Marathoning Reimagined #1

[This is not organized and not meant to be more than a thought experiment.]

Racers divide into two camps: those who go for records and those who go for wins. These are completely different mentalities and lead to different ways of training. The list of marathon world record holders and Olympic champions are quite different (Lopes, Bikila and Kohlemainen are the only cross-overs among men, Takahashi the only woman). This difference in mindset is present in writing as well; world record holders tend to write autobiographies about the struggles and obstacles overcome outside of running, whereas the Olympians and their coaches write the training manuals.

Training manuals, then, become plans that lead to running well on one particular day in the future, which is what most marathoners want; they pick a marathon and train for that particular race. The fastest runners, however, train to get in peak shape and then look for a race; if a 2:12 marathoner calls a marathon race director a week before a race, he gets in (free, and usually with expenses paid). While this is not possible for most runners and most races any more (I used to "jump in" to a marathon in the 1980's fairly often), I think something approaching that is still possible by racing marathons far more often than the 1 or 2 per year most people attempt.

If you're going to be a marathoner, I say don't dabble in it, do a lot of them. Every coach has pointed out that there is no training very specific to racing a marathon. That's why I think racing marathons as part of training is key - it's specific. One of the problems of marathoners is trying to guess what kind of racing shape they're in and shorter races aren't a good indicator, nor are the various other benchmark workouts (such as Yasso 800's); I say: start by running a marathon, then train to run that time. One never runs a perfect race, so by training to do better what one already can do, one should improve.

The frequency of marathon races is dependent upon training mileage. Racing 4-5% of one's mileage is reasonable. I've found - completely empirically, there's no theoretic basis at all - that marathoners do best when running an average of 75 minutes per day (assuming one run per day). This leads to the following frequency of races by finish time:

2:00 finish every 5-6 weeks
3:00............... 7-8
4:00............... 10-11
5:00............... 11-14
6:00............... 14-17

An old shorthand which works well for marathoners in the 2:30-3:30 range is that one trains at an average of one minute per mile slower than one's marathon pace. Using this and racing 5%, one comes up with

2:00 finish is done at 94 miles/wk and 5-6 weeks between races
3:00.................... 67...................8
4:00.................... 52...................10
5:00.....................42.................. 12-13
6:00.....................35...................15

Which matches the above well. For the slowest runners, this is not far different from the 17-20 week schedules commonly employed. [If one's running more than once per day, mileage can go up to 130 miles per week, which would lead to a race every 4 weeks, which by Jack Foster's rule of 1 day off per mile raced, would be the upper limit.]

This type of continuous racing works only if one acknowledges that one is only as good as one's last race. If one finishes a marathon slower than the one before, one needs the recovery of training more slowly. Areas with seasons of inclement weather automatically have a safety check, as one is forced to run more slowly - but one has to race locally as well. It's not possible to find a local marathon whenever one wants one, so one has to resort to time trials on a course of one's own devising.