Automatic Intervals -- How are they chosen? Is there any link to the algo?


I’m talking about these at the top of the ride. How are they selected and why?

1 Like

No one?

No one knows cos no one has the details of what you’re asking.

If you’re asking

  1. how are the intervals selected. As in, what’s the algo that indicates that “this is an interval”
  • ¯_(ツ)_/¯
  1. for each of these intervals. The fields. Can I change them.
  • this one I know the answer. There’s a bunch of options at the bottom of that screen. The one along the “map/chart” line. You can select what parameters you want.

Thanks. Im asking how the automatic intervals are chosen, as the post clearly states. If your response to this question is: " * ¯_(ツ)_/¯"

Does that mean youre just replying you dont know? No one knows? What are you communicating?

  1. I don’t know hence ¯_(ツ)_/¯
  2. as most who frequents this / read your post and you have received no answer hence most likely no one else knows hence ¯_(ツ)_/¯
1 Like

Okay. If anyone else knows the criteria for how these automatic intervals are selected, please let me know.

FWIW and apologies to David, is that Golden Cheetah does a better job at auto-detecting intervals. IIRC, David has said that updating is overdue.

Yes and you know, it’s not just that I want to know where the intervals were; I want to know what were the critical intervals. I want to know what interval did I do that was preceded or succeeded by some drop in HR efficiency.

As in, what interval caused my body to create aerobic decoupling. That would be a great question to answer.

Routinely ride longer? When doing a lot of 2-3 hour endurance and tempo rides, I’ve seen little to no decoupling out to 3+ hours. And little decoupling out to an hour at FTP pace. It’s possibly related to the amount of aerobic fibers (type 1 and type 2a). This is a guess, but from some classic research I think you might improve by doing more long threshold work. And longer rides. The Friel recommendation was something like “2x time at endurance for a 1x event.” Worked for me.

Good comment and thank you for the dialog. My assumption is that:

aerobic decoupling = bad.

Aerobic decoupling is caused disproportionately by certain efforts

We should uncover those efforts and train them.

And even if what you’re saying is true which I’m not doubting, uncovering intervals that cause aerobic decoupling and measuring their impact over time would contribute to improved performance.

What do you think?

I don’t think about aerobic decoupling much, because my events are 1-2 hours and after a season break, 2-3 months of quality base training will generally reduce my decoupling to below 3% over 1-3 hours all the way from endurance to threshold.

However going back to the basics, this metric was popularized by Joe Friel in the mid 2000s. This is the first article I remember reading:

http://www.trainingbible.com/joesblog/2009/11/aerobic-base-ride.html

To paraphrase and using the chart of the 2nd athlete in that post - if you have a 2 hour event and your endurance/zone2 power (relative to HR) starts quickly dropping after 1 hour - that is a bad signal and you will likely fade in a 2 hour race.

Which then raises training questions:

  • what is the right intensity to target the aerobic threshold?
  • what is the right duration?

Friel provides his point-of-view in this blog:

Which takes the position the duration for endurance rides is 2-4 hours. 4 hour race? Then do 4 hour steady endurance ride at or around your aerobic threshold. Crit racing? 2 hour endurance rides. Half-marathon? 2.5 hour endurance rides.

Do enough of those rides during base training until your decoupling drops below 5%. At that point your basic aerobic abilities are strong enough to move on to the next phase of training (late base or build).

Depending on time, slowly increasing your endurance riding (plus HIIT work) will help you build a bigger aerobic engine.

If you have a strong aerobic system, during the next phase of training you can look at decoupling on tempo and threshold rides. For example as explained in this blog post:

Uncovering efforts is actually very simple, when you look at it that way.

For myself, during late base / early build, about once a month I look at decoupling on long intervals (60-120 minutes) at 80-95% ftp.

1 Like

Yeah i mean, like i’ve said, Im not trying to just train aerobic decoupling. IM trying to see what effort specifically impacts it and creates aerobic decoupling.

https://www.sciencedirect.com/science/article/pii/S0024320521010961

(post deleted by author)

Everything posted by WindWarrior is relevant to your question. It’s explaining what aerobic decoupling is.

I.t.o. the algorithm, this is something David could answer, or not. Intervals started out by detecting intervals on a ride. I will try find the initial posts from 2018 where it’s described.

1 Like

This is where David introduced the Intervals project to the world, before there was a forum linked to the site. The detection might have changed, but originally it was based on having a short recovery period after time at work. I guess the work level needs a definitive value or % of threshold this be determined an interval.