five seasons: Scottie Barnes, LaMelo Ball, Ben Simmons, Malcolm Brogdon, and Derrick Rose. The only exceptions are Ja Morant (the Grizzlies were 34-39 while Nuinn’s Heat were 44-29) and LeBron James (the Cavaliers were 35-47 while Melo’s Nuggets were 43-39).
**Points per game is somewhat positively correlated to ROTY votes received (r = .52). Meanwhile, win percentage of the team is not at all associated with ROTY votes (r = -0.09).** As a meme, there is a negative correlation between ROTY votes received and FG% (r = -0.2), albeit very weak.
**In a quick multiple linear regression model I put together, points per game is the most significant predictor of ROTY votes, with total games played being a very distant second.** An [image](https://i.imgur.com/9VjETSM.png) of the table for your viewing pleasure. Win percentage appears to be useless in predicting ROTY votes. Other counting stats are also in the same boat. Curiously, I could only manage to make win share statistically significant in one specific configuration of all the models I tried.
Some caveats: the R-squared value is quite low for all models and the t-values also leave much to be desired. Further, this analysis does not say that the votes for the 2024 awards will follow the trends presented above. We cannot rule out the possibility that for this year specifically, voters will put emphasis on contributing to a winning team given the huge disparity in records between the Thunder and the Spurs. Of course, there are numerous issues with this position but that is a topic for another time.
TL;DR: PPG is a salient predictor of ROTY votes; winning and contribution to winning are somewhat significant.
by 10527189514
