Model lowers BWI departure delay risk 6–9 PM Friday to 78%
- Current state
- Resolved: closed without a new report (The forecast window has started; measured delays are reported separately as observed items.)
- Started
- Latest activity
- Updates
- 4
Tailodds reports what the FAA publishes, what we measure and what our model forecasts. It doesn’t control airport operations or change bookings; check with your airline about your trip.
Updates oldest first
-
Initial Forecast
Model forecasts 69% delay risk for BWI departures 6–9 PM Friday
Tailodds’ model gives the 88 departures scheduled from BWI 6–9 PM Friday an average 69% chance of leaving 15+ minutes late. With calm weather the same flights would average 49%. From the 7:00 AM EST forecast (model famA-fc-v3+famB-sd-v4i). This is a forecast, not a report of delays.
Evidence for this update
Target BWI departures 6–9 PM 2026-11-06 (15+ min late) Probability (calibrated, mean per flight) 69% over 88 flights Forecast cycle · horizon Fri Nov 6, 7:00 AM EST · 9.1 h ahead Model version famA-fc-v3+famB-sd-v4i Source event time Ingested Published here (2026-11-06 13:56 UTC) Sources Tailodds forecast Provenance forecast_blocks forecast-blocks-v1 · config 2026-10-17.1 · writer template-v1 Was this useful? -
De-escalation Forecast
Updated model puts BWI departure delay risk 6–9 PM Friday at 59%
What changed: 69% → 59% since the 7:00 AM EST forecast (model changed)
Tailodds’ model gives the 88 departures scheduled from BWI 6–9 PM Friday an average 59% chance of leaving 15+ minutes late. With calm weather the same flights would average 50%. From the 11:00 AM EST forecast (model famB-sd-v4i). The model changed since the previous item (from famA-fc-v3+famB-sd-v4i), so some or all of this change comes from the new model rather than new information. This is a forecast, not a report of delays.
Evidence for this update
Target BWI departures 6–9 PM 2026-11-06 (15+ min late) Probability (calibrated, mean per flight) 59% over 88 flights Forecast cycle · horizon Fri Nov 6, 11:00 AM EST · 6.1 h ahead Model version famB-sd-v4i (changed since the previous update) Source event time Ingested Published here (2026-11-06 16:56 UTC) Sources Tailodds forecast Provenance forecast_blocks forecast-blocks-v1 · config 2026-10-17.1 · writer template-v1 Was this useful? -
Escalation Forecast
Model raises BWI departure delay risk 6–9 PM Friday to 80%
What changed: 59% → 80% since the 11:00 AM EST forecast (same model, same window)
Tailodds’ model gives the 88 departures scheduled from BWI 6–9 PM Friday an average 80% chance of leaving 15+ minutes late. With calm weather the same flights would average 52%. From the 2:00 PM EST forecast (model famB-sd-v4i). This is a forecast, not a report of delays.
Evidence for this update
Target BWI departures 6–9 PM 2026-11-06 (15+ min late) Probability (calibrated, mean per flight) 80% over 88 flights Forecast cycle · horizon Fri Nov 6, 2:00 PM EST · 2.6 h ahead Model version famB-sd-v4i Source event time Ingested Published here (2026-11-06 20:26 UTC) Sources Tailodds forecast Provenance forecast_blocks forecast-blocks-v1 · config 2026-10-17.1 · writer template-v1 Was this useful? -
De-escalation Forecast
Model lowers BWI departure delay risk 6–9 PM Friday to 78%
What changed: 80% → 78% since the 2:00 PM EST forecast (same model, same window)
Tailodds’ model gives the 88 departures scheduled from BWI 6–9 PM Friday an average 78% chance of leaving 15+ minutes late. With calm weather the same flights would average 57%. From the 4:00 PM EST forecast (model famB-sd-v4i). This is a forecast, not a report of delays.
Evidence for this update
Target BWI departures 6–9 PM 2026-11-06 (15+ min late) Probability (calibrated, mean per flight) 78% over 88 flights Forecast cycle · horizon Fri Nov 6, 4:00 PM EST · 1.1 h ahead Model version famB-sd-v4i Source event time Ingested Published here (2026-11-06 21:56 UTC) Sources Tailodds forecast Provenance forecast_blocks forecast-blocks-v1 · config 2026-10-17.1 · writer template-v1 Was this useful?
Closed Fri Nov 6, 6:00 PM EST: The forecast window has started; measured delays are reported separately as observed items.