Race time calculator
3 watchers
Jul 2013
8:04am, 26 Jul 2013
371 posts
|
Moleshome
Having just put my 5k time into the calculator it predicted a 1/2 marathon time which was identical to one I ran at the beginning of the year - to the second. At first I thought "Wow! How accurate is that?" and then I thought "Hang on, if the best I can expect is the same time that I managed for my first half which came after a month of colds, flu and a chest infection and I'm now doing 25 miles a week training, I may as well not bother" What I think is happening is that the smart Riegel number calculator is taking my existing half marathon PB and skewing the Riegel number to match so it just turns into a circular argument. Perhaps what's needed is to apply an exponential reduction to the effect of PBs based on their age? |
Jul 2013
8:27am, 26 Jul 2013
17,379 posts
|
sheri3004
I don't think it does do that though (base the Riegel number on the existing PB), I think it's just a coincidence that the prediction was the same as your actual PB. My Riegel number has never changed as long as I've been on Fetch. I could be wrong, of course. |
Jul 2013
8:57am, 26 Jul 2013
4,690 posts
|
Wriggling Snake
don't take any notice of it. I ran 21.06 for 5k about 6 weeks ago, this sayd I can run 45.05 for 10k and 1:42 for a HM I have run 44.04 ans 1:39, so it is not that accurate, but I reckon it depend son your training, if you concentrating on fast stuff it will be out for the HM, if you are doing longer training runs, it will beour for the 5k, take it as a rough guide and not gospel. |
|
Jul 2013
9:00am, 26 Jul 2013
377 posts
|
Moleshome
I think it was changed recently based on the work Fetch did on marathon prediction. If you go to Train - Calculators - Riegel Calculator, it sayeth: "This calculator looks at all your PBs, and will attempt to work out your number." |
|
Jul 2013
9:00am, 26 Jul 2013
4,692 posts
|
Wriggling Snake
"attempt" 
|
|
Jul 2013
9:07am, 26 Jul 2013
378 posts
|
Moleshome
No, I understand that WS but it appears to be working on a flawed basis. By not applying a heigher weighting to my more recent 5k time it's just regurgitating my previous half time whereas I'll be gutted if I don't take about ten minutes off that time in September.
|
Jul 2013
9:29am, 26 Jul 2013
3,762 posts
|
paul the builder
Someone else could come along and argue that higher weighting for more recent races isn't what they want, because they were in heavy training, weren't full-on PB attempts, were ill, etc. etc. You've only got a small data set, with only one HM. What does it say it's taking as your Riegel value? Why not plug the 5k time in to a simpler model, like McMillan? It'll give you a HM pace something like 30-40sec/mile slower than 5k pace. Race your HM, put the result in your Fetch log, and then see what you get? (Or alternately, do a dummy version of that now to answer your own question) |
|
Jul 2013
10:01am, 26 Jul 2013
381 posts
|
Moleshome
But surely that's the point Paul, the more recent races are more indicative of your current level of fitness and are therefore a better indicator of expected performance in the near term. It might be small but it's perfectly formed and as my father in law always says "You wouldn't want it behind your ear for a wart". Riegel is 1.13, if I reset it to 1.06 it gives me a far closer time to what I'd anticipate doing. |
|
Jul 2013
10:05am, 26 Jul 2013
3,764 posts
|
paul the builder
And what about if you put in a dummy time for a dummy HM run last weekend, say? Does it work OK then? If you're challenging the algorithm that Fetch uses (it reads like you are), then gather plenty of information on it. Maybe you've spotted a possible improvement. |
|
Jul 2013
10:15am, 26 Jul 2013
382 posts
|
Moleshome
I'd rather not put any dummy runs in, it'll screw up totals that I'm tracking and even if I delete the run it doesn't get removed from other places. I wouldn't go so far as to say I'm challenging it, just raising some points that Fetch might want to ponder when he gets the chance. |
Related Threads
-
July 2021 - your predictions please
Jul 2020
-
Marathon Prediction White Paper
Jun 2019
-
Marathon Predictions from Half Marathon times
Mar 2017
-
Marathon time predictors... are they crap or am I
Mar 2016
-
Riegel Value for predictions on training log and race times calculator
Apr 2015
-
Scilly 60
Apr 2024
-
The virtual mile thread
Apr 2024
-
Do we have a race report thread?
Apr 2024
-
Runners trots
Mar 2024
-
What’s your “A” Race for 2024
Jan 2024
