I have the idea to let my intervals be guided by power numbers of much shorter time frames as the FTP.
So, I will track my 3, 5, 8, 10 min power. Then I will base my intervals on % of those metrics.
For example, if I do my 3min intervals, I’d do them either as a percentage of my 3min power or, if it works with my 5 or 8 min power.
This is my background thinking:
I think that the training methodology on the bike is the most refined compared to for example running. I think the reason for that is that you can measure your input on the bike much more easily, especially when you use an erg.
However, I want to generalise the elaborate “bike style” of training to other training modalities, for example the assault bike.
Additionally, I want to lower the entry point of using specific numbers in terms of time commitment. I think the 3-10 min numbers are more accessible, though I understand that for some people it is more daunting to test your 3 min power than to test your FTP.
To me, these power numbers are closer to the stimulus that I want to produce. Especially, as a bigger athlete (I weigh ~240lbs, 6’4’', fairly lean) the normal ratios can be a bit off. So, having metrics that are more similar in time duration should be more precise.
However, I don’t find any system that is similar enough, so I can learn from it. So:
Did this approach proved to be problematic in practice?
Am I to incompetent to find material and if yes can you point me to the right direction?
I’d appreciate any feedback or ideas. If you want more details, I’d be happy to provide them.
Setting interval targets at a %age of PRs is a perfectly sensible approach. 95% is the number you’ll tend to hear most often, so I suspect that works well for most people.
Of course, for VO2 Max improvements most people here (including me) are going to tell you to (mostly) ignore power anyway and just do them max repeatable at high cadence until you can’t breathe properly.
I think it’s definitely sensible to track 3-5min power and base intervals off that number (I don’t think it’s really worth doing vo2 stuff beyond 5min intervals, the 3-5min stuff seems to work for most people). Rather than use a percentage of those numbers, however, I’d use those as a target to try to surpass during your workout, if it’s a max effort you want to try a max number, even if it’s only for the first one. I think workout apps are guilty of making people think they have to nail every interval at a consistent power (and especially guilty of defining vo2 as a percentage of ftp), when experienced coaches like Kolie at Empirical Cycling will say it’s perfectly fine for the power output to drop, since you’re chasing a physiological state rather than a power target.
I wouldn’t bother trying to work off a fixed percentage. When I do 5 min intervals my recent best 5 min power is a target. But as long as I reach that controlled by heavy breathing each interval I fine.
If you go too hard, it can make it hard to repeat on subsequent intervals. But if you don’t go hard enough you are not getting to the heavy breathing till almost end.
I find it interesting looking at max heart rate in each interval, post workout. Often I’ll find I’ve reached 90-95% max HR in all intervals, but the average power has dropped.
I don’t go by a fixed percentage and often use the first interval to gauge what I think I can do.
This has been done in the running world with success for decades. Coaches prescribe efforts based on “goal pace” or “date pace.” For example you may have 10 x 300 @ 800 pace or 5 x mile @ 10k pace.
As @Helvellyn mentioned… “the best predictor of performance is performance.” If you are looking for a 2-3 minute effort you might do reps at your 5-8 min power. For longer reps of 5 minutes you might be looking at 10 or even 20 minute power depending on the amount of rest between reps. Generally the shorter the rest, the lower the intensity.
I don’t think that training methodology on the bike is more “refined” than running. I think that runners are pretty used to thinking in terms of pace on flat firm ground, so that’s pretty much like an erg. Riders don’t just ride on firm flat ground: they go up and down hills and are more affected by aerodynamic drag; the explosion in indoor training is an attempt to regularize cycle training so it looks more like running.
That said, what you’re looking for is pretty close to MMP modeling, the simplest version of which is the CP/W’ two-parameter model. If you get those two parameters right, you can predict reasonably well your MMP from around 3 minutes or so up to maybe 30 or 40 minutes and all durations in-between. It also gives you a way to see whether an effort at an oddball duration (like 16:34, or 16 minutes and 34 seconds) is “consistent with” an effort at 8 or 10 or 20 minutes, and gives you a natural way to update CP/W’ with these oddball duration efforts. There are more complex parameterizations than CP/W’, but that’s a reasonable place to start.
And Coggan used to talk about off-season training that he based off of a percentage of his “peak in-seaon MMP” so your idea has merit.
Many thanks for your answers. My fear is that those numbers will not correlate that nice with energy systems.
First, I thought that I will mimic what is provided by the prilepin chart for strength training. The various intensity brackets represent very specific stimuli.
However, I don’t think I will be that precise with the low power numbers. The reason is that I don’t think I learn a lot when I have my 5-minute power and cannot correlate it with thresholds in the energy metabolism or specific work loads for the heart.
I am in search for example for the intensity at which the preload is maximised. My thesis is that this is the “sweet spot” for training the heart, since maximal preload means that you maximised the “range of motion” of the heart.
Power is partly irrelevant. For max preload you want to do sets of intervals of between 3 and 5 minutes at high cadence (110rpm+) to drive your breathing to the “fish in a boat” state, because that’s the best proxy for being at or near VO2 Max, which should mean max preload. If you need a start point, try 150% of FTP and hang on for dear life.
Why are you trying to reinvent the wheel? You are taking the same old power duration model and trying to find new things in it. There is nothing magical about 3, 5, 10 minutes. If you just track those then you are tracking only above threshold power durations. That’s just a small part of your curve.
If you want to train energy systems, the basics are:
endurance (sub LT1, VT1, aka aerobic threshold, mid to low zone 2)
threshold (90-100% of your measured FTP) - long intervals extending the total work you can do during a build phase (3x10, 2x15, 3x12, 2x20, 3x15, etc. roll it however you like). An improvement on the classic 2x20
“vo2” intervals - 3 to 8 minutes - hard start, gasping for air. Pick a power you can sustain for all the intervals (12 to 25 minutes of work). The most famous is 5x5min. That’s a stiff workout so it’s best to build up to it. You probably still get 95% of the stimulus with just 4x3min. (Kolie Moore says the whole point is to train the cardiac preload. Seiler’s study found 4x8 to be the “best” by a tiny margin. Gains from 4x4 and 4x16 were very close behind.) Michael Rosenblat’s meta analyis of intervals found that the longer intervals (like 8 minutes) improved 40k TT times the best. (I’m not sure it matters which methodology you follow.)
anaerobic capacity - sprint intervals or short intervals- icing on the cake
Normally I’d keep my yap shut but since someone tagged me in… For the Prilepin chart, those brackets don’t really represent stimuli that are terribly specific. Either you’re building muscle or you’re building strength/skill, and sometimes a bit of both. And I won’t get started on the sacred cows of the 8min interval studies.
I personally find the approach of doing suprathrehsold efforts as a % of that duration’s MMP to be superior to the %ftp approach since it’s highly individualized, and I’ve used it on occasion with people. It’s also an approach that a lot of other coaches use to great success. I don’t really find it quite as effective as more maximal efforts in the 2-6min range, particularly at higher cadence. It therefore can be a stepping stone to that approach, or I’m happy to ride it out as long as I’m seeing a good training response especially if someone’s enjoying it.
There’s no need to get into more nitty gritty theoretical applications of stroke volume and preload and whatnot, and I feel like I can speak to this since that’s my original thesis based on common training methodology, available literature, and evidence/experience. Chasing niche physiological adaptations is almost always a waste of time for any of three reasons: the assumptions of the underlying mechanisms are incorrect; or it may be costing you in some other area of training; or it’s not actually as big a limiter/driver of performance as you think it is. In any one or combination of these cases, you can spend months going nowhere with your performance. Even if you got the theory right, because you’re missing either a more nuanced or integrated systems perspective. So as Coggan has said many times, train for performance and your physiology will sort itself out.
I think the issue on how to think about training is if you either use a black or a white box model approach. Giving up on aiming for specific physiological adaptations will lead you to a black box model approach. (I don’t mean that you completely ignore any of the physiological substrata of performance!)
Sometimes, you have to accept that you have to black box something.
However, the underlying model is only then correct if it maps the actual structures trained which are the physiological structures.
This is the theoretical problem of the VO2max training. In the end, it is a result of a black box model approach: You just design a workout program, test it, and then tweak it until you get the most response. But you can’t do very much with this model, since you can’t figure out how the interaction between this and other training modules (like Zone 2) is. The interaction happens on the physiological level. Only here, you can think properly what happens, if you train both.
What I am after is not to chase down a niche physiological adaptation, but rather use an incremental epistemiology to map workouts and weekly training schedules.
To give you a concrete example: Ignited by Peter Attia, zone 2 training made its way into a somewhat broader audience. There is the general advice on how much zone 2 is needed or wanted (currently, the main stream advice is 3-4 hours). However, the true individual zone 2 needs within such a program depends on the total volume that you already implemented in your regimen. If you for example included high volume, medium intensity work like sled work for your general conditioning (sample workout: Push forward, Pull backwards, Pull Sidewards; continuous work for 45min), you already got way more volume than compared to the typical gym rat. (those workouts provide enough ST fibre load and rhythmic contractions to elicit a stimulus for the heart)
I think this attitude ignores prior knowledge, we are standing on the shoulders of giants: our current training approach is based on a lot of trial and error.
Moreover, our knowledge is incomplete. E. g. I was talking to some people in our sports sciences departments. They are supporting German olympians in several aerobic sports that are not cycling, i. e. at least in that discipline, they know their stuff. I asked them what the state-of-the-art regarding super compensation is and the answer is surprisingly little if you measure performance outcomes (rather than, e. g. effects on glycogen usage). We in fact talked about whether it is possible to find better models for super compensation that predict performance on the bike or other things (I’m a mathematical physicists by trade, so those mathematical models are something I am familiar with).
What is the aim of your training, “general fitness” (however you want to quantify that) or “fitness on the bike”?
If it is the latter, what performance benefits do you expect for cycling? How would fatigue incurred during such a workout impact your training on the bike?
I’d be really surprised if that were true to be honest. Anything as complex as human physiology will be impacted by stimuli in more ways than one. Moreover, there are so many relevant factors that I can think of, which make that proposition quite unlikely. For example, you can shift the load (to a degree) from your muscles to your cardiovascular system by increasing the cadence.
See, there are already a couple misunderstands of physiology and its application in here, which is a danger of getting too granular with the “white box” approach. To overextend the metaphor, physiology is a “gray box” where we can certainly understand some direct drivers of stimuli quite easily and make very general training recommendations. In other ways physiology is very much inscrutable, if for no other reason than we don’t know everything yet, which is why they haven’t shuttered Science or, for our purposes, EJAP.
This is precisely the issue. We have no shortage of confounders in physiology, hence my caution above. And this is before we even consider individual variation. I have a couple clients who respond incredibly well to vo2max training that looks absolutely nothing like anything that anyone has ever heard of as vo2max training, but that doesn’t mean that everyone responds to it, otherwise the fallacy of division wins again.
Regardless, the reality is that we can theorize with the “white box” model but ultimately we must ensure that performance is improving, meaning we must fall back on the “black box”. Stimuli → meat grinder → measurable gainz, hopefully. Just the way grandma and grandpa used to do it. Just kidding, women weren’t allowed to do much more than tennis and golf back then.
Anyway, just did a podcast on something adjacent to this so it’s on the tip of the brain. Back to the topic at hand… there is decent research on cardiac preload as I elucidated in my vo2max podcast series, if we can forgive the error associated with the inherently imprecise measurement methods. But since you ask to be pointed in the right direction (or what I consider so), I shall point. First look up differences between cardiac and skeletal muscle function, then check the wiki for the frank-starling law, and lastly look up stroke volume for highly trained vs moderately/untrained individuals. Let me know what’s in the box after that.
I apologise for the delayed responses. I was gobbling down material by @empiricalcycling. By far, the best podcast I ever listened to.
I am not ignoring the prior knowledge. Quite the contrary: I treat the body of knowledge of the biking world with great gratitude, since here – at least, as far as I can estimate – people are actually nerding out in a good way.
The incompleteness of knowledge is in part of what is attracting me.
It is hard to describe my approach to other than a bottom-up septic tank. Opposed to mathematical modelling, I work more like a linguist would think about language. I try to figure out coherence, consistency and cohesion of the language used to map reality. This allows to speak meaningful about something that you don’t know.
I 100% get your critique, since normally the reasoning is mechanism → activating mechanism → win big. This is not what I am after.
For example: If there is a movement-specific intensity at which cardiac preload is maxed out, and you have good reasons that it is your heart that limits your performance, then you are justified to accumulate volume week over week around this intensity. Nevertheless, you’d have to also consider what you are leaving out, what you are adding etc.
Yes! You are not overextending the metaphor. White- and black boxing are both cognitive techniques which have their utility in different circumstances. This is what I mean.
I am not sure if the following example helps exactly to get my point across, but I try:
I don’t add “Zone 2” to a training plan as a physiological intensity. In many situations, the crucial input into the system is not physiological but psychological. I explain Zone 2 running as the speed at which you feel light, elastic and bouncy. The running rhythm should dictate or force the breathing rhythm (though you can synchronise them if you prefer). If you think about a problem, you should experience a slight shift towards convergent thinking, but you should be able to think freely.
If you can’t run like that, you can’t use running for zone 2. Like this, zone 2 running (in the small doses I use it) is almost free physiologically (as long as you feed yourself properly) and a net positive psychologically.
This is based on a phenomenological approach to training, since intentions are the first mover of training sessions. So, it is exactly, how you described it later on: You use a white box approach in the background thinking but then, in application, you have to blackbox to focus on both your input (how it feels) and the output (performance).
For sure.
I am already force feeding myself with your podcast. By far, the best podcast on training in general that I know!
