Race prediction is the sport of speculating about future pain. Runners obsess over finish times months out — 'if I can run a 22:00 5K in April, what marathon time can I target in October?' Peter Riegel's formula from 1977 remains the gold standard: T2 = T1 × (D2/D1)^1.06, where T is time and D is distance. It's simple, empirical, and startlingly accurate under the right conditions.

The right conditions matter. Riegel's exponent of 1.06 was derived from world-class runners at similar training states across their competitive distances. Applied to trained runners predicting from 5K to half marathon, accuracy is ±2%. Applied to a 5K runner predicting a marathon without marathon-specific training (20+ mile long runs, race-pace tempos, fueling practice), the prediction is 5–15% optimistic. The formula doesn't know if you've run long.

Common benchmarks worth memorizing: sub-4 marathon = 9:09/mile (5:41/km) average pace. Boston qualifier for 35-year-old man = sub-3:00 marathon = 6:52/mile (4:16/km). Sub-2 half marathon = 9:09/mile (5:41/km). Sub-25 5K = 8:03/mile (5:00/km). These times are within reach for most healthy adults with 3–6 months of structured training and a weekly long run appropriate for the distance.

The 'wall' at mile 18–22 in a marathon is glycogen depletion, not fitness failure. Muscle glycogen stores are limited to about 400–500 g (1600–2000 kcal), enough for 90–120 minutes of hard running. Beyond that, the body relies increasingly on fat metabolism, which is slower. Race-day fueling — 30–60 g of carbs per hour via gels, chews or drink — is what lets trained runners hold marathon pace to the finish. Skip fueling and even a well-trained runner blows up at 20.

Prediction confidence increases with training volume similar to race demands. A runner logging 25 miles/week predicting a marathon is guessing. A runner logging 45–60 miles/week with three long runs of 18+ miles has a prediction confidence of ±3 minutes. The calculator's training-adjusted mode weights predictions by weekly mileage, longest recent long run, and race-specific workouts completed. Use it for planning, respect its assumptions, and always target a pace 5 seconds/mile slower than predicted for the first 3 miles — negative splits win races.