aratea.dashboard

Predictor — learning loop

Aratea is a weather-factor discovery engine. Every named feature here is a hypothesis; every training run measures whether it carries signal. The bench is the same row-set kalshi_mid Brier — beat the market, on its own ground.

What Aratea thinks today

Daily low temperature · LAX

May 19, 2026 · B57.5

Bet open — outcome not yet resolved
Today's question

Will the lowest temperature in Los Angeles on May 19, 2026 fall in the 57.5°F bin?

KXLOWTLAX-26MAY19-B57.5

Aratea's chance the answer is YES
26%
Aratea is betting NO

Bet placed by the “vendor_ensemble” model (current champion).

Aratea thinks YES is 19.9 pts less likely than the market does.

Aratea estimates YES at25.6%
Market is pricing YES at45.5%
ConfidenceMedium

The model has a clear lean, but not extreme.

How to read this card

The big number is the probability Aratea gives to the answer being yes. The market number is what people on Kalshi are paying for the same bet right now. When the two disagree, Aratea takes the side it thinks is mispriced. This is paper-money only during Phase 1 — no real bet is placed.

Why this model and not another?

Aratea runs several models in parallel. Only the champion places real (paper) bets — challengers run in shadow until they beat the champion's track record. Other models may estimate a different probability — see Level 2 to compare.

Why does Aratea think this? →