Riegel Race Predictor

Peter Riegel's race prediction formula

Peter Riegel, an American engineer and runner, published his race prediction formula in a 1981 issue of Runner's World magazine. The formula is elegant in its simplicity: T2 = T1 × (D2/D1)^e, where T1 is the time at distance D1, T2 is the predicted time at distance D2, and e is a sport-specific exponent, 1.06 for running, 1.08 for cycling, 1.03 for swimming, and approximately 1.05 for rowing. The exponent captures a fundamental property of endurance performance: as distance increases, pace decreases predictably. The formula works because human fatigue over distance follows a power-law relationship, and Riegel's exponent quantifies that relationship for each sport.

Why an exponent greater than 1.0?

If the exponent were exactly 1.0, pace would be perfectly constant across distances, a runner who completes 5K in 25 minutes (5:00/km) would run a marathon at the same 5:00/km pace, finishing in 3:31. In reality, the marathon takes longer (3:31 is elite pace for women); a typical 25-minute 5K runner might run a marathon in 4:00-4:30, which is a pace of 5:40-6:25/km. The exponent 1.06 captures this slowdown: for a 5K-to-marathon prediction, (42.195/5)^1.06 = 8.44^1.06 ≈ 9.4, so the 25-minute 5K runner predicts a marathon of 25 × 9.4 = 235 minutes ≈ 3:55. That's a pace of 5:35/km, slower than 5K pace, as expected. The exponent quantifies the rate of slowdown per unit of additional distance.

Sport-specific exponents

The exponent varies by sport because the physiology of fatigue differs. Running (1.06) has a moderate exponent because running involves both cardiovascular and musculoskeletal fatigue, impact forces accumulate, glycogen depletes, and muscle damage builds over long distances. Cycling (1.08) has a higher exponent because cycling is more glycogen-dependent (cyclists ride at higher fractions of VO2 max for any given duration) and fatigue is more cardiovascular than musculoskeletal. Swimming (1.03) has a low exponent because swimming is non-impact and technique-driven; the slowdown with distance is modest. Rowing (~1.05) sits between running and cycling.

How Riegel derived the exponent

Riegel analyzed world record performances across multiple distances in each sport and plotted log(time) vs log(distance). The slope of the resulting line is the exponent. For running, the slope is approximately 1.06 across distances from 800m to the marathon. For cycling, the slope is approximately 1.08 across distances from 1 km time trials to multi-day stage races. Riegel's analysis showed that the exponent is remarkably consistent across ability levels, world records, age-group records, and recreational performances all follow the same power-law relationship, suggesting that the exponent captures a fundamental property of human endurance physiology.

Accuracy of Riegel predictions

Rigel's formula is most accurate when predicting performances at distances within 2-4× of the reference distance. Predicting a 10K from a 5K (ratio 2.0) is typically within 1-2% of actual time. Predicting a marathon from a 5K (ratio 8.4) is less accurate, typically within 5-10%, because marathon performance depends on long-run endurance, fueling, and pacing that 5K racing does not test. Predicting a 5K from a marathon (predicting shorter from longer) is generally more accurate than the reverse, because a marathon performance implicitly includes the endurance that 5K racing does not test.

Comparing Rigel to VDOT

Rigel's formula and Jack Daniels' VDOT are the two most widely used race prediction methods. They differ in their approach: Rigel uses a single power-law exponent for all runners, while VDOT uses a per-runner VO2 max equivalent derived from one race result and then predicts other distances using a more complex formula that accounts for the runner's individual fatigue profile. In practice, the two methods give similar predictions for distances within 2-3× of the reference race, but diverge for more extreme predictions. VDOT is generally more accurate for individual runners (because it's calibrated to that runner's actual performance), while Rigel is simpler and works well across populations.

Common use cases

Rigel's formula is widely used by recreational and competitive athletes to set race goals, plan training paces, and choose realistic target times. Common conversions include 5K → 10K, 10K → half marathon, half → full marathon, 10-mile cycling time trial → 40-mile time trial, and 1500m swim → 5K open water swim. The formula is also used by race directors to estimate cutoff times and by coaches to identify athletes whose performance at one distance is disproportionately strong or weak relative to other distances (a sign of either specific talent or training imbalance).

Limitations of Riegel

Rigel has several limitations. First, the exponent is a population average, individual runners may have slightly higher or lower exponents based on their physiology. A runner with exceptional endurance but modest speed may have an exponent closer to 1.04, while a sprinter-type runner may have an exponent closer to 1.08. Second, the formula assumes the athlete is equivalently trained for the target distance, a 5K specialist predicting a marathon will be optimistic because they lack the long-run training. Third, the formula does not account for course profile, weather, or surface, hills, heat, and trail all shift actual times substantially. Fourth, the formula breaks down at extreme distances (ultramarathons of 100+ miles) where fatigue, fueling, and sleep deprivation dominate over aerobic capacity.

How to use this calculator

Enter your reference race distance and time, then enter your target distance. The calculator will display the predicted time at the target distance, along with the predicted pace and the slowdown percentage relative to your reference pace. Use the predicted time as a realistic target for upcoming races, but adjust based on your training specificity, if you've been doing marathon-specific long runs, your marathon may be faster than the prediction from a 5K; if you've been doing 5K speedwork, your 5K may be faster than the prediction from a half marathon. Use Rigel alongside VDOT for cross-validation when setting race goals.

Riegel for predicting cycling times

Riegel's formula works for cycling with an exponent of 1.08 (vs 1.06 for running). The higher cycling exponent reflects the greater glycogen demand of cycling, cyclists typically ride at higher fractions of VO2 max for any given duration, so fatigue accumulates faster as distance increases. Common cycling predictions include 10-mile time trial → 25-mile time trial → 50-mile time trial → 100-mile time trial → 12-hour time trial. For a cyclist who completes a 10-mile TT in 25 minutes (24 mph average), Riegel predicts a 25-mile TT in 25 × (25/10)^1.08 = 25 × 2.69 = 67 minutes (22.4 mph average), a 1.6 mph pace drop, which matches real-world cycling performance data.

Cycling predictions are most accurate when the source and target distances use similar terrain and conditions. A 10-mile flat TT predicting a 25-mile flat TT will be accurate; a 10-mile flat TT predicting a 25-mile hilly TT will be optimistic because hills slow longer-distance cycling disproportionately (more climbing total, more fatigue per mile). For ultra-distance cycling (200+ miles), Riegel becomes less accurate as sleep deprivation, fueling, and saddle discomfort become the limiting factors rather than aerobic capacity. For these events, use Riegel as a very rough starting point and expect actual times to be 15-40% slower than predicted.

Combining Riegel and VDOT for race planning

Riegel and VDOT are complementary tools for race planning. Riegel is simpler, one exponent, no iterative calculation, and works across sports. VDOT is more accurate for individual runners because it's calibrated to your specific fatigue profile, but it's running-specific and requires more complex computation. For race planning, use both methods and compare: if they agree within 2-3%, you can be confident in the prediction; if they disagree by more than 5%, investigate why (perhaps your training is biased toward short or long distances, making one prediction more reliable than the other).

A practical race-planning workflow: Step 1: Enter your most recent race result in both calculators. Step 2: For your target race distance, compare the predicted times from Riegel and VDOT. Step 3: Choose the more conservative (slower) prediction as your A-goal pace, and the more aggressive (faster) prediction as your stretch goal. Step 4: During training, do workouts at the predicted pace to validate it, for example, if VDOT predicts a 10K in 45:00, run 5 × 1K at 4:30/km pace with 2-minute recovery. If the workout feels manageable, the prediction is realistic; if you're struggling, adjust the goal. Step 5: Re-run both calculators 2-4 weeks before race day with your most recent race result to confirm the prediction.

Limitations of Riegel for trail and ultramarathon

Riegel's formula was derived from road and track world records, where terrain is uniform and conditions are controlled. Trail races and ultramarathons deviate substantially from these assumptions, making Riegel predictions unreliable. Three factors cause the deviation. First, terrain, technical trail running is 20-50% slower than road running at the same effort, due to footing challenges, elevation changes, and obstacles. Second, fatigue accumulation, at distances beyond 50K, factors like foot swelling, chafing, gastrointestinal distress, and mental fatigue slow performance more than aerobic capacity predicts. Third, fueling and hydration, at ultra distances, the ability to consume and absorb calories becomes the limiting factor, not aerobic capacity. For trail and ultra races, use Riegel as a very rough starting point and expect actual times to be 15-50% slower than predicted, depending on terrain difficulty, elevation gain, weather, and your ultra-specific training. For more accurate ultra predictions, use race-specific time estimators that account for elevation gain (e.g. +1 minute per 100 feet of climbing) and terrain (multiply road pace by 1.2-2.0 depending on technical difficulty).