Interesting article on the (in)accuracy of race time predictors (Read 523 times)

scappodaqui


rather be sprinting

    Every runner has a sweet spot that is a combination of temperament and physiology.

     

    My best distance at my physical peak in my early 20s was 1000-1500m. I was substantially worse from 800m down, then my performance equivalents gradually tapered off as I went up in distance.

     

    I think that if I had kept training through my late 20s, my best distance would have been 5000m.

     

    By the time I returned to training in my 30s, my best distance had become the half marathon.

     

    I think these changes had more to do with my ability to concentrate and changes in my own personality and temperament than with physiological changes. I simply am not willing or able to go to the places I went as a young 1500m runner. But on the other hand, I am much more methodical and patient and tougher over the long haul than I was then.

     

    I wonder what Alan Webb would have to say about this sort of discussion.  He's finally trying distances other than the mile, I believe.  It  makes sense--the older you are the more your fast-twitch turns to slow-twitch, tendon elasticity decreases, etc.  That's why age is always going to be a factor in distance/performance calculators, too!  My friends who are over 30 can whup me over 5k, but under I am faster--because they have the aerobic buildup and I haven't run many lifetime miles.

     

    So I think it is partly mental but also very physical.

    PRs: 5k 19:25, mile 5:38, HM 1:30:56

    Lifting PRs: back squat 176 lb

      Eric's calculator right here defaults to 1.08, but you can plug in whatever coefficient you want.  Make the method fit the data, as did the author of the article.

       

      I'm tossing in my 2 cents because I'm taking a few minutes break from a project that's taking much longer than expected.  As many have pointed out, there are many factors that contribute to your race time.  For the sake of discussion, we'll assume the predictor only applies in ideal environmental and physical conditions.  Still, there is no one equation that can accurately and consistently predict your race time, never mind one that applies to everyone.

       

      While Fetch had a good idea to validate the Riegel formula against his data, his approach is wrong.  The problem lies in the distribution of the data.  In order for the 1.15 factor to be correct, the half and full marathon times must be narrowly clustered.  This can't be the case because the half marathon times he looked at probably ranged from 1 hour and change to all the way up to 4 hours or more.  The mean pace, or even the median pace is meaningless because the standard deviation is so large.

       

      Everyone slows down as the distance increases.  The slow down, or the degradation, is different for everyone.  Instead of trying to derive a degradation coefficient that applies to everyone, it is simpler and more accurate to derive the coefficient for each person.  RA's version uses a 1.08 factor, which I think might be a little low.  If you provide two recent races of differing distances, then it can calculate how much you slow down as the distance increases.  That should provide a better predictor than any fixed degradation coefficient.

      zonykel


        There is also a range of expected performance, as opposed to a precise time. We're not machines.

          There is also a range of expected performance, as opposed to a precise time. We're not machines.

           

          I could point to a few selected RA training logs that would indicate otherwise....

          Dave

          GC100k


             

            The problem lies in the distribution of the data.  In order for the 1.15 factor to be correct, the half and full marathon times must be narrowly clustered.  This can't be the case because the half marathon times he looked at probably ranged from 1 hour and change to all the way up to 4 hours or more.  The mean pace, or even the median pace is meaningless because the standard deviation is so large.

            I inferred that he compared half and full times for individuals, not medians or means:

             

            " I found a core of a thousand runners who (you'd hope) would have recorded a representative time at both distances. They range in ability from five-minute milers all the way down to twelve-minute milers."

             

            "I fed all of my 1071 runners through that formula, and found that only 49 of them managed to hold on to the tails of 1.06 - it was far more common to see a score of 1.15."

             

            So he did calculate an exponent for each person.  he goes on to discuss the difference between men and women.

              What time span are we talking about to make the coefficient realistic?

              Basically how close do the races have to be to come up with a valid factor such as 1.08 or 1.06 or 1.15?

                All predictors say the time is based on your optimal training. What is the optimal training? 70mpw? For how long? 18 weeks?

                 

                I am relying on the predictor for my MP. I have never run 26.2 miles. Many people describe their dreadful feeling in the last 6 miles. It is a bit scary for me.

                5k - 20:56 (09/12), 7k - 28:40 (11/12), 10k trial - 43:08  (03/13), 42:05 (05/13), FM - 3:09:28 (05/13), HM - 1:28:20 (05/14), Failed 10K trial - 6:10/mi for 4mi (08/14), FM - 3:03 (09/14)

                  Many people describe their dreadful feeling in the last 6 miles. It is a bit scary for me.

                   

                  It really doesn't have to be like that.  It is really more a factor of proper training, pacing and nutrition.  I've only run two fulls and am by no means fast, but if you look at my splits from Philly this past fall you can see the last 6 miles weren't dreadful at all.

                  https://lh3.googleusercontent.com/-Cjxh0Onuk-M/UXbrCmgK1aI/AAAAAAAABDw/zDdYfmd2E0k/s682/philly+splits.PNG

                    All predictors say the time is based on your optimal training. What is the optimal training? 70mpw? For how long? 18 weeks?

                     

                    I think using a shorter distance PR to predict a marathon time is sketchy business unless you've got a couple/few years of 2000 miles or more per year.

                    Runners run.

                       I think using a shorter distance PR to predict a marathon time is sketchy business unless you've got a couple/few years of 2000 miles or more per year.

                       

                      I think so too.  I'm going to be a test subject in less than two weeks (although I won't have multiple 2000 mile years).

                      Well at least someone here is making relevance to the subject.

                        I think using a shorter distance PR to predict a marathon time is sketchy business unless you've got a couple/few years of 2000 miles or more per year.

                         

                        And even then, it's still sketchy business, at least for some folks.

                        - Joe

                        We are fragile creatures on collision with our judgment day.

                          It really doesn't have to be like that.  It is really more a factor of proper training, pacing and nutrition.  I've only run two fulls and am by no means fast, but if you look at my splits from Philly this past fall you can see the last 6 miles weren't dreadful at all.

                           

                          I dont' know.... I think it does.  I think it has to be like that if you're going to run it close to your capability.  I've run >26.2 in training and been tired but generally OK = not dreadful, but in a race whenever I've taken it anywhere close to "the line" those final miles are always, indeed, at least moderately dreadful.

                          - Joe

                          We are fragile creatures on collision with our judgment day.


                          I'm back!

                            I am relying on the predictor for my MP. I have never run 26.2 miles. Many people describe their dreadful feeling in the last 6 miles. It is a bit scary for me.

                             

                             I dont' know.... I think it does.  I think it has to be like that if you're going to run it close to your capability.  I've run >26.2 in training and been tired but generally OK = not dreadful, but in a race whenever I've taken it anywhere close to "the line" those final miles are always, indeed, at least moderately dreadful.

                             

                            Well, I think the pain of running a marathon at max capability is not the same issue as  "many people describe their dreadful feeling in the last 6 miles" -- there, I would agree with FatSweatyBullDog; it doesn't have to be like that, with proper training & pacing.

                             

                            Running any race at "the line" is going to hurt -- if it didn't, you wouldn't be at the line. But the hurt is different in different kinds of races. I much prefer the marathon kind of hurt to the 10K kind of hurt. Paced well, it doesn't really get bad until 22-23 for me, and even then, it's different. But your first marathon is probably not going to be at this line anyway. 

                               Well, I think the pain of running a marathon at max capability is not the same issue as  "many people describe their dreadful feeling in the last 6 miles" -- there, I would agree with FatSweatyBullDog; it doesn't have to be like that, with proper training & conservative pacing.

                               

                              Inserted a word for you, since I know you didn't mean for the adjective "proper" to apply to "pacing".  Cool

                              - Joe

                              We are fragile creatures on collision with our judgment day.

                              DoppleBock


                                Agreed - I find that I never would look at one race ... even more than that, I think after years of training and running marathons, your speed work the last  recent block of training is a really good indicator.

                                 

                                There are times I run poor races, but know what level my speed work has been at, so I am confident that I can achieve a significantly better time than the race predictor.

                                 

                                 

                                 

                                I think using a shorter distance PR to predict a marathon time is sketchy business unless you've got a couple/few years of 2000 miles or more per year.

                                http://a-big-horse.blogspot.com/ 

                                2013 Goals ~ Mar < 3:00, 5M < 29, 10k < 35