The math, the variables, and an honest accounting of why one formula can't fit every marathoner. Built by a runner with 40+ years on the roads, open sourced for anyone to inspect, use, or improve.
Most race predictors give you one number and a lot of confidence. Runners know better. Two people with the same 5K time can finish a marathon half an hour apart, and the difference usually isn't talent. It's training, pacing, and how you handle the start. We wanted a calculator that admits this instead of hiding it, and that helps you plan the race you're actually going to run.
Pace and time are simple arithmetic: time equals distance times pace. The split table is that math applied mile by mile (or kilometer by kilometer), including mid-mile pace changes when you build a segment strategy.
Race equivalency uses the Riegel formula, published by Peter Riegel in 1977 and still the backbone of most race predictors:
Take a time you've run (T₁) at one distance (D₁), and it predicts your time (T₂) at another distance (D₂). The exponent is where all the interesting disagreement lives.
| Variable | What it is | Where it comes from |
|---|---|---|
| Distance | Marathon first (26.2 mi / 42.195 km), with half, 10K, 5K, and custom options. | You pick it. |
| Goal time or pace | Enter either; the other is computed. | You enter it. |
| Segments | Up to 5 blocks of the race at different paces. The last segment always auto-fills the remaining distance. | Your race strategy. |
| Riegel exponent | How much you slow down as distance grows. 1.06 is the classic textbook value. | The runner profile slider, 1.04 to 1.12. |
| Units | Miles or kilometers, converted throughout. | One toggle. |
Most calculators use 1.06 and stop there. That value assumes you've done the training the longer distance demands: the weekly mileage and, above all, the long runs. Research on recreational marathoners tells a different story. Large-scale studies have found that runners with low mileage or few long runs behave like an exponent of 1.10 to 1.15 when predicting a marathon from a shorter race. They don't lack speed. They fade.
The slider is really asking one honest question: did you do the training, or are you going to gut it out on shorter-race fitness? Both are legitimate ways to run a marathon. Only one of them should trust the textbook number.
The marathon punishes a fast start more than any other race. Going out ahead of goal pace to "bank time" is the most common way to turn the last 10K into a death march: the energy cost of early speed comes due with interest. That's why the calculator is built around split strategy, not just a finish time. Even splits are the gold standard. A controlled start with a planned finish surge is the realistic alternative. Banking time is not a plan.
No two marathoners run the same race. Here are the profiles we designed for. Find yours, and the slider setting that matches.
Fast 5K and 10K times, comfortable at the half, but training rarely goes past 13.1. Race day is an exercise in holding on. Capable of a great marathon on the right day, but the fade is real and the textbook prediction is a trap. This is our founder: 40+ years of running, quick short-race times, and marathons run on grit more than long-run mileage.
High weekly mileage, long runs on the calendar all year. Their 5K might be nothing special, but their marathon pace barely drifts from their half pace. They pass the specialists at mile 22, looking annoyingly comfortable.
Follows a standard plan: a weekly long run, a bit of speed work, sensible mileage. The textbook exponent was basically built for this runner, and for them it works well.
The goal is to finish strong, not to chase a number. Best served by a plan that starts slower than feels necessary, with the option of a faster final segment if the day goes well. The first marathon teaches you which type you are.
Fitness from a previous life, current training that hasn't caught up yet. Old race times overpredict badly. Enter a recent race, not a memory, and be honest with the slider.
Think we missed a type, or disagree with a setting? Good. That's what open source is for.
The whole calculator is plain HTML, CSS, and JavaScript in a single file. No frameworks, no build step, no tracking, no ads, no signup. It's MIT licensed: use it, copy it, embed it on your site, improve it. If you find a bug or think the math should work differently, open an issue or a pull request.