As I poorly stated in my initial comment I am aware of this problem. My hope was that by assuming virtually nobody ran the race 10 years ago, which I assume/hope is totally inaccurate, the two errors would cancel each other out.
It's a 'back-of-a-napkin' calculation. Certainly not accurate but a nice place to start for future refactoring.
I'm not saying his reasoning is sound, but your rebuttle isn't either.