need to raise my functional threshold grit
Even regular LT testing seems to be pointless as LT can change from day to day depending on substrate availability (glucose etc). Even if you knew your LT at any given point in time, it wouldnât tell you the duration at which you could maintain that effort before âblowing upâ. 40 minutes? 60 minutes? 70?
I agree partly. Going on HR alone is not optimal for pacing but I feel itâs useful for gauging low intensity aerobic efforts. If my goal is to stay aerobic and my target watts on point according to the prescribed %FTP zone but my HR is drifting upwards and my breathing is becoming more labored (for whatever the reason: fatigue, dehydration etc), I will lower the intensity because my body is responding to the effort in a manner not desired for the goal of my training session.
Yes this is good. In essence, you are using RPE and listening to your body, which is probably the best take away from all this discussion.
Very interesting article
PDF attached
Very interesting discussion guys. How do you guys think big data figures into all of this? TR and Zwift must have a gold mine of data that could be used to help us better understand how the body responds to training. Think of Uber and how they share data with economists.
Itâs overestimated just how much the current TR/Zwift data can teach us. They only currently store power/HR/cadence data, so can only establish trends between power/HR/cadence. At this stage they canât establish how the body responds to training as they donât store that kind of data.
Agreed there are major limitations to what these guys can show with their data. However if the right question is asked they have the power of numbers to show us something helpful.
They donât have a lab and they arenât measuring lactate values or even basing data on standardized power measurements. But the labs that are doing studies with this level of sophistication can only afford N of 10 or maybe 20 in their studies. Researchers drool over the kind of numbers TR and Zwift have.
So yes, they can only be so helpful but Iâd love to see them contribute.
They could ask how FTP (albeit a flawed metric) is affected by training consistency or time of day training is performed and use huge numbers to power their study.
Imagine if they outfitted us all with some sort of accurate lactate meter?
by the way, you may want to follow the aforementioned GC discussion thread. It seems they are actively working on the particular feature by âanalysing the open data (we have 3,500 athlete years and nearly 700k workouts)â ⌠If I remember correctly, as a user of GC you can opt-in to sharing your data with the devs
That is a limited view, IMHO.
They have all that info, and can correlate FTP info, TSS, time, training plans and workouts associated, and potentially other metrics generated from that data that we do not have access to see.
There are many, many ways you could slice that data if you look at it from a broader view.
The problem is the broader you start to look at the data, the more external factors (additional training / diet / sleep / life stressors etc.) start to muddy the waters.
They can accurately establish power/HR/cadence trends. Anything else becomes speculation.
It really depends on what you want to do with the data and how precise you want to get. Sure, itâs not a statistical study in a lab, but is that level of control really necessary to draw meaningful conclusions on some of these topics?
I donât think so.
The big show they plan to have will probably draw on this data and their other knowledge, to come to conclusions and recommendations that should be useful, IMHO. I doubt we will have anything earth-shattering or ânewâ.
Rather, I think we will see more practical results and demonstrations of the âcommon wisdomâ that shows how well they âcan workâ when certain things are done and followed well. It canât and wonât cover every aspect of things like diet, sleep, etc. since that isnât in the data. But things like plan adherence, total training time and TSS, and general trends in things like typical performance gains (FTP and the entire PR range) are likely worthwhile even if they arenât precise and tightly controlled.
If the sample is high enough then the external factors kind of take care of themselves.
Without having % of Max or threshold HR, the HR data is not beneficial.
Sure, but they may well be able to pull that from all the rides and tests.
I have not idea if they can or will do that fully, but it is a possibility despite the lack of direct input in some cases.
If you donât have the confounding variables available to assess or add to your statistical model then your results may still be affected by them regardless of sample size.
This is why you often hear people say that correlation doesnât equal causation.
Itâs not like they are trying to treat cancer based on results from a study looking at this data, so getting hung up on the confounders isnât a big enough hurdle for me to look at the data.
Sure there are limitations as there always are with any study. Results produced in labs (with plenty of attempts to limit confounders) with only 8-10 subjects also have serious limitations.
So weâd be crazy not to let researchers look at this data. I for one would love to know what happens in the fat over forty demographic with Trainerroad!
Iâm absolutely not saying that researchers shouldnât have access or that important questions canât be answered by using the TR dataset. Iâm a PhD student using administrative data so am well aware of the limitations and strengths of this type of work.
However, confounding must be taken into consideration and isnât any less of an issue in a large sample which was the assertion I was responding to.
Word
If you take the data a company such as Training Peaks hold - the full spectrum of training activity (bike/run/swim/strength), HR max/threshold data, HRV and sleep data pulled in through third party services, daily answers to subjective questions (how you feeling? / illness / injuries) etc⌠now thatâs the kind of data researchers would drool over with regards to the questions weâve been asking.
Donât get me wrong, the TR data set could be really powerful for more specific questions. Just not how we respond to training.
I donât think the âBig Dataâ will be of any use in the FTP Myth debate. Sure Zwift/TR/Garmin all have tons of user data in theory. But not much of it can be used very reliably, as none of is captured in a controlled environment. Thereâs just too much room for error. How do you know if power numbers from a Kicker or a Taxc Vortex is accurate compared to a Stages or Quarq. Or if Jerryâs watts / kg on Zwift is not adjusted to boost his ego. Maybe trends could be observed but I really donât know if the data is reliable enough to answer specific questions. These are just my initial thoughts from personal experience using Zwift/TR/Strava software along with a variance of hardware and the frustration Iâve had through the years with power drop-outs, spikes, calibration errors, GPS errors, HR flatlines, the list goes on. All that to say I would not trust ANY of the data from my workouts to be on a scientific study to answer fundamental questions regarding FTP or effectiveness of THR vs POL training. But Iâm no data scientist.
Good lord.