Show HN: PolyBridge – a causal model for forecasting global events

Posted by pbforesight on Tuesday July 28, 2026 @09:45AM
from the show-hn dept.

Hi HN, I’m working on PolyBridge. Ask a question about a future event and it returns a probability, a credible interval, and the evidence behind the result.

API/docs: https://polybridge.ai/api

It combines prediction markets with other data sources in a probabilistic graphical model. We currently index more than 170k event contracts across major prediction-market venues.

The model handles current-state queries, forecasts, conditionals, interventions, and counterfactuals.

For example, there may be a market on whether the Fed cuts rates in September, but not on how a spike in oil prices would change those odds. The model estimates the change using the relationships between rates, inflation, oil prices, recession risk, and other evidence.

We also publish Situation Rooms, which are curated pages for following a topic over time. Here is one for the 2028 presidential race:

https://polybridge.ai/situation-rooms/2028-race-to-the-white...

The API includes Search for finding relevant evidence and Forecast for returning a probability and the reasoning behind it.

I’d appreciate feedback on the API, the Situation Rooms, and the kinds of questions you would test.

Re:Show HN: PolyBridge – a causal model for forecasting glob... by zippolyon on Thursday July 30, 2026 @11:34AM (#49111454)

You mentioned that the system supports interventions and counterfactuals. I am curious: does it simply recalculate a predicted probability based on the intervention, or does it separately log the target, current context, intervention, actual outcome, and the residual between the outcome and the target? If these are logged, is the residual subsequently recalculated using the actual outcome once the event has occurred?

[ Parent | on HN ]