The ITRA Performance Index is calculated by obtaining the weighted mean of up to the 5 best scores achieved by a runner over the previous 36 months (for the general index or the index by category).
A weighted mean is used because we give more importance to the most recent results compared with older ones and we also give more importance to a runner’s best result, then to their second best, down to their 5 best ones.
Runners who do not have 5 results will still have a PI, which is calculated using their available results, but a runner with at least 5 results will have an advantage over someone with less results.
This period of 36 months is long enough to permit reliable statistical calculations and allow an injured runner to continue to appear in the performance index due to their previous performances.
We use a weighted average to calculate the Performance Index (PI), which gives more importance to:
Recent race results over older ones
A runner’s best races over their lower-scoring ones (up to their 5 best)
Even if a runner has fewer than 5 races, a PI is still calculated.
However, runners with 5 or more results may have an advantage because the calculation uses more data - the more races you run, the better we can estimate your real level!
Looking at races over the past 36 months helps ensure the PI is reliable while still including runners who may have missed some time due to injury.
Method of calculation
Overall Goal:
The process gives more importance to recent and better performances and adjusts for the runner’s racing experience.
Step-by-Step Process
Step 1: Collect Valid Race Scores
Gather all race scores from the past 36 months.
Select the top 5 best ITRA score performances
Ignore any races where:
The runner Did Not Finish (DNF).
The score is 0.
Step 2: Apply Time Weighting
Assign a weight to each score based on how long ago the race took place:
Race Age (Months Ago) | Time Weight |
|---|---|
0–11 months | 1.000 |
12–17 months | 0.995 |
18–23 months | 0.990 |
24–29 months | 0.985 |
30–35 months | 0.980 |
Multiply each race score by its corresponding time weight.
Step 3: Rank the Scores
Sort the time-weighted scores from highest to lowest.
Assign each result a rank based on its position in the list:
Rank 1 = Best score, Rank 2 = 2nd best, ..., up to Rank 5.
Step 4: Apply Experience Weighting
Simulate different scenarios assuming the runner had:
1 result in the last 36 months
2 results in the last 36 months
3 results in the last 36 months
4 results in the last 36 months
5 results in the last 36 months
For each simulation, apply experience weights to the top scores:
# of Races Used | Score Weights Applied |
|---|---|
1 | 1st = 0.97 |
2 | 1st = 0.99, 2nd = 0.98 |
3 | 1st = 1.00, 2nd = 0.99, 3rd = 0.99 |
4 | 1st = 1.01, 2nd = 1.00, 3rd = 1.00, 4th = 0.99 |
5 | 1st = 1.02, 2nd = 1.01, 3rd = 1.00, 4th = 1.00, 5th = 0.99 |
For each scenario, calculate the weighted average of the top scores using these experience weights.
Step 5: Select Best Simulation
Out of the 5 simulations (1 to 5 top scores), choose the one that gives the highest overall PI.
This PI becomes the runner’s final ITRA Performance Index level
Note on Distance Categories
Repeat the entire process separately for each Distance Category (e.g., 5K, 10K, Marathon).
In Step 1, only include races that belong to the relevant distance category.
Found this useful? Share it