Matches Revisited

RapDaddyo said:

Thanks to frenchy and wattsup? and peterwright, I have been trying to get my head around matches. My initial findings are that such analysis could be of significant value in post-ride analysis and, eventually, real-time on-bike decision making (assuming availability of relevant data). Granted, we're talking about anaerobic efforts. But, I have three specific questions about these efforts from a physiological perspective. (1) Above what threshold is one drawing on his anaerobic capacity? Is this what is meant by anaerobic threshold (AT), often equated with FT? (2) What is the relationship between power and utilization of one's anaerobic capacity? Is this relationship linear or non-linear? (3) Does one's anaerobic capacity partially or fully restore itself after some period of time at intensities below the threshold above? Any insights, especially supported by data or physiology, would be greatly appreciated. Thanks in advance.

It sounds like the first two questions can be answered by application of the Monod critical power pardigm http://www.velo-fit.com/articles/critical-power.pdf . In that model, below critical power (around FT), power is generated aerobically and this or lower power can be maintained for a very long time. Above critical power, all additional power comes from anaerobic metabolism and the total anaerobic work done is constrained by anaerobic work capacity. Since it is the total anaerobic work (P x t) that is constrained, it doesn't matter whether that is used up by being slightly over critical power for a long time or well over for a shorter time. As to the third question, experience tells me that anaerobic work content is restored at lower intensities (that's what let's me do a VO2max interval, recover, and then do some more) and while I believe the rate at which anaerobic work capacity is restored is related to aerobic fitness, I don't know the exact relationship.

asgelle said:

It sounds like the first two questions can be answered by application of the Monod critical power pardigm http://www.velo-fit.com/articles/critical-power.pdf . In that model, below critical power (around FT), power is generated aerobically and this or lower power can be maintained for a very long time. Above critical power, all additional power comes from anaerobic metabolism and the total anaerobic work done is constrained by anaerobic work capacity. Since it is the total anaerobic work (P x t) that is constrained, it doesn't matter whether that is used up by being slightly over critical power for a long time or well over for a shorter time. As to the third question, experience tells me that anaerobic work content is restored at lower intensities (that's what let's me do a VO2max interval, recover, and then do some more) and while I believe the rate at which anaerobic work capacity is restored is related to aerobic fitness, I don't know the exact relationship.

Good stuff. Like you said, different rates of recovery determined by aerobic fitness.

Also, one who is especially good at anaerobic efforts because of training and/or genetics will respond to anerobic efforts differently than someone who is not good at those efforts. For example, ask me to do 10 seconds at 3x my threshold power and I'd say, that's easy. Ask a different person to do 3x their threshold power and it may not be easy at all. I could do 20+ efforts like this while the other person might not manage even a few even though our threshold powers might be equal. 30 seconds at 175% of threshold power could see similar responses.

asgelle said:

Since it is the total anaerobic work (P x t) that is constrained, it doesn't matter whether that is used up by being slightly over critical power for a long time or well over for a shorter time.

Is there a need to consider a reduced output efficiency caused by lowering pH during extreme intensities? IOW, while the input energy is fixed, is it possible that the output work (which is what would be measured) does indeed depend on the intensity of the effort, duration of the effort, or possibly both?

frenchyge said:

Is there a need to consider a reduced output efficiency caused by lowering pH during extreme intensities? IOW, while the input energy is fixed, is it possible that the output work (which is what would be measured) does indeed depend on the intensity of the effort, duration of the effort, or possibly both?

"Critical power" isn't really a biological model; it's a mathematical model, and the math is pretty simple. How much work can a rider do in time X? CP * X + AWC, period. This assumes that "X" is long enough that you can actually exhaust all of your AWC. The model breaks down when you don't have enough time to squeeze all the toothpaste out of the tube.

Still, it's amazing how well this simple model works for efforts in the range of 3 minutes to 1hr. I plotted data from my MMP curve in cycling peaks in Excel, and tried to fit a line to the time/work plot. Sure enough, CP predicts a value for CP which is right on the high end of my estimate of my FT, and the AWC estimates are really close to the same for efforts between 5 minutes and an hour.

One surprise I've encountered fooling around with CP is how important AWC is for apparently aerobic events.

Take the San Bruno Mtn. Hill Climb, an approximately 20 minute hill climb. It's a solo effort, and 20 minutes long, so my training for this event was entirely focused on improving aerobic power. My CP (which seems to serve as a high reading of FT, for me) was around 260W. (Stop laughing. I weighed 60kg at the time.)

My pretty much untrained AWC appears to be around 11kJ. Even this puny AWC makes a surprisingly big difference in a 20 minute time trial for a 62kg rider! The work of getting me and my bike to the top of the hill was about 345kJ. My AWC did 11kJ of that work, so I was only asking my aerobic system to do 334kJ of work. My time for the event was about 21:34, or 1294 s. Sure enough, 334000J / 260W = 1284s. I consider 10 seconds an amazingly good prediction, and I'm doing it with round numbers and not a lot of data with which to estimate CP or AWC.

What if I had some sort of rare metabolic disease that limited my AWC to 0? Well, my aerobic system would have been on the hook to do all 345kJ of work, at a rate of 260W, costing me 42 seconds (1326). I regard 42 seconds as a huge difference in a 20 minute event. Maybe I should have been doing some focused L6 work before the event, in an effort to improve that 11kJ paramater. This is surprising to me, because I'd kind of gotten into the mode of thinking that any steady effort longer than a handful of minutes is all about FT.

So, question for the peanut gallery: how trainable is AWC? As a spindly little slow-twitch type guy, what range of improvements can be expected? 5%? 10%? 100%? How variable is AWC among trained cyclists? How badly does one compromise aerobic development by focusing on AWC? If you're interested in doing well in these shortish, but still very aerobically oriented events, what balance should we strike between AWC (L6/L7) and CP (L2-L4) development?

kmavm said:

Mt. San Bruno is a solo event only when you drop the group or you get dropped by the group. Of course everyone does there own pace as well as they can but a 22 minute time is ballpark 14 mph and is about when the aerodynamics of drafting would help on a windless day. Are you talking about the USCF New Year's race or something else?

http://mywebpage.netscape.com/rechung/wattage/sbhc/sbhc.html
FWIW, you can see Sterling's time in a few prior years and he weighs about 175 pounds.

kmavm said:

"Critical power" isn't really a biological model; it's a mathematical model, and the math is pretty simple. How much work can a rider do in time X? CP * X + AWC, period. This assumes that "X" is long enough that you can actually exhaust all of your AWC. The model breaks down when you don't have enough time to squeeze all the toothpaste out of the tube.

Sure, but part of the OP's question #2 is: if I squeeze really hard, do I still get all the paste out of the tube, just in a shorter amount of time? I agree that's more of a biological question than a mathematical one.

Take your puny 11kJ AWC. 😉 🙂 Could you use all that up in a 30-sec sprint? Surely even a spindly 62kg-er could manage 367w averaged over a 30-sec sprint, so does that mean your AWC is fully expended, or could you rest 5 minutes and do another one? Or would biological limitations mean that if you went full-out that you'd get less wattage output for your 11kJ input?

Do people feel that the CP model understates, or overstates, what they are capable of doing for short efforts?

kmavm said:

One surprise I've encountered fooling around with CP is how important AWC is for apparently aerobic events.

Which, if nothing else, illustrates the usefulness of the model as a teaching tool (which is why I've tried to bring it to the attention of the cycling community).

kmavm said:

Take the San Bruno Mtn. Hill Climb, an approximately 20 minute hill climb. It's a solo effort, and 20 minutes long, so my training for this event was entirely focused on improving aerobic power. My CP (which seems to serve as a high reading of FT, for me) was around 260W. (Stop laughing. I weighed 60kg at the time.)

My pretty much untrained AWC appears to be around 11kJ. Even this puny AWC makes a surprisingly big difference in a 20 minute time trial for a 62kg rider! The work of getting me and my bike to the top of the hill was about 345kJ. My AWC did 11kJ of that work, so I was only asking my aerobic system to do 334kJ of work. My time for the event was about 21:34, or 1294 s. Sure enough, 334000J / 260W = 1284s. I consider 10 seconds an amazingly good prediction, and I'm doing it with round numbers and not a lot of data with which to estimate CP or AWC.

What if I had some sort of rare metabolic disease that limited my AWC to 0? Well, my aerobic system would have been on the hook to do all 345kJ of work, at a rate of 260W, costing me 42 seconds (1326). I regard 42 seconds as a huge difference in a 20 minute event. Maybe I should have been doing some focused L6 work before the event, in an effort to improve that 11kJ paramater. This is surprising to me, because I'd kind of gotten into the mode of thinking that any steady effort longer than a handful of minutes is all about FT.

Or why I don't like using 20 min efforts to estimate functional threshold power: although they are predominantly aerobic in nature, there's still enough of a contribution from anaerobic work capacity* that using a fixed factor of, say, 0.95 only gets you in the ballpark.

*Note that, as defined by the critical power paradigm, anaerobic work capacity is really best translated as "resistance to fatigue during very high intensity, i.e., supra-critical power, exercise". That is, even though the work performed is mathematically treated as coming from anaerobic metabolism, it might actually be aerobically produced, especially in the domain between >critical power but <100% of VO2max.

kmavm said:

So, question for the peanut gallery: how trainable is AWC? As a spindly little slow-twitch type guy, what range of improvements can be expected? 5%? 10%? 100%? How variable is AWC among trained cyclists?

Like anything else, every person has their strengths and weaknesses, so there is significant variation between individuals (esp. if you pull the track racers into the mix). The good news, though, is that AWC is highly trainable, at least if you haven't focussed much on it previously. For example, I was able to raise my AWC by 33% just by adding some really hard level 6 intervals to my program.

kmavm said:

How badly does one compromise aerobic development by focusing on AWC? If you're interested in doing well in these shortish, but still very aerobically oriented events, what balance should we strike between AWC (L6/L7) and CP (L2-L4) development?

Those are really coaching questions, the answers to which depend too heavily upon the individual and their goal event to really address here (IMO).

frenchyge said:

Take your puny 11kJ AWC. 😉 🙂 Could you use all that up in a 30-sec sprint? Surely even a spindly 62kg-er could manage 367w averaged over a 30-sec sprint, so does that mean your AWC is fully expended, or could you rest 5 minutes and do another one? Or would biological limitations mean that if you went full-out that you'd get less wattage output for your 11kJ input?

Other way around, really: power will be higher than you'd predict just from AWC alone, because there will always be an aerobic contribution.

frenchyge said:

Do people feel that the CP model understates, or overstates, what they are capable of doing for short efforts?

The critical power model predicts that maximal power is infinitely high...so I'd say that it overestimates. 😉

frenchyge said:

Take your puny 11kJ AWC. 😉 🙂 Could you use all that up in a 30-sec sprint? Surely even a spindly 62kg-er could manage 367w averaged over a 30-sec sprint, so does that mean your AWC is fully expended, or could you rest 5 minutes and do another one?

In the CP model, I get to use my aerobic power all the time, so if I got all the toothpaste out I'd be pumping out 367 + 260 = 627W. That would be, to put it mildly, a PB for 30 seconds. My real 30s PB is around 500W (don't have CP here to play with).

Is the CP model the right model here? I am looking at anaerobic capacity as a sort of 2nd gas tank (or matchbook), with a limited volume of gas or matches. As I understand the anaerobic energy system, one begins using this resource at intensities above, say, FT and it supplies an increasingly large percentage of energy up to max power. But, is the rate at which this resource is used truly linear with respect to time at power? So, if one's FT is 300w and VO2MAX is 120%FT=360w, then is 1 min at 360w equal to 2 mins at 330w in terms of utilization of anaerobic capacity? Or, is it that the relationship is non-linear (as in the MP/duration curve) but the CP model is linear for mathematical simplicity? And, the model doesn't deal with the rate of renewal of the capacity as a function of time at power < FT. So, it leaves a gaping hole with respect to post-ride match analysis much less real-time match analysis.

RapDaddyo said:

Is the CP model the right model here? ...

I'd say the best way to answer that is to try it and see. Perform the analysis you're considering using the simplest implementation of CP (fixed AWC with no replenishment) and see how well the analysis matches up to data. If it works, great; if not, at least you'll have information on how the model breaks down and possible avenues for improvement.

acoggan said:

The critical power model predicts that maximal power is infinitely high...so I'd say that it overestimates. 😉

🙂

I was thinking more of 'time at power' performance rather than peak power output. 😉

RapDaddyo said:

Is the CP model the right model here? I am looking at anaerobic capacity as a sort of 2nd gas tank (or matchbook), with a limited volume of gas or matches. As I understand the anaerobic energy system, one begins using this resource at intensities above, say, FT and it supplies an increasingly large percentage of energy up to max power. But, is the rate at which this resource is used truly linear with respect to time at power? So, if one's FT is 300w and VO2MAX is 120%FT=360w, then is 1 min at 360w equal to 2 mins at 330w in terms of utilization of anaerobic capacity? Or, is it that the relationship is non-linear (as in the MP/duration curve) but the CP model is linear for mathematical simplicity? And, the model doesn't deal with the rate of renewal of the capacity as a function of time at power < FT. So, it leaves a gaping hole with respect to post-ride match analysis much less real-time match analysis.

A few semi-random responses:

1) despite its simplicity, the critical power model is actually fairly robust. If you want something more complicated, though, there are alternatives, as reviewed here:

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=16284785&query_hl=4&itool=pubmed_docsum

2) as I alluded to earlier, the best way to view AWC in the critical power model is "resistance to fatigue during very high intensity, i.e., supra-critical power, exercise". If you realize that's what it really represents, you're less likely to be confused when, e.g., the absolute values for AWC and maximal accumulated O2 deficit don't match up even when expressed in the same units, as you might expect them to, or when critical power turns out to be less than power at VO2max, even though it theoretically represents the maximal sustainable aerobic power output.

3) various manipulations have been employed to test the interpretation of the slope and intercept of the work-time relationship, including (relevant to the present discussion) different exercise protocols, e.g.,

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=16195993&query_hl=4&itool=pubmed_docsum

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=12665540&query_hl=9&itool=pubmed_docsum

In general, the results of these studies support the validity of the critical power concept, although again it is still clearly a very simplified model of reality.

RapDaddyo said:

And, the model doesn't deal with the rate of renewal of the capacity as a function of time at power < FT. So, it leaves a gaping hole with respect to post-ride match analysis much less real-time match analysis.

Hmmm.... maybe it's time to breakout the WKO file from "The Day RapDaddyo did 47 L6 Intervals"..... 😄

frenchyge said:

Hmmm.... maybe it's time to breakout the WKO file from "The Day RapDaddyo did 47 L6 Intervals"..... 😄

LOL. I had forgotten all about that workout. It could have been labeled, "How to keep training when your glutes are trashed." But, it definitely proved the regeneration theory.

RapDaddyo said:

Is the CP model the right model here?

Play around with it and see. It's fairly easy to copy raw data out of your MMP curve, pick 5 or 6 points between 3 minutes and a half hour, and fit a line to them. Like you, I had the impression that the relationship between work and time would be all complicated and stuff. For me, at least, CP is shockingly robust.

kmavm said:

Play around with it and see. It's fairly easy to copy raw data out of your MMP curve, pick 5 or 6 points between 3 minutes and a half hour, and fit a line to them. Like you, I had the impression that the relationship between work and time would be all complicated and stuff. For me, at least, CP is shockingly robust.

I plan to do that this weekend. But, I think at most I will get a functional (even robust) capacity utilization relationship of time at power for a short elapsed duration of the ride (during which power > AT). For a complete matches model, I still need a physiology model for regeneration of capacity as a function of time at power < AT. If the capacity were permanently depleted, Andy couldn't do his 15s on/off rides at FT+100w/FT-100w for an hour.