Marathon Time Predictor (from 5K, 10K, or Half)

Marathon Prediction Guide

What should I know about The Riegel Formula?

Peter Riegel's 1977 formula: T2 = T1 × (D2/D1)^1.06. Where T = time, D = distance, exponent 1.06 represents the typical race performance fatigue factor. Used by virtually all major running calculators. Accuracy: very accurate from 5K to half-marathon predictions. Less accurate for marathon prediction — the 1.06 exponent works for shorter distances but marathon has additional 'wall' factor most runners experience. Standard Riegel marathon predictions are systematically optimistic by 10-30 minutes for most runners, which is why experienced coaches often apply a correction factor specifically for the marathon distance.

What should I know about Typical Predictions?

From a 22-minute 5K (4:30/mile): Riegel marathon = 3:31. Realistic for high-mileage runner: 3:35-3:45. Realistic for moderate runner: 3:45-4:00. Realistic for low-mileage: 4:00+. From a 45-minute 10K: Riegel marathon = 3:30. Note: 10K time gives more accurate marathon prediction than 5K. The aerobic component is more similar. From 1:35 half-marathon: Riegel marathon = 3:20. Closer prediction — half marathon physiology much more similar to marathon. Half-marathon time × 2 + 10-15 minutes gives a more realistic marathon estimate for most runners than the pure Riegel formula.

What should I know about Adjustments for Training Volume?

Mileage matters MORE than speed for marathon prediction. Two runners with same 10K time: 60-mile/week runner: closer to Riegel marathon prediction. 25-mile/week runner: 15-30 minutes slower marathon. The aerobic engine needed for marathon is built by total time on feet, not interval workouts. Long run distance: peak long run within 6 weeks of marathon should reach 18-22 miles. Without this: very high risk of crashing in last 6 miles. Marathon training plans: minimum 16-week build, 4-5 runs per week, is the widely recommended baseline for a first-time marathoner aiming to finish comfortably rather than simply survive.

What's the key thing to understand about Race-Day Pacing?

Negative split strategy: run the second half slightly faster than the first. Conservative first 6-10 miles at goal pace + 10-15 seconds. Settle into goal pace miles 10-20. Accelerate slightly final 10K if energy allows. Positive split (first half faster): nearly always ends in crashing. Almost every sub-3-hour marathon is negative split. Pace bands: many marathons sell pace bands showing splits for target time. Useful for staying disciplined. Pacing groups: official pacers at major marathons, such as London, run at a steady, practised pace and carry a sign showing their target finish time — a reliable way to hit a specific goal without constantly checking your own watch.

Why Does Riegel's Formula Become Less Accurate for Very Different Distances?

Riegel's formula (T2 = T1 × (D2/D1)^1.06) works well for predicting times between distances that stress similar energy systems — for example, predicting a marathon from a half-marathon — but its accuracy declines the further apart the two distances are, because longer races increasingly depend on factors a shorter race barely touches: glycogen depletion, fuelling strategy, heat management, and musculoskeletal fatigue from sustained pounding. Predicting a marathon time from a 5K result, for instance, tends to produce an overly optimistic marathon prediction for most recreational runners, because the 5K says little about a runner's ability to fuel and pace over 3+ hours, whereas a recent half-marathon or 30K time trial gives a much more reliable marathon prediction since it already involves meaningful glycogen use and sustained effort similar to what the marathon demands.

How Much Does Marathon-Specific Training Change the Prediction?

Riegel-style predictors assume the runner has trained specifically and adequately for the target distance, which is a significant assumption for the marathon in particular — a runner who has only run half-marathon-length long runs in training will typically run slower relative to their predicted time than a runner who has completed several 30-32km long runs, because the marathon's back quarter is disproportionately affected by glycogen depletion and fatigue resistance that shorter long runs do not fully train. This is why coaches generally treat a pure formula-based prediction as an upper-bound best case rather than a guaranteed outcome, and recommend runners with limited long-run experience target a time 5-10 minutes slower than the raw prediction, while well-prepared runners with a strong long-run base and practiced fuelling strategy can reasonably expect to run close to or even slightly inside the formula's prediction.

How Does Riegel Compare to Other Prediction Models?

Riegel's formula isn't the only way to predict a marathon time from a shorter race, and comparing it with alternatives shows where each tends to over- or under-predict. VDOT, developed by Jack Daniels from a large dataset of trained runners, uses a race-specific physiological model rather than a single exponent, and tends to be more accurate across a wider range of fitness levels because it accounts for the different ways aerobic and anaerobic systems contribute at different distances. The Cameron formula adds separate terms for distance and a fixed component, and in some studies predicts marathon times slightly better than Riegel for runners who haven't trained specifically for the marathon distance, since Riegel's fixed exponent assumes endurance scales the same way for everyone. In practice, all of these formulas agree closely for predictions across short-to-middle distances (a 5K to a half marathon) but diverge more as the target distance gets further from the race used to predict it — which is exactly the marathon-from-5K scenario where Riegel is known to be most optimistic. For a runner with limited marathon-specific training, it's sensible to treat Riegel's number as a ceiling rather than a target, and to weight it down slightly using VDOT or Cameron estimates as a sanity check before setting a goal pace.

Worked Example: Predicting a Marathon Time From a 10K

Riegel's formula is T2 = T1 × (D2/D1)^1.06. Starting from a 10K time of 45:00 (2,700 seconds) and predicting a full marathon (42.195 km vs 10 km, a distance ratio of 4.2195): (4.2195)^1.06 = e^(1.06 × ln 4.2195) = e^(1.06 × 1.4396) = e^1.5260 ≈ 4.601. Multiply through: T2 = 2,700 × 4.601 ≈ 12,423 seconds ≈ 3:27:03 — a reasonable marathon prediction for that 10K time, assuming marathon-adequate endurance training.

How the Riegel Exponent Changes the Prediction

The exponent 1.06 is an average across runners; well-trained endurance athletes fade less over distance (a lower exponent), while speed-biased runners with less distance training fade more (a higher exponent). Using the same 45:00 10K time:

The roughly 24-minute spread between exponents shows why Riegel predictions should be treated as a starting estimate rather than a guarantee — your actual long-run mileage and fade pattern in training say more about which end of that range you sit on than the formula alone can.

Riegel's Formula Was Originally Validated on Elite Runners

Peter Riegel's 1977 paper derived the 1.06 exponent from world-record and elite performance data across distances, which is one reason recreational runners sometimes find the prediction slightly optimistic for the marathon specifically — elite athletes maintain a more consistent percentage of their aerobic capacity across distances than less-trained runners, who tend to fade more over the final 10km of a marathon due to glycogen depletion and under-trained long-run endurance, regardless of their 10K speed.

Getting a More Reliable Prediction From Multiple Race Results

A single recent race result, fed into a prediction formula such as Riegel's, gives a reasonable estimate, but it reflects the conditions, pacing, and form of that one specific day. Coaches commonly get a more reliable marathon prediction by averaging estimates from two or three recent races at different distances, for example a recent 10K and a recent half marathon, rather than relying on just one. This helps smooth out the effect of an unusually good or unusually poor single performance, a race run in unhelpful weather, or a result from a distance that does not stress the same systems a marathon will. It does not replace marathon-specific training in the final predicted number, since even an accurate multi-race average is still estimating a performance the athlete has not yet actually completed, but it generally narrows the range of likely error compared with extrapolating from a single shorter race, which the companion marathon pace and split calculator can then help translate into a mile-by-mile or kilometre-by-kilometre pacing plan.