All Articles

Most runners think of their race times as isolated data points — a 5K time, a half marathon time, a marathon time. Each one measured separately. Each one useful only for that distance.

Jack Daniels saw them differently. A race time at any distance, he argued, encodes a precise measurement of your aerobic fitness. And once you have that measurement, you can project it to any other distance with high accuracy. The system he developed to do this — VDOT — is the most widely used race prediction and training prescription tool in endurance coaching.

Here's how it works.

What VDOT Is — and Isn't

VDOT is not the same thing as VO2max, though it's closely related. VO2max measures the maximum volume of oxygen your body can consume per minute per kilogram of body weight — your raw aerobic ceiling. VDOT measures something Daniels called effective VO2max: the aerobic capacity you actually express in racing conditions, accounting for both raw capacity and running economy.

Running economy is the efficiency with which you use that oxygen. Two runners can have identical VO2max values but very different race times if one of them converts oxygen into forward motion more efficiently than the other. VDOT captures both variables in a single number derived directly from race performance.

"VDOT is a measure of running ability that takes into account both aerobic capacity (VO2max) and the economy with which a runner uses that capacity. A runner with excellent economy can achieve a higher VDOT — and thus faster race predictions — with a lower actual VO2max."

— Daniels, J. (2014). Daniels' Running Formula, 3rd edition. Human Kinetics.

The practical implication: VDOT derived from race performance is a more accurate predictor of your actual racing ability than a laboratory VO2max test would be. It measures what you do, not what your physiology theoretically allows. A lab test might show a higher ceiling; VDOT shows what you can actually run to today.

The Equivalence Table: What Your Race Time Predicts

Given a VDOT value, Daniels' tables project equivalent performances at any standard distance. Runners at the same VDOT are considered aerobically equivalent — they should be able to produce the same times, distance for distance, on equivalent courses under similar conditions.

The full VDOT tables are published in Daniels' Running Formula. Here's how equivalence works in practice, using three worked examples:

Example 1 — Beginner runner: You ran a 30:00 5K. Using the Daniels formula, this places you around VDOT 28. Your equivalent performances: ~62:30 10K, ~2:18 half marathon, ~4:49 marathon.

Example 2 — Intermediate runner: You ran a 22:00 5K. This places you around VDOT 39. Equivalents: ~45:40 10K, ~1:41 half marathon, ~3:30 marathon.

Example 3 — Advanced runner: You ran a 17:00 5K. This places you around VDOT 50. Equivalents: ~35:20 10K, ~1:19 half marathon, ~2:45 marathon.

To find your VDOT: enter a recent race time into a VDOT calculator (or look it up in Daniels' published tables). Then read the equivalent performance at any other distance. Runners at the same VDOT are considered aerobically equivalent — they should be able to produce the same times, distance for distance, on equivalent courses under similar conditions.

These predictions assume equivalent preparation and execution. A runner who trains for 5K but enters a marathon untrained will run significantly slower than the VDOT prediction. The equivalence describes potential, not a guarantee.

Training Pace Zones from VDOT

The second major use of VDOT is prescribing training paces. Each pace zone corresponds to a specific physiological stimulus, and VDOT anchors those zones to your actual fitness rather than to arbitrary percentages of maximum heart rate or subjective effort descriptions.

Zone Purpose Pace Relative to Race Pace When Used
Easy Aerobic base, recovery, long runs ~1:00–1:30/km slower than 5K race pace Most runs (80%+ of volume)
Marathon Pace Race-specific endurance Goal marathon race pace Tempo long run segments, marathon-specific workouts
Threshold (Tempo) Lactate threshold improvement ~15–20 min race pace equivalent Tempo runs, cruise intervals
Interval VO2max stimulus Approximately 5K race pace Track intervals, quality sessions
Repetition Running economy, speed Faster than 5K pace, short reps Short repeats (200m–400m), strides

The easy pace range — 1:00 to 1:30 per kilometer slower than your 5K race pace — is the one most runners get wrong. For a runner at VDOT 35, that's roughly 7:00–7:30/km on easy days. Most runners at that fitness level run their easy days at 5:30–6:00/km, which is threshold territory. They're in the gray zone on days designated for recovery. The VDOT system makes the correct easy pace explicit and non-negotiable.

How the App Generates Predictions Passively

Most runners don't have a recent race time at the distance they're training for. Someone training for their first marathon may only have a 5K from two years ago. Someone who hasn't raced at all needs a different approach entirely. Pheidi generates VDOT estimates using a four-tier confidence hierarchy:

  1. Recent race time (any distance, within 6 months): highest confidence. A race time is the cleanest VDOT input — it represents actual performance under racing conditions. The system calculates VDOT directly and projects to the target distance.
  2. Logged training paces (5+ data points from the last 4 weeks): medium confidence. Consistent workout paces at known effort levels allow VDOT inference. More data points reduce the uncertainty range.
  3. Plan adherence + experience level + completed weeks: low confidence. Without pace data, the system can estimate fitness progression based on training completion and stated experience. This produces a range rather than a point estimate.
  4. Insufficient data: the prediction is withheld entirely until 4 weeks of logged workouts have accumulated.

Predictions are never shown before 4 weeks of training. A VDOT estimate before a meaningful training baseline exists is statistically unreliable and potentially misleading. Better to withhold the number than to show a confident-looking projection that has no real grounding.

— Pheidi prediction display policy

The result is a prediction that updates as evidence accumulates, becomes more precise as training data grows, and displays an honest confidence level rather than false precision. A runner in week 5 might see "Based on your training, you're on track for approximately 4:05–4:15." A runner in week 16 with consistent pace data might see "Based on your training, you're on track for approximately 4:08."

Why the Marathon Is "Slower" Than the Ratio Suggests

A common misconception among newer marathon runners is that marathon time should be roughly 5 times their 5K time. This would imply a 1:5 ratio between 5K and marathon performance. VDOT tables show the actual ratio is closer to 1:5.8 to 1:6.2 for most runners.

1:5.8–6.2 Actual 5K-to-marathon ratio for most runners (not 1:5 as commonly assumed)

The gap between the intuitive ratio and the real ratio reflects several compounding factors unique to the marathon distance:

  • Glycogen depletion. The marathon is long enough that most runners exhaust their glycogen stores, forcing a shift to slower fat metabolism. No shorter race distance produces this constraint. The last 10–15km of a marathon are physiologically different from the first 30km.
  • Pacing complexity. Running 42.2km at the correct even pace is a cognitive and physical skill. Most runners, especially in early marathons, go out too fast and pay for it in the final third. The VDOT prediction assumes optimal execution; most race-day performances fall short of that.
  • Accumulated fatigue. Four-plus hours of continuous effort creates systemic fatigue that 20 minutes of racing does not. The degradation of running form, muscle coordination, and pain tolerance over marathon duration has no equivalent at shorter distances.
  • Course and conditions. Running the tangents — the shortest geometric path through a marathon course — saves approximately 200 meters over a standard 42.2km course. Runners who don't run tangents effectively run longer than the listed distance. Weather, hills, and road surface also compound at marathon distance in ways they don't at 5K.

The practical implication: runners targeting a marathon time based on their 5K should use the VDOT table projection, not mental multiplication by 5. The VDOT projection already accounts for the ratio mismatch. Using 5x as a target will produce a race plan that is approximately 10–15 minutes too aggressive for most runners.

How the Prediction Updates Over Time — and What to Do When It Dips

A VDOT-based prediction isn't a single calculation. It's a continuous estimate that updates as more training data accumulates. The trend over time is more meaningful than any individual week's projection.

A healthy improvement curve looks something like:

  • Week 6: 4:18 (range: 4:12–4:25)
  • Week 8: 4:12 (range: 4:07–4:18)
  • Week 10: 4:07 (range: 4:03–4:12)
  • Week 14: 4:02 (range: 3:59–4:06)

The narrowing range reflects accumulating data confidence. The improving projection reflects real fitness gains from consistent training.

If your prediction worsens in a given week — say, 4:07 becomes 4:11 — this is normal and expected, not a sign of declining fitness. Deload weeks, accumulated fatigue, extreme heat, illness, and disrupted sleep all temporarily suppress training paces. A single week's data point does not represent your fitness ceiling. The trend across 3–4 weeks is the signal; any individual week is noise.

The response to a dipping prediction is never to train harder. It's to ask what suppressed the output: recovery debt, illness, heat, life stress. Address the suppressor. The fitness is almost certainly still there.

RPE Mode: Training Without Pace Data

Not every runner wants to train by pace. Some runners don't use GPS watches. Some run trails where pace is meaningless. Some are early in their running career and find pace targets create anxiety rather than guidance. For these runners, Rate of Perceived Exertion (RPE) is a complete alternative.

RPE describes workout intensity in effort language rather than pace language:

  • Easy: Conversational pace — you can speak in full sentences without pausing to breathe
  • Moderate: You can speak in short phrases but not extended sentences
  • Comfortably hard (tempo): You can say a few words; speaking feels effortful
  • Hard (interval): You can barely speak; this is near-maximal sustained effort
  • Race effort: Maximum sustainable effort for the distance

RPE mode allows runners to follow the full training plan structure — easy days genuinely easy, hard days genuinely hard — without requiring a pace target or GPS data. Workouts are described in effort terms, not time-per-kilometer targets.

In RPE mode, pace-based VDOT predictions aren't available. The app instead estimates fitness progress from training adherence and plan completion, generating a range rather than a point estimate. This honest uncertainty is the correct approach: without pace data, a specific projection number would be false precision.

Runners can switch between VDOT mode and RPE mode at any time. A runner who switches from RPE to pace mid-plan will begin generating pace data immediately; predictions will transition from range-based to point-based as enough data accumulates.

Pheidi calculates your VDOT automatically

Enter a recent race time and get your full training pace prescription instantly. No pace data? Use RPE mode — every workout is described in effort terms, no watch required. The plan adapts as your fitness data grows.

Get Your Free Plan

Key Takeaways

  • VDOT measures "effective VO2max" — both raw aerobic capacity and running economy — derived directly from race performance
  • Any race time at any distance produces a VDOT value that predicts equivalent performances at all other distances
  • A 24:03 5K corresponds to VDOT 35, predicting a 1:50:24 half marathon and 3:52:07 marathon
  • VDOT also prescribes training paces: easy runs are 1:00–1:30/km slower than 5K race pace — slower than most runners expect
  • The app generates predictions passively through a four-tier confidence hierarchy: race time (highest) → training paces → adherence/experience → insufficient data (withheld)
  • Predictions are never shown before 4 weeks of training
  • The 5K-to-marathon ratio is 1:5.8–6.2, not 1:5 — glycogen depletion, pacing complexity, and accumulated fatigue widen the gap
  • A single week of slower paces doesn't indicate declining fitness — look at the 3–4 week trend, not individual sessions
  • RPE mode allows full plan execution without pace data; predictions show as ranges based on training adherence

References

  • Daniels, J. (2014). Daniels' Running Formula, 3rd edition. Human Kinetics. (Original VDOT methodology and tables.)
  • Daniels, J. & Gilbert, J. (1979). "Oxygen power: Performance tables for distance runners." Privately published. (Original VDOT research basis.)
  • Foster, C. et al. (2001). "A new approach to monitoring exercise training." Journal of Strength and Conditioning Research, 15(1), 109–115. (RPE-based training monitoring.)
  • Borg, G. (1998). Borg's Perceived Exertion and Pain Scales. Human Kinetics. (RPE scale foundations.)
  • Midgley, A.W., McNaughton, L.R. & Jones, A.M. (2007). "Training to enhance the physiological determinants of long-distance running performance." Sports Medicine, 37(10), 857–880.