or Connect
Cycling Forums › Forums › Bikes › Power Training › Parsing Time (& TSS)
New Posts  All Forums:Forum Nav:

Parsing Time (& TSS)

post #1 of 186
Thread Starter 
I am currently a beta tester for a new application designed to parse power meter files into segments based on descriptive training schema (i.e. identifies sections of a ride which the meet minimum intensity and duration criteria to be considered an L4 effort, L5 etc). These segments can then be totalled for time and TSS at each level.

Any schema can be used such as Andy Coggan's levels based on FTP, or Ric Stern's levels based on MAP. As we know they are fundamentally similar schema in terms of the intensity and minimum duration applicable, just some variances in the number of levels and "bandwidth" of each level.

I thought I would share some charts developed as a result of the output of this data. This is by no means the only way of looking at the parsed workout data but I thought it may provide some insights which perhaps didn't previously exist in a visual format.

There will no doubt be debate over the validity of the file parsing approach so certainly interested in what forums members make of this lot....

The setting of levels is arbitrary (you define lower limits for duration and intensity), so there is some interpretation as to what the limits are. This approach also tends to make you focus on levels as being discrete in their physiological impact, when in fact we all know this isn't the case, indeed this in itself is probably the biggest challenge to overcome when translating our "analogue" bodies' response into a "digital" display.

Nevertheless, this provides one way of visualising the mix of workout types from all rides, in particular non-structured rides.

The charts cover my own training since 1 August. All road and track riding, nothing on trainer/rollers (except for the odd short warm up before a race).

Chart 1 shows
weekly ride time in minutes parsed into "levels" based on Ric's MAP based levels.

Chart 2 shows weekly ride TSS parsed based on same "levels" and plotted with CTL attained at end of week.

Thoughts?

Alex

LL
LL
post #2 of 186

Re: Parsing Time (& TSS)

Alex - those graphs are great.

I am testing too - would love to get my hands on a templat to produce those graphs from my data - can you send me anything ? (I am very poor on creating graphical output)

Thanks
post #3 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by peterwright
Alex - those graphs are great.

I am testing too - would love to get my hands on a templat to produce those graphs from my data - can you send me anything ? (I am very poor on creating graphical output)

Thanks
Alex is quite the chartmaster, isn't he?
post #4 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by RapDaddyo
Alex is quite the chartmaster, isn't he?
Isn't he just


I think he is just showing off..
post #5 of 186

Re: Parsing Time (& TSS)

I've got a question about this approach.

Let's say I do a ride L3 for x minutes, L4 for y minutes, L3 for x minutes.

where x < min L3 duration, y > min L4 duration, x + y + x > L3 minimum duration. What are the results going to be?

In other words, since L4 is greater intensity than L3, does it count towards L3, and should it?

So, would I get:
L3 = x + y + x
L4 = y

Looking at Coggans table of physiological adaptions, things get a bit complicated for L4/L5/L6, however, the rapid decline in time spent at these levels probably doesn't introduce much of an error.
post #6 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
I've got a question about this approach.

Let's say I do a ride L3 for x minutes, L4 for y minutes, L3 for x minutes.

where x < min L3 duration, y > min L4 duration, x + y + x > L3 minimum duration. What are the results going to be?

In other words, since L4 is greater intensity than L3, does it count towards L3, and should it?

So, would I get:
L3 = x + y + x
L4 = y
I am very glad you put the "in other words" part in...

It filters the data so that all data runs through a check to see if it fulfills your preset criteria for the various zones - it starts with the highest intensity and works down so that no data segment can be a member of more than one zone. In other words .. L4 does not count towards L3 etc.

I am sure Rap will explain this better (it is his app after all)
post #7 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by peterwright
I am very glad you put the "in other words" part in...

It filters the data so that all data runs through a check to see if it fulfills your preset criteria for the various zones - it starts with the highest intensity and works down so that no data segment can be a member of more than one zone. In other words .. L4 does not count towards L3 etc.

I am sure Rap will explain this better (it is his app after all)
So, what you are saying is that the results of my example would be
L4 = y
L3 = 0
post #8 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
So, what you are saying is that the results of my example would be
L4 = y
L3 = 0
One way to solve this would be to add a
"Nested levels" switch that can be turned on and off.

Personally, I'd turned it to on, meaning that I prefer total time per level to be higher than the total workout time.
post #9 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
I've got a question about this approach.

Let's say I do a ride L3 for x minutes, L4 for y minutes, L3 for x minutes.

where x < min L3 duration, y > min L4 duration, x + y + x > L3 minimum duration. What are the results going to be?

In other words, since L4 is greater intensity than L3, does it count towards L3, and should it?

So, would I get:
L3 = x + y + x
L4 = y

Looking at Coggans table of physiological adaptions, things get a bit complicated for L4/L5/L6, however, the rapid decline in time spent at these levels probably doesn't introduce much of an error.
It's a function of how you define the levels (zones). Each level (zone) has a minimum power criterion and can have a minimum duration criterion, plus a basis (AP or NP). Segments are identified that meet all of the criteria (power and/or duration) based on the basis (AP or NP). It gets a little complicated when there are "stacked" segments that meet the criteria. For now, let's just say that it's a zero sum game (i.e., no part of the ride file can be a member of more than one segment). The app sorts out these conflicts.
post #10 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
So, what you are saying is that the results of my example would be
L4 = y
L3 = 0
No, the results would be
L4 = y
L3 = 2x

This is what I was referring to as "nested segments," in which there is one long L3 segment (x + y + x) with a nested L4 segment (y) in the middle. The L4 segment would be "lifted out" and the remainder of the long L3 segment (x+x) is what's left.
post #11 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by RapDaddyo
No, the results would be
L4 = y
L3 = 2x

This is what I was referring to as "nested segments," in which there is one long L3 segment (x + y + x) with a nested L4 segment (y) in the middle. The L4 segment would be "lifted out" and the remainder of the long L3 segment (x+x) is what's left.
Okay, so the L4 segment contributes nothing towards the overall L3.
post #12 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
Okay, so the L4 segment contributes nothing towards the overall L3.
Correct. It's as though the L4 segment was sliced out of the ride file and you're left with the first and second parts of the L3 segment. However, if the L4 segment did not meet the minimum duration requirement for an L4 segment, then the whole thing would be one long L3 segment (x + y + x).

Technically, the app gives both results (i.e., "gross" and "net").
post #13 of 186

Re: Parsing Time (& TSS)

A further conclusion therefore, is a segment at a certain intensity level that does not meet the duration criterion, is demoted to a lower level.

To illustrate:

I ride at L3 for x, and L4 for y, where x < L3minduration and y < L4minduration but x + y > L3minduration I would get:

L4 = 0
L3 = x + y

But, should y > L4minduration I would get:

L4 = y
L3 = 0

Sorry, I'm a programmer at heart and I love testing algorithms!
post #14 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by RapDaddyo
Technically, the app gives both results (i.e., "gross" and "net").
What do you mean by the "gross" and the "net"?
post #15 of 186

Re: Parsing Time (& TSS)

Quote:
Originally Posted by Bruce Diesel
A further conclusion therefore, is a segment at a certain intensity level that does not meet the duration criterion, is demoted to a lower level.

To illustrate:

I ride at L3 for x, and L4 for y, where x < L3minduration and y < L4minduration but x + y > L3minduration I would get:

L4 = 0
L3 = x + y

But, should y > L4minduration I would get:

L4 = y
L3 = 0

Sorry, I'm a programmer at heart and I love testing algorithms!
Stop being a propeller head and testing algorithms and start training ..
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Power Training
Cycling Forums › Forums › Bikes › Power Training › Parsing Time (& TSS)