Most runners train by feel: run easy days easy, run hard days hard, do the long run on weekends. This is not wrong, but it leaves training intensity targets vague. Without specific pace zones derived from current fitness, hard days are often not hard enough to drive adaptation, and easy days are often not easy enough to enable recovery.
Race prediction formulas -- particularly Riegel's formula, which powers most online race predictors -- give you a data-based method for deriving specific training pace targets from a recent race effort. This article walks through how to use prediction outputs practically to set tempo, threshold, interval, and easy pace zones for an entire training block.
Step 1: Establish Your Current Fitness Baseline
The starting point is a recent time trial or race run at full effort. The input must reflect your genuine current fitness, not a training run or a race where you paced conservatively for other reasons.
Acceptable inputs:
- A recent 5K race result
- A recent 10K race result
- A 1-mile or 2-mile time trial on a track
- A half marathon run at full effort
Run within the past 4-6 weeks is ideal. Fitness changes enough over 8-12 weeks that older race results may generate training targets that are either too easy (if you have improved) or too aggressive (if fitness has declined due to injury or reduced training).
For detailed context on how the formula works and where its 1.06 fatigue exponent comes from, read Riegel's Formula Explained: How Race Time Prediction Actually Works.
Step 2: Generate Predicted Race Times
Using your recent race time, run the prediction formula for several distances:
For a 5K input of 25:00 (5:00/km, or about 8:02/mile), Riegel's formula gives:
- Predicted 10K: approximately 52:04
- Predicted half marathon: approximately 1:54:15
- Predicted marathon: approximately 3:55:00
The Pace & Race Time Calculator -- use it here -- performs these calculations automatically and outputs the corresponding per-kilometer and per-mile paces for each distance.
These predicted times anchor your training zones.
Step 3: Derive Pace Zones from the Predictions
Classic training zone frameworks use percentages of threshold or VO2 max pace to define easy, moderate, tempo, and interval targets. Race predictions give you the reference points needed to calculate these zones.
Easy / recovery pace: 65-75% of effort, or approximately 75-90 seconds per mile slower than 5K race pace. For a 25:00 5K runner (8:02/mile average), easy pace is approximately 9:20-10:00 per mile.
Aerobic / general endurance pace: 75-80% effort, approximately 45-75 seconds per mile slower than 5K race pace. For the same runner: 8:47-9:17 per mile.
Marathon pace: The predicted marathon pace from the formula. For a 3:55 marathon prediction: approximately 8:58 per mile. This is the pace to practice in the later portions of long runs and in marathon-pace workouts.
Half marathon pace: Predicted half marathon pace. For a 1:54:15 prediction: approximately 8:43 per mile. This is the target for longer tempo runs of 40-60 minutes.
Tempo / lactate threshold pace: Approximately 10K to half marathon effort pace, around 85-90% intensity. For this runner: approximately 8:15-8:43 per mile.
Interval / VO2 max pace: 5K race pace or slightly faster. For this runner: 8:02 per mile or slightly faster. These are for shorter, intense repeats (400m-1200m) with recovery between.
Photo by Vladvictoria on Pixabay
Step 4: Structuring the Training Week
With pace zones derived, a structured training week might look like this for a marathon-focused block:
Monday: Rest or easy recovery run (9:20-10:00/mile)
Tuesday: Interval work at 5K-pace or faster (400m-800m repeats)
Wednesday: Easy run (9:20-10:00/mile) + strides
Thursday: Tempo run at half marathon pace (8:43/mile) for 30-40 minutes
Friday: Rest or easy run
Saturday: Long run with progression (easy first half, marathon pace second half: 8:58/mile)
Sunday: Easy recovery run (9:20-10:00/mile)
Every pace target in this week is derived from the single 5K baseline input. This is the advantage of prediction formulas: one recent race result produces a calibrated, internally consistent set of pace targets for the entire training block.
Step 5: Tracking Progress Across the Training Cycle
The other use of prediction formulas is tracking fitness change. Run a 5K time trial at the start of a training block, then run another one 8 weeks later. The change in predicted times at other distances tells you how your fitness has evolved.
If your 5K time trial drops from 25:00 to 24:15 over 8 weeks:
- 5K improvement: 45 seconds
- Predicted marathon improvement: roughly 3-4 minutes (the formula translates a 5K improvement proportionally to longer distances)
This gives you a quantitative measure of training effect, which is more precise than comparing the feel of two long runs separated by weeks of different weather and fatigue levels.
Runner's World training plans and most structured marathon programs build in time trial assessments for exactly this purpose. Tracking the prediction trend across a cycle tells you whether training is working and whether your goal for the race is still realistic. Reviewing Strava activity data alongside time trial results provides additional context: average pace in long runs, heart rate trends, and training load all inform whether fitness gains are translating to improved predictions.
Step 6: Adjusting Goals as the Race Approaches
Race predictions are not fixed targets. They should be revisited as new race results or time trials come in.
If you started a 16-week marathon block with a 3:55 prediction from a 5K time trial and your 8-week 10K tune-up race produces a 3:48 prediction, your goal has improved. Revise the marathon goal and recalculate your pace plan accordingly.
If the tune-up produces a slower-than-expected time, investigate before downgrading the goal. One bad race under poor conditions is not necessarily a fitness measurement. Two bad results in a row warrant goal revision.
For predictions to be useful tools, they need to be treated as live inputs that get updated as new data comes in -- not as fixed commitments made at the start of a training cycle.
Photo by RyanMcGuire on Pixabay
Common Mistakes When Using Prediction Formulas
Using a slow input race: A 10K where you paced conservatively for a training run rather than racing hard produces a pessimistic prediction. Use only full-effort race results.
Ignoring training specificity: A 5K time from a track-focused runner predicts a marathon time that assumes full aerobic development at the marathon distance. If long run volume is low, adjust the prediction upward by 10-20 minutes.
Treating predictions as guarantees: The formula predicts what is physiologically possible given recent fitness. Race conditions, fueling, and execution determine where you actually land. The formula is a ceiling, not a floor.
Not updating predictions: Using a 4-month-old race result as the input during a training block where fitness has changed significantly gives you stale pace targets. Run a time trial or use a recent tune-up race every 6-8 weeks to keep your training zones calibrated.
The Practical Value of Data-Based Training
The alternative to data-based training is effort-based training calibrated by feel. Feel is valuable, especially for experienced runners with good self-awareness. But feel does not give you a number to write in a training log, share with a training partner, or compare across weeks. A prediction-derived pace zone does.
For the mathematical foundation behind race time predictions and why the 1.06 exponent was chosen, read the full breakdown at Riegel's Formula Explained: How Race Time Prediction Actually Works. To generate your own predictions and derive your training pace zones from a recent race result, visit the Pace & Race Time Calculator.
The data is already there in your last race result. This is how to use it.
Top comments (0)