RandomXCObserver wrote:
Pretty sure Speed Ratings compare runners to runners not runners to courses.
Yep.
Speed rating is a ratio of velocities (or inverse ratio of course times).
SR_i = 100 x V_i/V_m = 100 x (D/T_i)/(D/T_m) = 100 x T_m/T_i
where i = individual runner being measured
m = median runner in the race that day. (Median is middle person in race - 50/100, 20/40, etc.)
So if SR = 150 , that person's velocity was 50% faster than the median. (And the median time was 50% slower).
There is not much theory to it, except that there should be many runners near the median,
so the median time should have a low variance.
And the median time is not sensitive to the fastest or slowest runners.
It may be able to adjust for weather and course conditions, since those should affect both the person
being measured and the median runner proportionally.