Skip to content
Slow Run Fast

Training Volume Race Predictor

Enter your average weekly training distance and training pace over the last 8 weeks to predict your marathon time, using the Tanda (2011) model.

Calibrated for roughly 30 to 130 km per week of consistent training over 8 or more weeks.

Predicted marathon time

3:31:18

5:00 /km

Other distances (less certain)

5K, 10K, and half marathon times are derived from the predicted marathon time using Riegel's formula, not the Tanda model directly. Treat them as more uncertain than the marathon prediction itself.

DistancePredicted timePace /km
5K22:024:24
10K45:564:36
Half Marathon1:41:214:48

This estimate uses a generic formula. Connect your Strava/Whoop/Garmin data for a version personalised to your actual training.

Get early access

What a training volume race predictor is

Instead of predicting a race time from another race result, this calculator predicts your marathon time directly from your training itself: how much you have been running each week, and how fast those training runs have been. It answers a different question than the Race Time Predictor does: given your actual training, not a recent race, what marathon is realistic right now.

How it is calculated

This calculator uses the Tanda (2011) model: marathon pace in seconds per km = 17.1 + 140.0 × e^(-0.0053 × K) + 0.55 × P, where K is your average weekly training distance in km over the last 8 weeks, and P is your average training pace in seconds per km over the same period. Marathon time is that predicted pace multiplied by 42.195 km.

5K, 10K, and half marathon times are not part of the Tanda model, which is fit to marathon performance specifically. Instead, this calculator feeds the predicted marathon time into Riegel's formula, the same one used by the Race Time Predictor, to estimate those shorter distances. That extra step makes them less certain than the marathon prediction itself.

Why runners use it

It gives a training-based marathon target without needing a recent race result, and it reflects the training block you are actually in, rather than a single result that may be weeks or months old.

Sources and limitations: see our methodology.

Worked example

Take a runner averaging 60 km a week at a 5:30 /km training pace, both averaged over the last 8 weeks.

8-week average weekly distance60 km
8-week average training pace5:30 /km
Predicted marathon pace5:00 /km
Predicted marathon time3:31:18
Predicted half marathon (less certain)1:41:21

The Tanda model takes this runner's 60 km weeks at 5:30 /km straight to a predicted 3:31:18 marathon. The half marathon figure is a second-order estimate on top of that, Riegel's formula applied to a prediction rather than a real result, so treat it as a rough guide rather than the main output of this calculator.

FAQ

What is the Tanda formula for marathon prediction?

The Tanda formula, from Giovanni Tanda's 2011 paper, predicts marathon pace from training data rather than a race result: marathon pace (seconds per km) = 17.1 + 140.0 × e^(-0.0053 × K) + 0.55 × P, where K is your average weekly training distance in km over the last 8 weeks, and P is your average training pace in seconds per km over the same period.

How accurate is a marathon prediction from training volume?

The Tanda model is fitted to recreational marathon runners training in roughly the 30 to 130 km per week range over 8 or more weeks of consistent training. Inside that range it gives a reasonable training-based estimate, but like any fitted model it gets less reliable the further your training falls outside the data it was built from.

What counts as my weekly training distance for this calculator?

Use your average total weekly running distance over the last 8 weeks, not just one week. A single unusually high or low week will not show up correctly here since the model expects an 8-week average, not a single data point.

What pace should I enter as my training pace?

Enter your average training pace, meaning the typical pace of your regular training runs over the last 8 weeks, not your race pace or your fastest interval pace. The Tanda model was built from everyday training paces, so a race-pace input will skew the prediction faster than it should be.

Can I use this to predict a 10K or half marathon time?

Yes, but treat those numbers as secondary. This calculator predicts your marathon time directly from the Tanda model, then feeds that prediction into Riegel's formula to estimate 5K, 10K, and half marathon times. That is an extra layer of estimation on top of the marathon prediction itself, so it is less certain than the marathon number.

How is this different from the Race Time Predictor?

The Race Time Predictor starts from one race result you already have and scales it to other distances with Riegel's formula. This calculator starts from your training itself, weekly volume and training pace, with no race result required, and predicts a marathon time using the Tanda model instead.