Race time predictor
One recent result — equivalent performances at every distance, from the Daniels/Gilbert curves the RunPlan engine uses.
Equivalent-performance predictions (Daniels/Gilbert). They assume you train for the distance — a 5K runner doesn’t owe anyone a marathon.
A prediction says what your engine is worth if you train for the distance — the free training plans are how you collect, and the training pace calculator sets day-to-day paces from the same result.
Common questions
How accurate are race time predictors?
For nearby distances — a 5K predicting a 10K — very. The further apart the distances, the more the prediction assumes: a marathon prediction from a 5K is only honest if you actually do marathon training, with the long runs and volume that implies. The math gives equivalent fitness; the training makes it real.
Why is my marathon slower than the predictor says?
Almost always endurance, not speed: the prediction assumes distance-appropriate preparation. If your longest run was 25 kilometers, the last 10 of a marathon will not follow the curve. Fuelling and pacing errors are the other classic reasons.
What result should I enter?
Your most recent all-out effort — a race or an honest time trial, ideally within the last couple of months. Fitness moves; a result from last year describes last year.
The math is open source: training_plan_kit on GitHub.
