Clearly this topic has come up before, but II can’t find an answer that actually address the question: let’s say that I start doing 4min intervals (above TP) and then recover until I’m back at 100% reserve capacity, then repeat again.
Once my MPA is fully restored, I should be able to replicate the effort again and again according to the model, which obviously is not the case in real life.
Intuitively I’d say that two things should happen:
The MPA should decline during the course of workout or activity
Recovery time constant should extend as the activity progresses
Clearly both of these time constants are very subjective, but I would think they are ways to extract and model that from past activities
This is easily illustrated by turning a Seiler 4x4 into a 50x4 (which I dare anyone to complete) where after 5h15m of torture the MPA is modeled to be back to 100% within a few minutes (where it would probably take weeks)
Since difficulty seems to monitor closely the energy expenditure, why not using it to adjust the effective MPA or, I’d call it MPRA (Max Power REALLY Available)?
MPA is an abstraction of the energy actually available to perform work, and it’s composed of multiple contributions from different energy systems, with different time constants: it seems like currently only the “faster” systems are modeled.
Seems like these ideas are already on your board (together with temperature and altitude impact on MPA), and I’m hoping to see them soon implemented.
We’ll get into this a bit more during a podcast. Knowing how much MPA is affected during longer activities is essentially trying to map how your TP, HIE and PP change over time based on second-order fatigue processes. Some of these are dependent on other information like how much you’ve eaten during the ride so accuracy will depends on this and the results will likely only be estimates. In some ways, predicting your fitness signature after hours of riding could be considered the “holy grail” of fitness data analysis.
Could you put in some user defined function that would provide a means of decaying your MPA over a period it time, similar to how you are able to individually set your signature decay.
I think people might be a little confused, so I’ll clarify things a little with how we thing things would/will work.
Your MPA (anchored to your PP when “recovered”) would not decline much, since PP doesn’t really decay over time (Peak Energy is supplied by PCr, not Fats/Carbs). Instead, your TP and HIE will decay far more. At the end of a long ride your MPA would still be at/near PP, but MPA would decline faster during harder efforts (since HIE is compromised), and it would start dropping at a lower value (when endurance energy is depleted, anything above LTP would drop MPA).
If I understand correctly then this is currently not modeled: again in the case of the 50x4, the MPA decay during the effort is the same on interval #1 and #50 (in the workout builder at least), while it should be incrementally faster (as it is IRL)
This means also that the chances hitting a breakthru decrease exponentially as a workout get longer.
Yeah, its not currently modeled, as I stated. These are ideas of how we think it will/might work.
To be fair to Xert, no one on the market is close to being able to predict MPA in short duration exercise, much less over longer duration activities, due to the huge amount of possible variation (nutrition, intensity, muscle glycogen levels, etc.). As Armando mentioned above, being able to be remotely close to predicting this would be the holy grail of endurance exercise.
Oh absolutely, I’ve been paying attention to the various platforms that are coming out and I’m convinced Xert is on the right track. I’m looking forward to see how far you’ll be able to take the human modeling.
I would think that data analysis may be able to extract fundamental time constants and modeling variables that are unique to each athlete, and possibly get closer to that holy grail.
There are many reasons for not being able to hit a breakthru on a given day, including mental ones. Personally I rarely hit it on the trainer, it is always on a group ride or some type of event. Others seem to have enough drive and motivation to push hard indoor.
It’s not that I am obsessed with the idea of hitting a BT, but it seem like it is necessary to keep the fitness signature up to date. Again for me, going out on a spirited group ride does it, but sometimes having a reliable way to induce it on trainer would be nice (and I’m not a big fan of the ramp test)
Appreciate the time you guys spend answering our questions and the hard work you put in this platform, keep it up!
I’m very familiar with that article (and that whole research team), as it’s relevant to my field of research.
I can see why they did it, but I’m not a fan of the 3MT for determining fitness signature. In my opinion, in order to get the best result, the test needs to be performed perfectly - good gear selection, good cadence, etc.Additionally, it almost looks like there were ‘responders’ and ‘non-responders’ to decreases in performance after 2 hours of heavy exercise. This figure shows in the individual data… to me it looks like there are many athletes who showed almost the identical responses before and after 2 hours of heavy exercise, while others were more severely affected by the exercise:
Interestingly, they had a relatively fit cohort (relative VO2max of ~57 mL/kg/min), so would be curious to see this experiment replicated in an elite/highly fit group of athletes (Vo2max in excess of 65 mL/kg/min).
Reading Scott’s and Armando’s comments I understand why it’s difficult to mathematically predic failure after 3h of hard afford.
But…
1 - At the moment, if I’m 30 minutes into a workout and try a 2 minutes max afford and I fail at minute 1, you will tell me that I’m probably tired or I didn’t refuel correctly or not mentally prepared (if signature and power source are correct). So I can’t see this be a problem after 3h.
2 - if my signature have a 4th parameter(constant?) “intra workout signature decay” not so most based on a perfect mathematical model but more in data points of my workouts (breakthroughs) Xert will be closer to reality than is now. This have the plus of a few more BTs opportunities
3 - Xert already have my power curve… stating that I can hold my TP for 5h doesn’t make sense.
I understand why moving more into a statistical model is not so appealing but is closer to the truth. The workout example modelled by the OP is clear on that.
Now that I sorted the
When can we expect an update?
Jokes aside, being using Xert since August and I’m loving it! Please keep going
We have a concept called “Endurance Energy” which is already in use partially. You can see the effect on your power curve (calculator) where longer durations show power below TP.
Our approach to this continues to evolve and one of the reasons why it’s not in the system at this time. It is a very complex to model, quantify and depict properly. There is already ample complexity in the 3 parameter fitness signature concept for most athletes using the system already and introducing yet more isn’t a simple matter.
An approach we are considering is to be able to analyze activities manually by accounting for various 2nd order affects like Endurance Energy and others like cadence, altitude, gradient, heat, etc. All these can have an affect on an athlete’s abiility to reach MPA and having a way to interpret the data with these would be useful at the competitive and even professional level.
after more than 3hrs of social pace
