Tiers for Fears? Partnow's Predictions for us
Posted: Tue Aug 22, 2023 2:17 am
Long post incoming. Please to read, and react.
Some of you may know, I'm a fan of TheAthletic's Player Tiers Project, authored by Seth Partnow. You may also remember that I make my own season projections based on this list.
Basically, my idea is that a team's best player is responsible for 40% of their wins in the coming season. A team's 2nd-best player correlates to 30%, a 3rd banana to 20%, and a 4th guy to 10%. Beyond that, every team's depth has about the same amount of impact to their team's success, 0%; they essentially cancel each other out.
So, if you have a top-5 player in the world, you'll get a lot of credit for wins based on that 40% alone. But if the rest of the team is shaky, that weak 60% may still not amount to much overall (see: Mavericks, Dallas). Conversely, a team might have an overabundance of top-100 players, but not a single top-30 player. That would cost you in my system because just about every other team has a better "best player" than you (see: Pelicans, New Orleans). I think this plays out in real life.
Finally, a player's value is only fixed by what role they have on their given team. The best player in the NBA is assigned 30 points, while the worst #1 option gets 1 point. That point then is divided by their role, to get the # of wins above a replacement player (WARP) they generate. Jayson Tatum is the 10th player on Seth's list (he swears it isn't a ranking but cmon). Devin is the 11th player. Jayson is the 10-best #1 option, worth 21 points, while Devin is THE best #2 option, 30 points. Tatum yields 21 x .4, 8.4 WARP for Boston. Booker yields 30 x .3, 9.0 WARP for Phoenix.
SO WHAT DOES ALL THIS MEAN FOR BROOKLYN???????
Well, it's not good. There are only 1230 wins to go around each season for all 30 teams. My model based on Seth's rankings yields 29 wins and 53 losses, 11th in the East. Based on some other projection models, I have seen us at 31 or even 32 wins, but this is a far cry from the lofty aspirations I've seen for us. Mikal is real, and he might even outplay his tier level if we're lucky. Still, our supporting cast is one of the worst among teams vying for the playoffs. That's gonna hurt us on nights where Bridges is out or just doesn't have it.
Other key takeaways: Seth's projection is HUGELY skewed toward the Warriors. I don't think he's a Golden State fan or anything, but the Tiers filtered through my system yields 24-25 more wins for GSW than other advanced statistical models. Seth admits he tweaks his rankings based on proven success, but all of the indicators are that he's gonna be very wrong on this one. Having Curry rated as a top-3 player, when other models ranked him just outside the top 10, was curious. Listing Draymond as a top-40 player, when most outlets barely gave him a top-125 impact, was egregious. He had Paul, Wiggins and Klay much higher than their statistical impact, too. It's just... weird.
He's also 14 wins higher on Minnesota (47-35), and that looks like a poor bet based on their clunky fit. Edwards may be a superstar soon, but he's not on the level of Jaylen Brown and Donovan Mitchell quite yet, in terms of knowing how to win. And Rudy as a top-35 player? IONO.
Finally, Seth's project through my filter yields a 32-50 record for the New York Knicks. I would love it, but that's a 14 wins under other projections through my filter. Julius is a notorious regular season stat padder and playoff choker, and Seth's list values playoff performance more than the regular season. Still, I thought his positioning was harsh.
Thanks for reading. I've spent a lot of time with this, so let me know where my blind spots are!
Some of you may know, I'm a fan of TheAthletic's Player Tiers Project, authored by Seth Partnow. You may also remember that I make my own season projections based on this list.
Basically, my idea is that a team's best player is responsible for 40% of their wins in the coming season. A team's 2nd-best player correlates to 30%, a 3rd banana to 20%, and a 4th guy to 10%. Beyond that, every team's depth has about the same amount of impact to their team's success, 0%; they essentially cancel each other out.
So, if you have a top-5 player in the world, you'll get a lot of credit for wins based on that 40% alone. But if the rest of the team is shaky, that weak 60% may still not amount to much overall (see: Mavericks, Dallas). Conversely, a team might have an overabundance of top-100 players, but not a single top-30 player. That would cost you in my system because just about every other team has a better "best player" than you (see: Pelicans, New Orleans). I think this plays out in real life.
Finally, a player's value is only fixed by what role they have on their given team. The best player in the NBA is assigned 30 points, while the worst #1 option gets 1 point. That point then is divided by their role, to get the # of wins above a replacement player (WARP) they generate. Jayson Tatum is the 10th player on Seth's list (he swears it isn't a ranking but cmon). Devin is the 11th player. Jayson is the 10-best #1 option, worth 21 points, while Devin is THE best #2 option, 30 points. Tatum yields 21 x .4, 8.4 WARP for Boston. Booker yields 30 x .3, 9.0 WARP for Phoenix.
SO WHAT DOES ALL THIS MEAN FOR BROOKLYN???????
Well, it's not good. There are only 1230 wins to go around each season for all 30 teams. My model based on Seth's rankings yields 29 wins and 53 losses, 11th in the East. Based on some other projection models, I have seen us at 31 or even 32 wins, but this is a far cry from the lofty aspirations I've seen for us. Mikal is real, and he might even outplay his tier level if we're lucky. Still, our supporting cast is one of the worst among teams vying for the playoffs. That's gonna hurt us on nights where Bridges is out or just doesn't have it.
Other key takeaways: Seth's projection is HUGELY skewed toward the Warriors. I don't think he's a Golden State fan or anything, but the Tiers filtered through my system yields 24-25 more wins for GSW than other advanced statistical models. Seth admits he tweaks his rankings based on proven success, but all of the indicators are that he's gonna be very wrong on this one. Having Curry rated as a top-3 player, when other models ranked him just outside the top 10, was curious. Listing Draymond as a top-40 player, when most outlets barely gave him a top-125 impact, was egregious. He had Paul, Wiggins and Klay much higher than their statistical impact, too. It's just... weird.
He's also 14 wins higher on Minnesota (47-35), and that looks like a poor bet based on their clunky fit. Edwards may be a superstar soon, but he's not on the level of Jaylen Brown and Donovan Mitchell quite yet, in terms of knowing how to win. And Rudy as a top-35 player? IONO.
Finally, Seth's project through my filter yields a 32-50 record for the New York Knicks. I would love it, but that's a 14 wins under other projections through my filter. Julius is a notorious regular season stat padder and playoff choker, and Seth's list values playoff performance more than the regular season. Still, I thought his positioning was harsh.
Thanks for reading. I've spent a lot of time with this, so let me know where my blind spots are!