Marathon Prediction White Paper

14 watchers
Jun 2019
3:14pm, 3 Jun 2019
21,974 posts
  •  
  • 0
fetcheveryone
Those of you with funny data - I'd forgotten to exclude planned sessions from the totals - so you were getting lots of extra mileage for no extra time. That should be fixed now.
Jun 2019
3:18pm, 3 Jun 2019
41,835 posts
  •  
  • 0
Velociraptor
That wouldn't have applied to me because I don't put anything in my log till I've done it.
Jun 2019
3:20pm, 3 Jun 2019
41,836 posts
  •  
  • 0
Velociraptor
(It possibly just indicates that I run a lot of miles relative to the level at which I perform.)
Jun 2019
3:20pm, 3 Jun 2019
21,975 posts
  •  
  • 0
fetcheveryone
I mean the folks who were getting a training pace of 3 mins per mile :-)
Jun 2019
3:21pm, 3 Jun 2019
3,111 posts
  •  
  • 0
StuH
Ah, that's better, still looks like I'm not trying hard enough!
Predictions: 3:39 to 3:34
Outcomes: 4:15 to 3:46
Jun 2019
3:23pm, 3 Jun 2019
7,720 posts
  •  
  • 0
larkim
I'm no statistician, but early on in his paper he dismisses # training days per week and Max distance per week as having no correlation with marathon outcomes. Yet, the correlation coeffieicnt r for both of these are above 0.7, which I'd always understood to be the starting point for some form of strong correlation. Admittedly, not as strong as the 0.81 and 0.85 for the total distance and average pace, but surely those other characteristics ought to have been brought into play?

Anyway, it's a useful calculator for Fetch to have included - another data point to bounce off!
jda
Jun 2019
3:30pm, 3 Jun 2019
4,573 posts
  •  
  • 0
jda
I can't imagine anyone finding this sort of predictor useful if they have a recent half marathon to go off.

I've predicted my race time to about 2 mins for most of the last 5 (and only once missed by 5 mins). We all know roughly how a half converts, right (especially once you've done it once or twice)?
Jun 2019
3:40pm, 3 Jun 2019
21,976 posts
  •  
  • 0
fetcheveryone
I've added a calculator to the page now, so you can try adjusting mileage and pace to see what differences the Tanda calculator produces.

fetcheveryone.com/genovastudy.php
Jun 2019
3:46pm, 3 Jun 2019
21,977 posts
  •  
  • 0
fetcheveryone
I've also moved the calculator to this link:

fetcheveryone.com/training-calculators-tanda.php

and included a link to it under Training > Calculators.
Jun 2019
3:50pm, 3 Jun 2019
2,291 posts
  •  
  • 0
B Rubble
Prediction 3:00:43 to 3:14:42. Outcome 2:59:00 to 3:26:26 (removing one "outlier"). Seems fairly good, but I agree with jda on the conversion from a recent 1/2 marathon.

About This Thread

Maintained by fetcheveryone
I was pointed at the following publication from Giovanni Tanda at the University of Genova:

https...

Related Threads

  • events
  • marathon
  • prediction









Back To Top
X

Free training & racing tools for runners, cyclists, swimmers & walkers.

Fetcheveryone lets you analyse your training, find races, plot routes, chat in our forum, get advice, play games - and more! Nothing is behind a paywall, and it'll stay that way thanks to our awesome community!
Get Started
Click here to join 112,115 Fetchies!
Already a Fetchie? Sign in here