Page 1 of 1

Understanding marathon shape better

Posted: Fri 14. Jun 2019, 16:26
Hi

I've tried to understand marathon shape, but it is baffling me! I know that the current figure is a function of volume of km (miles) run and long runs (long jogs) done, as well as VO2Max, but I just don't "get" what I would (in theory) need to be doing to improve the figure.

My effective VO2max is roughly 50, so I understand that that gives me approx 85 km as the volume and 30 km as the long runs figure. But over what period? Do I get a score of 100% if over a certain period of time my effective VO2Max remains 50, and I achieve at least 85km per week and have a weekly long run of at least 30km per week? Is that one weekly run of 30km, or is that a series of long runs which accumulate to at least 30km per week?

The (?) next to the marathon shape figure on the main page suggests that there is a 6 month window over which this is evaluated, but I've read elsewhere that there is a 70 day window.

I know achieving the highest marathon shape score isn't an objective in itself, but if I understand the parameters better I can evaluate for myself what the graph is showing me - at the moment, I have no clue!!!

Thanks!

Re: Understanding marathon shape better

Posted: Fri 14. Jun 2019, 16:33
It's configureable: Have a look at Configuration -> General Settings -> Marathon Shape.
https://runalyze.com/glossary/marathon-shape

Re: Understanding marathon shape better

Posted: Fri 14. Jun 2019, 16:53
THanks - I can see what I can change, but the descriptions against the settings don't really tell me what the various parameters mean / do.

So for my own clarity:-
"Minimal distance for longjogs" - this is the shortest distance run that can count towards being a long run for the purposes of the marathon shape calculation
"Days for longjogs" - this is the period over which long runs are accumulated. Are they then averaged per day or per week?
"Days for week kilometer" - this is the period over which the total volume is accumulated. Again, are these then averaged over a day or week?
"Days for week kilometer (minimum)" - no idea what this is supposed to do!
"Factor for week kilometers" - is this in relation to the split of the weighting between volume and long runs? So if I change this to 0.5, then volume and long runs are equally accounted for in the calculation?

So to get a perfect score for long runs, I'd need to do 1 or more runs of at least 13km accumulating to 30km (based on my VO2Max) over a 7 day period to max out the weekly score, and do that over a 70 day period to max out the score overall - and this would comprise 33% of the total score.

And to get a perfect score for volume, I'd need to do at least 85km in a week to max out the weekly score, and do that consistently over 182 days to max out that score overall - and this would comprise 67% of the total score.

Re: Understanding marathon shape better

Posted: Fri 14. Jun 2019, 22:34
larkim wrote:
Fri 14. Jun 2019, 16:53
"Minimal distance for longjogs" - this is the shortest distance run that can count towards being a long run for the purposes of the marathon shape calculation

Yes. And this setting does highly influence on how long runs longer that the target (of currently 30 km) scale.

larkim wrote:
Fri 14. Jun 2019, 16:53
"Days for longjogs" - this is the period over which long runs are accumulated. Are they then averaged per day or per week?

Each long run gets its own rating. For a target of 30 km, a 30 km gets exactly 1.0 points - weighted on it's date (i.e. 2.0 points if it's today and 0.0 points if its just out of the specified period). There is no averaging or anything like that. Those points are just accumulated.

The score for a 25k run (or a 35k run) depends on the minimal distance from above. For a minimal distance of 13k, a 13k run gives 0.0 points and the distance range from 13k to 30k scales exponentially to 1.0. Setting the minimal distance up to e.g. 20k increases the 'gradient' and causes a 35k run to get a much higher score than with a minimal distance of 13k.
larkim wrote:
Fri 14. Jun 2019, 16:53
"Days for week kilometer" - this is the period over which the total volume is accumulated. Again, are these then averaged over a day or week?

What kind of averaging are you talking about? The total sum (of e.g.182 days) is compared to its target (in your case: 84k/week --> expected sum of182 * 84 / 7).
larkim wrote:
Fri 14. Jun 2019, 16:53
"Days for week kilometer (minimum)" - no idea what this is supposed to do!

It's only relevant if there's no historical data.

larkim wrote:
Fri 14. Jun 2019, 16:53
"Factor for week kilometers" - is this in relation to the split of the weighting between volume and long runs? So if I change this to 0.5, then volume and long runs are equally accounted for in the calculation?

Correct.
larkim wrote:
Fri 14. Jun 2019, 16:53
So to get a perfect score for long runs, I'd need to do 1 or more runs of at least 13km accumulating to 30km (based on my VO2Max) over a 7 day period to max out the weekly score, and do that over a 70 day period to max out the score overall - and this would comprise 33% of the total score.

As mentioned above, every single run get's its own score and those points should sum up to 10.0 (for a 70 day period).

The third icon in the 'Calculations' panel on the right (magic wand icon) links to an information popup that shows some of those details.

Re: Understanding marathon shape better

Posted: Fri 14. Jun 2019, 22:40
That's great info, many thanks!!

Matt

Re: Understanding marathon shape better

Posted: Wed 15. Apr 2020, 08:05
Each long run gets its own rating. For a target of 30 km, a 30 km gets exactly 1.0 points - weighted on it's date (i.e. 2.0 points if it's today and 0.0 points if its just out of the specified period). There is no averaging or anything like that. Those points are just accumulated.

The score for a 25k run (or a 35k run) depends on the minimal distance from above. For a minimal distance of 13k, a 13k run gives 0.0 points and the distance range from 13k to 30k scales exponentially to 1.0. Setting the minimal distance up to e.g. 20k increases the 'gradient' and causes a 35k run to get a much higher score than with a minimal distance of 13k.
I managed to "reconcile" to the target KM per week. I just ran an average over the last 26 weeks and was very close.
I think I had the wrong expectation of the long run piece. My minimum is 13km and I ran a 13km yesterday thinking that would start to bump up that side of the equation. But no.
My window shows - Target long run (10 weeks) 29 done by 0% (0 points) x0.33.
My run yesterday was 13.5. Am I right in saying that this is then calculated as (13.5- 13.0) / (29 - 13), so 0.5/16 = .03125 ?? And any score that is less than 0.2 is ignored. If I had run 16.2 km, then that would be 3.2 / 16 =0.2 ?

Also did not realise that the target values were based on the current evaluated VO2max. Does that mean that the targets will move over time?

Re: Understanding marathon shape better

Posted: Fri 17. Apr 2020, 07:19
keithsherman wrote:
Wed 15. Apr 2020, 08:05
My minimum is 13km and I ran a 13km yesterday thinking that would start to bump up that side of the equation. But no.
My window shows - Target long run (10 weeks) 29 done by 0% (0 points) x0.33.
My run yesterday was 13.5. Am I right in saying that this is then calculated as (13.5- 13.0) / (29 - 13), so 0.5/16 = .03125 ?? And any score that is less than 0.2 is ignored. If I had run 16.2 km, then that would be 3.2 / 16 =0.2 ?

Also did not realise that the target values were based on the current evaluated VO2max. Does that mean that the targets will move over time?

In your calculations you miss one point of my statement:
... the distance range from 13k to 30k scales exponentially to 1.0 (the exact term would be quadratic):
((13.5 - 13.0) / (29 - 13))^2 = (.03125) ^2 = .00097...
((16.2 - 13.0) / (29 - 13))^2 = (.02) ^2 = .04

And keep the following in mind:
weighted on it's date (i.e. 2.0 points if it's today and 0.0 points if its just out of the specified period)

And yes, the targets will move over time. The targets of a 2:30h runner are different than for a 4h runner.

I can give you a short preview of our new 'Marathon Shape View' that will be release as new beta feature soon:
There you can see the points for each run (by hovering it, the bar length depends on the distance only)
2020-04-17_Preview-Beta-Marathon-Shape.png (37.54 KiB) Viewed 6628 times

Re: Understanding marathon shape better

Posted: Mon 20. Apr 2020, 05:15
Thanks for the lesson.
This is a great piece of work, and it has motivated me to keep striving to push the number higher. Just need to get stronger to make some progress.