Race Score - An "Race Score” is given to each runner for every race they complete. Each person’s runner file profile consists of many different scores, corresponding to each race they have completed.
ITRA Performance Index – the individual scores of each runner are used to calculate their Performance Index. The performance index changes with the addition of new results and over time as races become older.
ITRA Points – These points are awarded for finishing a race (0 – 6 points), each person who finishes the race receives the same number of points. This article does not discuss ITRA points, it describes only the Race Score.
The ITRA score is calculated based on a runner’s finish time VS the specific characteristics of the race he has participated in.
Scores are built on a scale up to a maximum of 1000 score with the top of the scale corresponding to the theoretical best possible performance.
It's calculated in two steps, and takes into account different parameters :
1) Measurable factors : The distance & elevation data are turned into a flat equivalent distance (km-effort)
- Distance
- Elevation change (up & down)
- Minimum, average, and maximum altitude
For each flat equivalent distance, there is a corresponding world record, based on official athletics world records.
2) Non objective factors : These factors are translated into a coefficient of adjustment after ITRA algorithm's analysis
- Terrain type (dry, muddy, snowy, forested, sandy, etc.)
- Racing conditions (temperature, weather, humidity, wind, etc.)
- Any other factors directly influencing performance
These non objectives factors are mathematically translated into a coefficient of adjustment that reflects all the factors affecting a runners' performance.
The ITRA score calculation : Race's finish time for a flat equivalent distance + race condition coefficient of adjustment = ITRA score
1) Flat equivalent world record
A runner's finish time is compared with the theoretical flat equivalent best possible performance for that race. It gives the raw ITRA score value.
2) Coefficient of adjustment
A statistical analysis taking into account all the non objectives factors influencing the performance is done thanks to ITRA's algorythm.
We name this step the "coefficient of adjustment"
This coefficient adjusts the raw ITRA score value previously obtained and calculates the final ITRA score.
Our calculation uses an indirect method, based on the statistical analysis of our database of more than 8 million individual results, to also account for this factor.
The same algorithm is used to calculate the score for all races.
Each race has its own adjustment coefficient, which is applied equally to all runners in that race.
This method has two major advantages:
- we can consider any factor influencing the race time.
- our data base and scoring system is constantly improved with every race.
Please note that the following factors do not affect the score obtained during a race :
- Your race ranking
- Number of withdrawals (dnf)
- Presence of elite runners or not
- Event's fame
Not necessarily !
Our algorithm estimates a runner's performance, rated on a scale from 0 to 1000, based on several factors:
- Distance
- Elevation change
- Minimum, average, and maximum altitude
- Terrain type (dry, muddy, snowy, forested, sandy, etc.)
- Racing conditions (temperature, weather, humidity, wind, etc.)
- Any other factors directly influencing performance
These mentioned factors can significantly affect a runner's time and performance.
Some of these factors can be measured, some are processed through statistical analysis.
- Distance, elevation, and altitude are measurable factors. (using GPX files and ITRA's race evaluation service)
- Terrain technicality and racing conditions are not objectively measurable (what’s technical for one person may not be for another).
Our algorithm uses statistical analysis to account for these variables by evaluating the performances of all runners in a race.
It compares performances from that race to previous performances of the same runners on similar km-effort courses.
This analysis translates non-quantifiable factors into an adjustment figure, creating a performance score equivalent to a world record reset.
For each race, the algorithm identifies a trend:
“The runners on this course are on average 10% faster or slower than on the same exact course or equivalent courses in a different year.”
This adjustment mathematically reflects all the factors affecting runners' performance - and the final ITRA score obtained.
Typically, scores on an identical course from one year to the next are very similar.
However, if racing conditions significantly impact performance, the same time may not yield the same score.
This is because the runner's time is directly influenced by the unique challenges of trail running.
Example:
A technical route (limestone terrain, forest, narrow dirt path) of 20km with 1000m elevation gain.
Year N: Ideal conditions—dry ground, no wind, temperature around 15°C, good visibility.
Mr X finishes the race in 3:05:00 and scores 590.
Year N+1: Same exact course with difficult conditions—muddy, slippery terrain, continuous rain, temperature around 5°C.
Mrs Y finishes in 3:15:00 but still scores 590 because the conditions impacted performance.
In conclusion, your score is based on your time relative to the course characteristics and its own unique conditions
The following factors do not affect your score regardless of the race:
- Your race ranking / Age group / Your gender
- Number of withdrawals
- Presence of elite runners or not
- Event's fame
No, the score for a single race does not change over time*
This makes it possible to compare directly any of your race scores over any time period.
*HOWEVER :
An adjustment may be applied to the score of an individual race upon the receipt of new information, most commonly due to a change of course from the planned route to actual route or a mistake in the result file.
An adjustment can take place at any time, but they are extremely rare.
A score is calculated for all races in our calendar upon the receipt of complete and correct results files from the event organizer as well as accurate race data (distance - vertical gain / loss).
Exceptions: At the moment, no score is calculated for the following races:
- Races where the km-effort (distance + (elevation profile/100)) is less than 5.
Examples:
2km/+100m = km-effort – No score.
2km/+500m = 7 km-effort – Yes , score given
- The race has too many stages
- Virtual races
- The number of finishers is too low (less than 5)
- The data in the result file contains too many errors
Publishing of the results file is a free service; however, the results must be submitted according to the ITRA results format.
Our years of race analysis tells us that there is no such thing as a similar course.
The Race Score relates to the performance of a runner for a specific race in specific conditions and will depend many things not just on the distance and elevation profile.
Two races may be advertised with the same distance and the same elevation profile, but in reality differ quite considerably (technical difficulty of the course; the relationship of elevation gain to distance (4 x 1000 m or 8 x 500m) etc.), the humidity, the type of terrain, the average altitude... BAsically, any factor having an influence over the performance !
Our statistical correction coefficient makes allowance for these differences as well as many other technical aspects.
Therefore, the comparison of time between two apparently similar races is not possible.
That's exactly the purpose of our algorithm, to compare what apparently looks to be comparable, but is not in reality.
