HM Pace vs. Marathon Pace (Read 1041 times)

A runner who’s screen name is Andre_Media conducted a study of 300 runners who ran a HM-marathon combination spaced 6-8 weeks apart, such as the Philadelphia Distance Run and the Philadelphia Marathon, to determine the actual relationship between their HM and marathon performances as compared to the projections of running calculators. The results, which he reported in an excellent essay, are interesting. They clearly indicate that the marathon time projections of running calculators are overly aggressive for most runners….and the commonly used rules of thumb of doubling HM time and adding either 10-minutes or 10% are even worse for those mid-to-back of the packers who run HM’s slower than 1:40. He arrived at a ratio of actual marathon:HM performances that should be a useful guideline for many, if not most, runners. For his report, which I posted on my Running Page with his permission, see HM Pace vs. Marathon Pace.

Interesting, though the real issue with these projections is people not being fully trained for the longer distance.

The projections are usually pretty good for me. Within a minute or two. Must be I am running those HM's too slow!

Very good analysis Jim, I have been struggling with this question and knew that 4:10 based on my 2:00 HM was a bit agressive projection for me. I was basing this on anecdotal resluts from other runners on this site. Now I have facts supporting the hunch. One other race that one can study is the Thunder Road marathon in Charlotte, which has a Dowd Y Half marathon about 6 weeks before it. The first half of the marathon is about 80% same course as the Half, with the start being less than a half mile apart.

If you tell me you are at least in the range of 50-60 miles per week average, then I would say no ... If under 50 average, then its unknown and I would encourage a more conservative approach for the 1st 1/2 and see how you feel.

Interesting, though the real issue with these projections is people not being fully trained for the longer distance.

I think it's this, plus I think first time marathoners underestimate how different these two races are, both from a training and racing standpoint.

VI was basing this on anecdotal resluts from other runners on this site. Now I have facts supporting the hunch.

hehe *snigger* sorry... go back to discussing important running topics.

Unexplained here is how the formulas in the calculators were obtained in the first place. If they were derived by the analysis of running results of a large sample of runners, then the first question to ask is whether the discrepancy across studies can be explained by differences in the samples. I assume, but do not know, that the various calculators are based on empirical studies, not simply made up out of whole cloth. That said, most of the calculators online and elsewhere do state the caveat noted by several posters here; one needs to actually train for the longer distances.

Unexplained here is how the formulas in the calculators were obtained in the first place. If they were derived by the analysis of running results of a large sample of runners, then the first question to ask is whether the discrepancy across studies can be explained by differences in the samples. I assume, but do not know, that the various calculators are based on empirical studies, not simply made up out of whole cloth. That said, most of the calculators online and elsewhere do state the caveat noted by several posters here; one needs to actually train for the longer distances.

I believe that the calculators and VDOT tables were created by sampling enough elite runners and elite masters until a linear pattern appeared, and then projecting downwards onto non-elites. I think the theory is that you can't get statistical significance when sampling non-elites because there is too much variance in training, racing goals, run/walk, and the "crapshoot" factor of the marathon. The side effect of this is that the calculators offer an answer to how you would do if you trained for the marathon equivilently compared to how you trained for the test race. The problem is: the marathon is more training dependent than any other race, and the volume of training required is higher. If a 30-35 mpw runner runs a 20-minute 5k, and then trains for a marathon at 40-50 mpw, they will fall short of what the calculator predicts. This is due to the fact that they were much better trained for the 5k than they were for the marathon, an equivilent amount of training for the marathon for them would be in the 80 mpw range. The cold hard truth is that the vast majority of casual runners do not train adequately for "racing" (as opposed to "running") the marathon, and as such the calculators do not work for them.