wellnow wrote:
I will never be a believer in mathematical formulae for predicting race times or training times.
I'll try to tone it down a little, but here's a practical example of applied mathematics.
I know a guy, who's a pretty decent masters runner. Here are some recent race times, over the last few years, and the corresponding VDOTs:
800 02:06.3 67.34
1500 04:00.7 69.87
5000 15:15.7 68.22
10000 31:56.0 67.64
21.0975 1:09:24 67.02
42.195 2:29:50 66.40
As you can see, according to his VDOT scores, he is quite consistent at all distances, around 67-68, showing a little preference for the 1500m. Even 800m is in line, but note VDOT is not as reliable for 800m.
You might expect this kind of broad consistency from a seasoned veteran.
Note, he describes himself the same way, as a stronger middle distance runner than marathon runner, so VDOT is not way out of line for him.
The same table with Purdy points:
800 02:06.3 678.32
1500 04:00.7 819.76
5000 15:15.7 805.17
10000 31:56.0 809.1
21.0975 1:09:24 808.55
42.195 2:29:50 803.44
Using a completely different reference, we see the same consistency across the board, with a showing of strength at 1500. (Note Purdy is more reliable for 800m, so a 2:06, while still a very respectable time for a masters athlete, is relatively worse than the other times).
Now, for example, if we saw high numbers for 800/1500m, and a consistent degradation at longer distances, as is often the case with younger runners, we might conclude he has what "Hadd" calls, a "loose" or "no relationship" between his race times, and would benefit from a period of Hadd's Phase 1 training, or Lydiard's marathon training to build endurance and stamina. This usually improves times across the board, while bringing the relationship of the longer events closer.