Track record
How often our go / no-go calls hold
Every evening we record tomorrow's verdict for the same sample of cities and activities. After the day, we apply the same rules to the weather that was actually recorded and publish whether the call held. Nothing is left out.
Calls that held, last 30 days
93%
665 of 714 day-ahead go / no-go calls
Last 90 days
93%
665 of 714 calls · 12 places, 6 activities
When we said “go”
94%
of 582 go calls, the day was a go
When we said “no-go”
88%
of 132 no-go calls, the day was a no-go
Average difference between the forecast score and the score for the recorded weather: 4.5 points (out of 100). Updated 2 Nov 2026 10:56 UTC.
By activity (90 days)
| Activity | Calls | Held |
|---|---|---|
| Crane lifts | 119 | 96% |
| Cycling | 119 | 97% |
| Golf | 119 | 89% |
| Outdoor event (BBQ, party, wedding) | 119 | 92% |
| Pouring concrete | 119 | 87% |
| Running | 119 | 99% |
By place (90 days)
| Place | Calls | Held |
|---|---|---|
| Belfast, United Kingdom | 60 | 92% |
| Berlin, Germany | 60 | 85% |
| Cardiff, United Kingdom | 60 | 92% |
| Chicago, United States of America | 60 | 97% |
| Dublin, Ireland | 60 | 92% |
| Edinburgh, United Kingdom | 60 | 93% |
| London, United Kingdom | 60 | 93% |
| Madrid, Spain | 60 | 100% |
| Manchester, United Kingdom | 60 | 85% |
| New York, United States of America | 60 | 100% |
| Paris, France | 60 | 95% |
| Sydney, Australia | 54 | 94% |
Latest checks
| Day | Place | Activity | Our call the evening before | Recorded weather | |
|---|---|---|---|---|---|
| 1 Nov | Belfast | Pouring concrete | go great, 97 | go great, 98 | Held |
| 1 Nov | Belfast | Crane lifts | go great, 100 | go great, 100 | Held |
| 1 Nov | Belfast | Cycling | go great, 97 | go great, 99 | Held |
| 1 Nov | Belfast | Golf | go great, 90 | go great, 92 | Held |
| 1 Nov | Belfast | Outdoor event (BBQ, party, wedding) | go good, 74 | go good, 77 | Held |
| 1 Nov | Belfast | Running | go great, 100 | go great, 100 | Held |
| 1 Nov | Berlin | Pouring concrete | go great, 98 | go great, 100 | Held |
| 1 Nov | Berlin | Crane lifts | go great, 100 | go great, 100 | Held |
| 1 Nov | Berlin | Cycling | go great, 98 | go great, 100 | Held |
| 1 Nov | Berlin | Golf | go great, 91 | go great, 96 | Held |
| 1 Nov | Berlin | Outdoor event (BBQ, party, wedding) | go good, 75 | go good, 82 | Held |
| 1 Nov | Berlin | Running | go great, 100 | go great, 100 | Held |
| 1 Nov | Cardiff | Pouring concrete | go great, 99 | go great, 100 | Held |
| 1 Nov | Cardiff | Crane lifts | go great, 100 | go great, 100 | Held |
| 1 Nov | Cardiff | Cycling | go great, 100 | go great, 100 | Held |
| 1 Nov | Cardiff | Golf | go great, 95 | go great, 98 | Held |
| 1 Nov | Cardiff | Outdoor event (BBQ, party, wedding) | go good, 80 | go good, 84 | Held |
| 1 Nov | Cardiff | Running | go great, 100 | go great, 100 | Held |
| 1 Nov | Chicago | Pouring concrete | go great, 100 | no-go great, 92 | Missed |
| 1 Nov | Chicago | Crane lifts | go great, 100 | go great, 100 | Held |
| 1 Nov | Chicago | Cycling | go great, 100 | go great, 100 | Held |
| 1 Nov | Chicago | Golf | go great, 100 | go great, 100 | Held |
| 1 Nov | Chicago | Outdoor event (BBQ, party, wedding) | go great, 95 | go great, 99 | Held |
| 1 Nov | Chicago | Running | go great, 100 | go great, 100 | Held |
| 1 Nov | Dublin | Pouring concrete | go great, 98 | go great, 99 | Held |
| 1 Nov | Dublin | Crane lifts | go great, 100 | go great, 100 | Held |
| 1 Nov | Dublin | Cycling | go great, 98 | go great, 100 | Held |
| 1 Nov | Dublin | Golf | go great, 91 | go great, 93 | Held |
| 1 Nov | Dublin | Outdoor event (BBQ, party, wedding) | go good, 75 | go good, 78 | Held |
| 1 Nov | Dublin | Running | go great, 100 | go great, 100 | Held |
How we measure it
- Fixed sample, chosen in advance. 12 cities with different climates (Belfast, Berlin, Cardiff, Chicago, Dublin, Edinburgh, London, Madrid, Manchester, New York, Paris, Sydney) and 6 activities with different sensitivities (Crane lifts, Cycling, Golf, Outdoor event (BBQ, party, wedding), Pouring concrete, Running). Every day is recorded; none are removed afterwards.
- The call. From 18:00 local time we record tomorrow's verdict from the forecast, exactly as the API would have given it then, with the activity's usual hours.
- The check. After the day is over (from 04:00 the next morning) we take the hour-by-hour weather that was recorded for that day from the same weather service and apply exactly the same rules. A day is a “go” when there is at least one window of the activity's minimum length in which every hour scores 60 or more.
- Like for like. Air quality, pollen and weather warnings are left out of both sides, because the recorded data doesn't include them. Days without complete recorded weather are skipped, not counted as held.
- What it shows. How well a decision made the evening before held up against what happened. Recorded weather is itself an estimate for the place (from nearby stations and models), not a measurement at your exact site, and results for your own place and activity may differ. Treat it as a guide, not a guarantee.