Accuracy Report
Our forecasts are off by 100% on average
We compare every posted forecast against what actually happened. 1 day of data, 1 matched pairs.
Median absolute error
100%
Lower is better
Within 20% of actual
0%
Higher is better
Matched pairs
1
across 1 days
Updated weekly. MAPE = mean absolute percentage error vs. median posted wait. "Within 20%" = fraction of forecasts within ±20% of the realized wait. Data covers operating-status attractions only; closed/down rows excluded.
How we measure this
Every forecast WaitScout issues is timestamped. Once the day closes, we compare the posted median wait (p50) against the realized median from our crowdsourced actuals corpus. We compute MAPE — mean absolute percentage error — which penalizes proportional misses, not raw-minute misses. A 5-minute error on a 10-minute wait hurts more than a 5-minute error on a 60-minute wait.
We do not cherry-pick. Every forecast that has a matching actual is included. If a ride was closed, we skip it — forecasting closure is a separate signal. We publish this page because hiding accuracy numbers is a trust signal in the wrong direction.