A model with R-squared near 0 can still give valid 90% prediction intervals - here's why (and the catch)

I recently calibrated a recovery-rate model that had only two weak features. Its point accuracy was almost nothing — R² basically zero. I expected its uncertainty estimates to be junk too.

Source: [Dev.to](https://dev.to/whatsonyourmind/a-model-with-r-squared-near-0-can-still-give-valid-90-prediction-intervals-heres-why-and-the-31jp)