In all human endeavors repeatability and a sense of control are generally sought after. Even in a game as chaotic as Aussie Rules Footballs; strategies, techniques, fitness and numerous other elements are all developed to maximize control and win.

Running performance has been traditionally based on either a VO2max test, or previous running. This is somewhat a limiting data set when trying to set out and create proper goals.

Bundle, Hoyt and Weyand; 2002 demonstrated that using max speed runs, they were able to use the anaerobic speed reserve to accurately predict runs of sub 4min (and were 97% accurate at 100m, 200m and 400m).

The authors point out early in their article that the best sprinters run twice as fast as the best milers, but that the best milers are not much faster than the best marathoners. Clearly, anaerobic factors make up the difference. Realizing that tests for runs 10-120min had been well established on the literature, they set out to study the relationship between anaerobic reserve and running speed.

Using all out treadmill runs and track runs, they found a generalized relationship:

Spd(t) = Spdaer + (Spdan – Spdaer) e(Kt)

T is the duration of the all-out run Spd is the speed maintained for that run (generally found to be about 8 steps or 3 secs), Spdan is the maximum speed supported by anaerobic power while Spdaer is the max speed supported by aerobic power, e is the base of the natural log, k is an exponent used to describe the decrements in observed speed and 0.013 was found to fit best.

The subjects ran for 5 min at 2.5m/s rested for 5 min and then the speed was increased bby 0.5m/s. This was continued until they were unable to complete a 5min run. Air from minutes 4 and 5 were collected and analyzed. Using linear extrapolation, they found metabolic cost of the maximum running speed above the maximum oxygen uptake.

Maximal anaerobic power was determined to be the max speed subjects could maintain on a treadmill for 8 steps without any backward drift. Again, linear extrapolation was used to determine metabolic cost.

Maximal aerobic power was determined by a progressive, discontinuous treadmill test (as above), where the highest completed minute value was used as the base for the extrapolation.

Subjects then performed 10 all out runs on a treadmill at speed while harnessed. The runs lasted from 3s to 240s. After this was completed, track runs of 100m, 200m, 300m and 400m were done to determine maximum over-ground speed. A gate system and 50m and 55m were the basis of this trial.

The R^2 value for the subjects comparing predictive treadmill running to track running was 0.91. It was found that the treadmill based equation was accurate to within 3.7% of the predicted value and the track 3.3%. Both of these methods being well above 95% can be accurately used to assess speed running.

Several track teams in the US now base their pre-season training on this model and have used it with great success (of note, David Cox and Chris Jones).