How we measure our pass rate
Most DMV apps advertise a pass rate somewhere between 95% and 97%. Almost none of them say how it was measured, over what period, or how many people it is based on. This page describes our method before we publish a figure, so the number can be judged rather than just believed.
Where the number will come from
After a scheduled written-test date passes, the app asks a single question: did you pass? Answering is optional, and the answer is counted anonymously — a pass or fail is added to a running total for your track and nothing else is transmitted. There is no account, no device identifier, and no way to connect two answers to the same person. You can turn the reporting off entirely in Settings.
What we will publish
- The pass rate itself, as a whole percentage
- The number of responses it is based on
- The date range those responses cover
- The breakdown by track, where each track has enough responses to support one
The threshold before we publish anything
We will not quote a pass rate until at least 385 responses exist for the claim being made — the sample size needed for roughly a five-point margin of error at 95% confidence. Below that we will say we do not have enough data yet. We will not publish a provisional figure, a rounded-up one, or an “early data suggests” version, because a number like that gets screenshotted and outlives its caveat.
What this figure will not be
It will be self-reported and opt-in, which means it carries a selection bias we cannot remove: someone who passed may well be likelier to come back and say so. Treat it as a floor on how our users did, not as a controlled study, and not as a prediction of your own result. We would rather publish a smaller honest number with its weaknesses attached than a larger one without them.
It is also not a guarantee. No app can promise you a pass, and any app that implies otherwise is telling you something it cannot know.
Corrections
The figure will be recomputed and re-dated on a fixed schedule. If it goes down, it goes down — a statistic that only ever moves upward is how the numbers elsewhere in this category lost their meaning. Our sourcing and correction policy is described in the content methodology.