Dear Slate:

This may be helpful

Actually, it makes a lot of sense.  Imagine two people with a 25:00 5K PR.  We'll assume that both people set their PR's within the last month on a flat course with good weather conditions.  However, one person is a 25 year old male who runs 15 miles per week while the other is a 40 year old female who runs 50 miles per week.  McMillan would spit out the exact same predicted time for them in a marathon but we know that the 40 year old female in this case would crush the 25 year old male in a marathon because he doesn't have the mileage base needed to run it.

What this person is trying to do is to build a calculator based on a statistical model with numerous factors impacting the model including age, gender, mileage per week, and whether you do speed work.

I understand that I am just not convinced that just because you are female or older you would do better at distance based solely on those two factors.

 

If for instance you currently specialize in the 800, it is not automatic that you will do better at the marathon because you aged.

 

Although to the articles credit they did say relatively speaking (I initially read right past that) so there is that.

Exactly. Would the 40 year old woman do better at the marathon because she is female and older, or because she runs 50 miles a week, unlike the young guy who only runs 15? I can see gender and age being a factor in speed. Being a young male is ideal. But in distance, how can being old and/or a woman help you? You can say that women prefer training for distances, rather than speed. And that older people have more time to train. But that doesn't make them "better" at distances.

 

 But that doesn't make them "better" at distances.

No, it makes them worse at shorter distances. Statistically if a 25yo male and a 40yo female can both run a 3 hour marathon, chances are the male is faster at 5k. That is, as we age, the difference in pace b/w distances gets closer (i.e. worse relatively at shorter stuff). So, it's not that age/gender makes one better at longer, but "worse" - again, relative to performance degradation - at shorter stuff.

Though also, women, statistically, crush men overall at really long distances (see wins, high placement, at ultras).

 

No, it makes them worse at shorter distances. Statistically if a 25yo male and a 40yo female can both run a 3 hour marathon, chances are the male is faster at 5k. That is, as we age, the difference in pace b/w distances gets closer (i.e. worse relatively at shorter stuff). So, it's not that age/gender makes one better at longer, but "worse" - again, relative to performance degradation - at shorter stuff.

 

Though also, women, statistically, crush men overall at really long distances (see wins, high placement, at ultras).

Yes, I agree 100% with the first part of your post.

For the second statement, I wonder though if it's not simply that some very good female runners (are we talking very young runners or more like late 30s, 40s?) tend to gravitate around ultras because they know they cannot shine anymore in shorter distances. A man who runs a sub-20 5k is hugely congratulated (I'm talking about any recreational runner here, on our forums), especially if he's older. He's a stud, etc, etc... But a woman runs a 21:00 5k and we pat her on the head and say good work little girl. So, damn it, we older females have to turn to ultras for our minute of glory.  (though not me, I would stink at both 5ks and ultras)  They can outperform the younger crowd because they have the time for proper training, the running experience (both mental and physical) and their bodies can still operate close to 100% when it comes to distance. But again, they are not necessarily "better" than anyone younger who would put in the same amount training. But maybe I'm wrong. You have the stats and I don't. I was just rambling. I'm at least good at that. 

I don't know that he is making the assumption that women and older folks are better at longer distances.  Maybe they are and maybe they aren't.  What he appears to be attempting is to collect data from thousands of runners.  Using that statistical data, you could then make a better race prediction for someone.  So, rather than just inputting a time like you do with most running calculators, you would input a race time, age, gender, miles per week, whether you do speed work, etc.

What I can guess though is if LRB kept training for the 5k and I the marathon, I know he'd smoke me in the 5k regardless of what the current calculators predict for my 5k time.

While you would smoke me in the marathon regardless of what my predicted time was.