2017 Updated Age Factors

The age factors in Table 1 in Fair (2007) have been updated through 2017 for the "run" category, which is all running events between 5km and the marathon. The updated estimates are based on world single age records for men compiled by ARRS: World Single Age Records. Four road races were pooled---5km, 10km, half marathon, marathon---along with two outdoor track events---5000 meters and 10000 meters. The data are as of June 2, 2017.

The same methodology was used as in Fair (2007), and this discussion will not be repeated here. Only non dominated times were used, and the six races were pooled. There were 8 observations for ages 92 and above. The oldest age was 94. The total number of observations was 216: 33 for 5km, 33 for 10km, 37 for HMar, 36 for Mara, 41 for 5000m, and 36 for 10000m. The decline rate is assumed to be linear in percent terms up to some transation age, which is estimated, and then quadratic after that. The (nonlinear) regression is constrained to have all the estimated errors on or above the line. The estimated transation age was 77.2, which compares to 75.1 for the 2007 results. In general the results are fairly close. The following compares the 2017 estimates to the 2007 estimates. This table is similar to Table 1 in Fair (2007). The 2007 row is the same as the "Run" row in Table 1.

                    2017 versus 2007 Estimates
alpha^age^delta^R40R50R60 R70R80R90R100
20070.008075.10.001641.041.131.221.32 1.492.244.68
20170.007677.20.002571.041.121.211.31 1.442.316.24

The R's are the age factors at the various ages. The age factors are quite close until about age 90. The updated estimates are more pessimistic about decline rates in the very old ages. There are few observations beyond age 90, and some of the existing records are undoubtedly soft. The estimates beyond about age 90 are thus not trustworthy. More observations at the very old ages are needed before much can be said.

Go to calculations