Race simulator

The race simulator estimates a finish time and kilometre-by-kilometre splits from the course, the athlete's history, and current form. It is available from the Race simulator menu.

Forecast, not a promise

A simulation supports race strategy; it does not guarantee a result. Weather, actual technical difficulty, aid stations, current condition, and race-day circumstances can all change performance.

Before you begin

The forecast becomes more relevant when MINDRAW has:

• recent and comparable running or trail activities;
• enough analysed activities to build the runner profile;
• a recorded reference VO₂max;
• a check-in completed on the day of the simulation.

MINDRAW first looks for comparable runs among the latest 60 activities. If none are available, it uses the runner profile, followed by the reference or estimated VO₂max. If none of these sources is available, it uses a generic pace and confidence remains limited.

Describe the course

There are two ways to provide the course.

Import a GPX file

Import the official or prepared route as a .gpx file, up to 10 MB. MINDRAW extracts:

• distance;
• elevation gain and loss;
• the elevation profile;
• the location of climbs and descents along the course.

When a GPX is loaded, its values replace the manually entered distance and elevation. The race name can still be customised.

Enter the values manually

Enter:

• the race name;
• a distance between 0.1 and 500 km;
• elevation gain and loss;
• terrain: road, rolling, trail, technical, or mountain;
• optionally, the race date.

Without a GPX, MINDRAW does not know where the actual climbs and descents occur. Elevation is therefore distributed uniformly in the forecast. The date is currently informational and does not yet change the calculation.

Run and read the simulation

Select Simulate my race. MINDRAW produces three scenarios:

Scenario	Intended use
Ambitious	an aggressive target with a higher margin of risk
Probable	the central estimate from the available data
Conservative	a cautious strategy with more margin

Each scenario shows a finish time and average pace. Selecting a scenario also updates the kilometre split table.

Each split includes:

• suggested pace and cumulative time;
• elevation gain and loss;
• average gradient;
• expected effort type: flat, uphill, steep climb, or descent.

With a GPX, the elevation chart helps identify the sections that will influence the strategy most.

Understand confidence

Confidence is displayed between 35% and 90%. It notably increases with:

• the number of comparable runs, up to eight;
• a runner profile built from at least ten sessions;
• a real GPX profile;
• a recent check-in.

The number of comparable runs used is displayed with the result. High confidence means MINDRAW has more context; it does not turn the forecast into a certainty.

What affects the time

The local engine notably accounts for:

• distance and endurance observed in the athlete's history;
• terrain type;
• the gradient of each section when a GPX is available;
• mountain experience recorded in the runner profile;
• training load and form balance from the runner profile;
• energy, sleep, stress, motivation, and focus reported in the check-in.

Progressive fatigue is also applied near the end of the course. The model does not, however, know future weather, trail conditions, or the athlete's fuelling strategy.

AI analysis and PDF report

After choosing a scenario, Analyse with AI proposes a race strategy covering the start, important sections, the middle of the race, and the finish. This feature requires an AI provider to be enabled by an administrator.

AI interprets the selected local simulation. It does not replace or recalculate the times produced by the MINDRAW engine.

Download PDF report exports the course, all three scenarios, confidence, kilometre splits, and the AI analysis when one has been requested.

Good practices

• Prefer the official GPX over advertised elevation when available.
• Complete the daily check-in before simulating.
• Compare the probable and conservative scenarios first.
• For trail races, keep extra margin for technical terrain and stops.
• Run the simulation again after adding new representative activities.