Want more ways to catch up on the latest in Bay Area sports? Sign up for the Section 415 email newsletter here and subscribe to the “Section 415″ podcast wherever you listen.
No matter how the Warriors’ season ends, the franchise has a lot of work to do this summer.
As The Standard’s Tim Kawakami noted, Golden State has just four healthy players signed for next season: Steph Curry, Brandin Podziemski, Gui Santos, and Will Richard. Everyone else is either injured, has a player option to decide on, is set to enter restricted free agency, or is about to hit the open market.
Perhaps the two biggest decisions involve new center Kristaps Porzingis and veteran guard De’Anthony Melton. Each player’s circumstance is vastly different.
Porzingis’ injury history and bouts with POTS have limited his availability. But he has had one of his healthiest and most productive stretches of the season recently, in the middle of which he called Warriors director of sports medicine and performance Rick Celebrini “the GOAT.” He has averaged 18.1 points per game in March, playing in 10 games – his most in any month since last year’s playoffs, when POTS zapped his energy.
Porzingis is on an expiring $30 million deal, meaning he’s a pending unrestricted free agent. One league source told The Standard that he wouldn’t expect a team would be willing to offer a deal longer than two years given his health bill. Another mused that a team could look to use more flexible contract language including partially guaranteed money tied to incentives, like the Pelicans did with Zion Williamson (opens in new tab) (Jonathan Isaac and Lonzo Ball are other recent examples).
Melton, meanwhile, has impressed in his first season back from his torn ACL, averaging 12.7 points in 43 games (19 starts). He remains an excellent option in the backcourt next to Steph Curry, but has struggled to consistently lead Golden State’s offense without its star.
Melton will almost certainly decline his $3.4 million player option to enter free agency. He’s an analytics darling, and compares well with Nickeil Alexander-Walker, Cason Wallace, and Ayo Dosunmu. Their contracts range from $7 million to $15 million annually.
Unless the Warriors get creative with their cap management, it’s possible that they can only bring one of Melton or Porzingis back.
But that’s assuming both Melton and Porzingis are each worth most, if not all, of the $15 million non-taxpayer midlevel exception. They probably both are, but informed guesses (including mine) around the NBA aren’t ironclad.
Educated speculation is nice, but more data points are always helpful. So I did what many 27-year-olds in San Francisco do: I started to vibe-code.
The goal was to create a proprietary model that factors in past and current performance, league trends, the CBA, and injury risks to predict future contracts for players like Porzingis and Melton.
I have zero coding experience and consider myself an AI skeptic. Before this week, I’d never used machine learning for anything of substance. But I realize the tools can be powerful, especially when analyzing large sets of data. Heck, pro teams — even the Warriors (opens in new tab) – use them.
After going back and forth with Claude for a few hours at Caffe Trieste in North Beach, then troubleshooting a bit with Claude Code and briefly consulting my techie roommate, KRISTAPS (Knowledge-Rich Intelligent System for Tracking and Analyzing Player Salaries) was born. (opens in new tab)
The algorithm accounts for raw statistics and some advanced ones, plus injury history, age, and accolades. Sometimes it’s better at pulling stats through than others. Prediction results can vary because the tool does a fresh search for each query, sampling probabilistically.
It’s trained on league trends and the CBA rules, but appears to either occasionally misinterpret or over-index certain dynamics.
Today
4 days ago
Tuesday, Mar. 24
For instance, Rockets guard Amen Thompson will be eligible for a rookie extension this summer. Based on league precedent, a player of his caliber and trajectory would almost certainly be in line for a max extension. Instead, KRISTAPS projected a four-year, $112 million deal.
Many of the system’s predictions, like that for Thompson, feel like they’re on the conservative side. But, in the apron era, perhaps teams will start to spend more pragmatically.
The model also seemed to struggle with historical outliers like LeBron James. At 41, he’s about to become an unrestricted free agent. He’s among the most decorated players in league history and has never taken a significant pay cut, but the tool spit out a $2.5 million prediction. That’d be great for contending teams like the Warriors, but is wholly unrealistic on its face.
Those caveats aside, KRISTAPS seemed to do a fairly nice job with the notable Warriors – Porzingis, Melton, and rookie extension-eligible Podziemski.
Porzingis’ estimated contract was three-years, $37.5 million. That annual value of $12.5 million would sop up most of the non-taxpayer midlevel exception. A comparable contract it spit out was the three-year deal Harrison Barnes inked when he was 31.
Melton, meanwhile, earned a three-year, $28.5 million projection. That $9.5 million annual value fits in the range of complementary starting guards or super bench options. Guards currently making roughly $9.5 million include Alex Caruso, Sam Merrill, Aaron Wiggins, and Ty Jerome.
Podziemski’s rookie contract extension negotiations could be rather interesting. He has largely struggled as the lead ballhandler of Golden State’s Curry-less offense, but has proven to be a trusted, valuable contributor in his three seasons. He plays strong team defense, is an excellent rebounder for his size, and is growing as a shot creator.
The Warriors have Podziemski signed through 2026-27, but working toward an extension this summer would avoid the possible pains of restricted free agency two summers from now.
What does KRISTAPS think a possible extension could look like? A four-year, $74 million pact.
“Podziemski is a 3YOS extension candidate with improving efficiency and elite perimeter shooting as a winning archetype,” KRISTAPS reasoned. “At (23), he represents a cost-controlled extension window before restricted free agency. Four-year structure typical for extensions; $18.5M AAV reflects: strong 3-point volume shooter, defensive versatility, proven health, and fit in winning systems. Below star tier but well above replacement level. Positions team for cap flexibility with structured deal.”
Sounds about right. For the Warriors, at least.
Teams use models assuredly much more advanced than our KRISTAPS passion project, but A.I. isn’t coming for Mike Dunleavy’s job any time soon. It’ll be up to him and the Warriors’ front office to retool next year’s roster in the real world, outside a silly algorithm.