The Sub 3:15 Marathon Thread
334 watchers
Jul 2017
11:56am, 27 Jul 2017
26,990 posts
|
HappyG(rrr)
This is the study overview spectrum.ieee.org but I can't find a link in it to the study itself. I'll keep looking. This overview is probably techie enough. Anything with a formula like: "The Salzburg formula for the average Overestimation of Distance (OED) is then, OED = (d02 + Vargps - C)1/2 - d0 The variance is always positive, so if the autocorrelation is lower than the variance, the overestimation of distance will always be positive. And the autocorrelation is generally lower than the variance." is good enough explanation for me!  G
|
|
Jul 2017
11:57am, 27 Jul 2017
26,991 posts
|
HappyG(rrr)
It was in the second paragraph... tandfonline.com |
Jul 2017
12:28pm, 27 Jul 2017
1,738 posts
|
jdarun
I've always set up workouts on the watch mostly for timing purposes (e.g. so I get a beep on 6 mins rather than having to keep checking for it!) which also gives me easy analysis of how I did after the run. Doing intervals by time rather than distance wouldn't work well with |
Jul 2017
12:54pm, 27 Jul 2017
2,434 posts
|
larkim
Wow, that's good science stuff, thanks! My only caveat having skimmed the doc is that there must be edge case scenarios (e.g. plotting around a circle) where the overestimation doesn't apply. For example, imagine a 100m diamter circle being run around with a GPS device. If only 5 measurements are taken, at the start, 1/4 round, 1/2 round, 3/4 round and finish the GPS trace will have created a "square" for which each recorded point will either be within or outwith the circle circumference. But even if all are outside of the circle by a margin of error, the perimeter of that square will be shorter than the circumference of the circle (up to a point). It's that logic which tends to make me consider tracks likely to measure short as the sampling interval of a watch may only take 5 or 6 measurements as the runner goes around each bend. If I only understood the maths in their paper well enough I'd have a proper read!! |
Jul 2017
12:54pm, 27 Jul 2017
6,171 posts
|
paul the builder
larkim - you assume that GPS inaccuracy only occurs in the longitudinal direction (by which I mean - the direction you're moving). It can also place you L and R of true position, so creating a false 'zig-zag'. But the effect is pretty small if you stay on a straight path, true. G - you commit one the Fetch cardinal sins, for me. Not putting separate training entries in your log for each segment of the training. If I do 8 with 4 tempo, then you'll see 2 WU, 4 tempo, 2 WD. It's then a piece of piss for me to compare tempo efforts across any given spread of time, and average HR for them too, all on one page. |
|
Jul 2017
12:56pm, 27 Jul 2017
6,172 posts
|
paul the builder
x-post; I was referring to your earlier discussion about 1603m and all that.
|
|
Jul 2017
1:01pm, 27 Jul 2017
2,435 posts
|
larkim
True, ptb. As I said, miles have got "longer" for me in the period that I've been using GPS devices as their smoothing algorithms get better, but I do trust a GPS to more closely / accurately measure a mile in a straight line than I do in a circular route as outliers have less of an impact and the GPS watch does a better job of smoothing out some of these inaccuracies.
|
|
Jul 2017
1:09pm, 27 Jul 2017
2,436 posts
|
larkim
Ref not splitting runs down - I'm equally guilty of that, I very rarely can be bothered to stop start the watch at the changeover times (though I did do that last weekend for the MP run). Would be nice if there was a facility on the site to mark parts of runs as WU / CD etc as I prefer the neatness of 1 run per day, unless there is a decent period of time between each aspect (e.g. when I run a mile to parkrun and then hang around for 5 minutes, the WU gets logged separately).
|
|
Jul 2017
1:26pm, 27 Jul 2017
6,173 posts
|
paul the builder
But it's so easy to press stop, save the run, and then press start again. Sometimes I don't even actually stop running to do it (although I typically do, and have a bit of a stretch before inflicting any quicker running on these short old muscles).
|
|
Jul 2017
1:32pm, 27 Jul 2017
2,437 posts
|
larkim
That's exactly what I did last weekend (not stop running). But it makes the log so messy
|
Related Threads
-
RW sub3/3.15
Jan 2025
-
Sub 3Hr Marathon
Jan 2025
-
The sub 3.30 marathon thread
Jan 2025
-
Sub-4hr marathon support and celebration thread
Dec 2024
-
Sub 2:30 Marathon
Oct 2024
-
The sub 2.45 marathon thread
May 2024
-
The sub 3.45 Marathon Thread
Oct 2023
-
The sub-5 hour marathon thread
Apr 2022
-
5:43 Marathon to Sub 5hrs
Dec 2018
-
Sub 3.05 Marathon Plan / Advise - Help please
Aug 2016