Does Age Grading Work? (Part 2 of 3)
A Deep Dive utilizing the 2017 Chicago Marathon Data
Introduction
In the previous post, I presented the finishing time distributions by age group for the 2017 Chicago Marathon. The distributions were compared before and after age grading adjustments.
In this post, I’m taking a closer look at the age graded results for the faster runners within each age group. My arbitrary cutoff for “faster” runners is an age graded time of 3:00 or less for men and 3:30 for women.
Age Graded Results
In the tables below, I calculate the percentage of finishers within each age group who achieved the age graded equivalent of a 3:00 marathon time for men and a 3:30 marathon time for women.
Keep in mind this table is prepared after age grading all results. For example, if a runner ran a 4:00 marathon and the age grading factor for his age group was 0.75 (this is the approximate factor value for a 69 year old male), then the age graded time would be 3:00.
A reasonable expectation would be that the percentage of athletes achieving these cutoff times would be fairly consistent across age groups if the age grading adjustments are working well.
For the men’s table, the results are fairly consistent. The first row shows that 6.2% of all men achieved an age graded 3:00 time or better, and most of the individual age groups are quite close to this with the largest outliers occurring for age groups with relatively few participants.
For the women’s table, the first row shows that 6.4% of all women achieved the cutoff time, but the results are less uniform across age groups. The percentage of women achieving the cutoff time increases for the older age groups (over age 50). I’ve circled these values in red in the table.
Conclusion
When looking at the fastest runners, the age graded results for men show reasonable consistency across age groups. The results for women are less consistent with a higher percentage of finishers hitting the age graded 3:30 threshold within the 50+ age groups. This could indicate the age grading adjustments are overcompensating for women within this age range.
The WMA tables do indicate a much larger adjustment for women’s times than men’s times for runners over 50. For example, the age grading factor for a 70 year old man is 0.7405, and for a 70 year old woman it is 0.6596. This difference could be due to real differences in aging between the men and women, but it may also be that the women’s world records (used for calculating the factors) for these older ages are soft due to lower historical participation.
In the next post, I am taking a final look at the age graded results. I will be focusing on the overall top finishers before and after age grading. My initial expectation is that, in a race with a strong elite field like the Chicago Marathon, there should typically be very little (if any) change in the top 10 finishers after age grading. The elite runners have both extreme talent and high incentives to train themselves to approach their physiological limits. I think it should be quite rare that an age grouper will be able to match them on both of these dimensions.
