sohambhatia

Cal '18, UMich MBA '27. Growing @ForecastLabsAI

[ForecastLabs AI @ForecastLabsAI]

@WashWizards @ @ATLHawks tonight at 7PM ET. @TheTraeYoung revenge game after a loss in DC earlier this week. Your heart says he drops 40. The line still has the Hawks around 73%. Your heart doesn't have a Brier score. Predict yours @ https://t.co/LMIRt80zml

[ForecastLabs AI @ForecastLabsAI]

The WNBA regular season is over. We graded every forecast against it — our model, and the humans on our season board.

dakotamonster finished #1: 0.1952 Brier. The closing line: 0.1962.

A thousandth of a point off the sharpest number in sports, for a full season.

[ForecastLabs AI @ForecastLabsAI]

Most sports models share one massive blind spot: they work at the team level, putting one number per jersey. They can tell you how good a team has been, but they cannot see who is actually wearing the jersey tonight. Today, we are fixing that player-shaped hole. 🧵👇

[ForecastLabs AI @ForecastLabsAI]

Overconfidence is the human factory setting. In classic calibration studies, when people claimed they were 90% sure of something, they were only right about 75% of the time.

There is no scoreboard in ordinary life for the sentence "I'm sure". So we built one. 🧵👇

https://t.co/SqMl8Cgi5A

[ForecastLabs AI @ForecastLabsAI]

RAVEN went 2-1 in the WNBA last night. Brier 0.126.

This season: 183-89 across 272 graded forecasts, Brier 0.210. On pace with the market's final price.

Every pick is timestamped before tip-off and graded in public: https://t.co/QHsElE52um

[ForecastLabs AI @ForecastLabsAI]

Sight is not foresight.

When we launched ForecastLabs, our logo was a telescope. Philip Tetlock’s Superforecasting was our bible, and the telescope felt inevitable—the ultimate instrument for seeing further.

After a year of building, we realized why that logo was fundamentally wrong. 🧵👇
https://t.co/lZdyU21pCt

[sohambhatia @sohambhatia]

@thejaikannan Skill issue

[sohambhatia @sohambhatia]

@Selenagra13579 @gggiants You want to trade devers for relievers??

[ForecastLabs AI @ForecastLabsAI]

Most sports forecasting relies on a single eye: a standalone model, market consensus, or pure gut feeling. But depth requires a completely different architecture.

Introducing RAVEN: Three eyes. One signal. 🧵👇
https://t.co/8IFvVo489y

[ForecastLabs AI @ForecastLabsAI]

New SMS feature for game-time updates or changes! Check it out here.... https://t.co/8EmxNLzfl6

[ForecastLabs AI @ForecastLabsAI]

Most sports models have a massive, player-shaped hole. Standard models work at the team level—one number per jersey. Pull an MVP right before tip-off, and the model doesn't flinch. It keeps calling them a title team until enough losses drag it down. To fix this, we built PRISM. Here is the math behind how we price individual player impact. 🧵👇

[ForecastLabs AI @ForecastLabsAI]

The wait is over. The WNBA is officially LIVE on ForecastLabs!

With the 2026 season tipping off TODAY, there has never been a better time to test your prediction skills. Whether you're backing the defending champion Liberty, or analyzing the debuts of the Golden State Valkyries and Toronto Tempo, we’ve got you covered.

Spar with Brier to pressure-test your WNBA predictions, analyze player matchups, and find your edge against the market.

The 2026 @WNBA season is going to be massive. Make sure your prediction game is ready for it.

https://t.co/dYbfOe8fuf

[sohambhatia @sohambhatia]

@KimoNKC @HeritageUSA69 @FCA_World @IfindRetards

[sohambhatia @sohambhatia]

@Timccopeland Nightmare blunt rotation

[sohambhatia @sohambhatia]

@HankPowell6 Had smu +5 thank fuck 🙏🏾

[sohambhatia @sohambhatia]

@QTRResearch Where we trimming lol

[sohambhatia @sohambhatia]

@SleeperSFGiants @CheddarRod

[sohambhatia @sohambhatia]

@matthew_sigel Neuberger's background at GloriFi isn't a concern?

[sohambhatia @sohambhatia]

@timkawakami lfggg

[sohambhatia @sohambhatia]

@Melt_Dem @TheStalwart Any shortcomings?