Race Time Predictor
What a Race Time Predictor Actually Does
A race time predictor takes a performance you’ve already run and projects what you’d finish at a different distance. Enter a 5K of 22:30, and it estimates what a half marathon or 10K might look like. The number is only as good as the formula behind it, and there are two meaningfully different approaches worth understanding before you trust either output.
The Two Formulas: Riegel vs VDOT
The Riegel formula has been in use since 1981:
T2 = T1 × (D2 / D1)^1.06
T1 is your known time, D1 is the known distance, T2 is the predicted time, D2 is the target distance. The exponent 1.06 captures the physiological reality that pace slows as distance increases. It’s simple, mathematically clean, and still embedded in most online tools, including the Runner’s World race time predictor. The marathon-specific failure modes of Riegel are the main reason it’s slowly being replaced in serious training tools.
The VDOT method, developed by Dr. Jack Daniels, works through an intermediate fitness score. Your race result implies a VDOT value, a proxy for your current aerobic capacity. That score then maps to predicted performances at other distances using oxygen cost data Daniels assembled from thousands of runners over several decades.
The practical difference shows up most clearly when predicting across a large distance gap. A runner using a 5K to estimate a marathon will get significantly different answers from Riegel versus VDOT. Riegel tends to produce the more optimistic number. VDOT tends to be more conservative and, in practice, closer to what actually happens on race day. That’s the main reason the Race Time Predictor on Pacesmith uses VDOT rather than Riegel.
What Makes a Good Input Race
The quality of any prediction depends almost entirely on the quality of the input.
Recency. Your fitness changes from month to month. A result from the past six to eight weeks is the most useful input. A PR from last fall reflects who you were then, not where you are now.
Full effort. Predictors assume you actually raced the input distance hard. A controlled tempo or a paced effort where you finished with plenty left will produce optimistic numbers across every target distance you plug in.
Proximity to the target. A 10K-to-half-marathon prediction is reliable. A mile-to-marathon prediction crosses several distinct physiological thresholds, and uncertainty compounds with each one. The further apart the two distances are, the more conservatism you should build into your race plan.
Course and conditions. A certified flat road 5K and a muddy cross-country 5K with significant climbing are not equivalent inputs. Predictors don’t account for terrain, heat, humidity, or wind. If your input race had difficult conditions, the output will be more optimistic than reality warrants.
How Accurate Are Race Time Predictions?
Accurate enough to be useful. Not accurate enough to treat as a guarantee.
The most common complaint runners have about predictors is that they’re too optimistic for longer distances. This is usually a training-specificity problem rather than a formula problem. A solid 10K finish reflects aerobic capacity; it doesn’t reflect the muscular durability needed to run 26.2 miles efficiently. A runner with strong 10K fitness but minimal long-run mileage will almost always underperform a VDOT-based marathon prediction, because the formula assumes your training was appropriate for the target distance. Often it isn’t.
The Garmin race predictor, which uses VO2 max estimated from heart rate and GPS pace data, carries a similar optimism bias. Easy pacing can be fast for all sorts of reasons: cool weather, a favorable route, fresh legs early in a training cycle. Any of those push the VO2 max estimate up, and the projected race times follow. Garmin’s predictions are useful as a rough directional signal, but they deserve the same skepticism you’d apply to any VO2 max-based race time predictor.
The most honest framing: a race predictor tells you what a runner with your aerobic fitness could run at the target distance under good conditions with appropriate training behind them. Whether all three of those things apply to your situation is a question the formula cannot answer.
Race Time Predictor for the Half Marathon
The half marathon is where predictions tend to perform best. The physiological overlap with 5K and 10K fitness is high enough that the formula has solid ground to work from, and the distance gap isn’t large enough to compound errors significantly.
A recent 10K is the best input for a half marathon prediction. A 50:00 10K (8:03/mile) maps to approximately 1:51:00 to 1:52:00 under the VDOT formula, around 8:29 to 8:33 per mile. That’s a meaningful target if your long runs have been appropriate. If your only reference is a 5K and your weekly mileage has been modest, add a few minutes of buffer.
Once you have the predicted time, the practical next step is building a split plan from it. Knowing your target is useful; knowing your per-mile splits is actionable. A race split planner turns your predicted finish into a mile-by-mile or kilometer-by-kilometer plan, with options for even splits or a negative split structure you can print and carry to the start line.
Predictions for Shorter Distances: 800m and Mile
Below 5K, the dynamics shift. Aerobic capacity still matters, but lactate threshold and raw speed become larger factors. A runner with strong endurance training but limited track work will often underperform VDOT predictions at the 800m and mile, because those distances require fast-twitch adaptations that long easy miles don’t develop.
For an 800m prediction specifically: treat the VDOT output as an optimistic ceiling, not a target. The more reliable check is whether your interval workout paces in training are consistent with the predicted race pace. If 400m repeats at that speed feel genuinely hard but manageable, the prediction is probably honest. If they feel unrunnable, your fitness hasn’t caught up to the number yet.
The VDOT Calculator shows your training zone paces alongside race predictions. Your Interval and Repetition zone paces correspond most directly to 800m-through-5K racing fitness, so those are the zones worth comparing against your actual workout data. The training pace calculator guide walks through what each of those zones is for in more detail.
What About Bike Race Time Predictors?
The Riegel and VDOT formulas are built entirely around running physiology: the relationship between aerobic capacity and pace across different running distances. Bike race time prediction is a separate domain, working from power output, aerodynamics, and course gradient. The variables don’t translate. A running predictor applied to a cycling event produces a meaningless number.
Turning the Prediction into a Race Plan
A predicted finish time is the start of race planning, not the end.
Work backward from the predicted time to a target pace per mile. Then think about how to distribute effort. Most runners go out too fast in the first mile because goal pace feels comfortable when you’re fresh, which is exactly how it should feel. The problem is interpreting that comfort as a signal to push harder. A moderate negative split, running the second half marginally faster than the first, consistently produces better outcomes than even splits, and far better outcomes than a fast start that unravels in the final miles.
The prediction also serves a training calibration function that often goes unnoticed. VDOT connects your race result directly to training zone paces: Easy, Marathon, Threshold, Interval, and Repetition. If your actual training paces don’t match what your race fitness implies, you’re either leaving adaptation on the table or accumulating fatigue you don’t need. The prediction reflects where you actually are, not where you assume you are. That’s one of the more honest things a number can do.
If you want to work through all of this in one place, Pacesmith puts the race predictor, VDOT zones, and split planner together for a one-time $1.99, no subscription, no account, and it works offline.