Carbonfiberboy 

How'd you get it to recalculate?

SethAZ 

Told it to use the dashboard settings, set the date range to Aug. 31st to the present, and then hit the Recalculate TSS button.

The problem I now have with the graph is that it looks like only half or so of my activities actually came over from Garmin Connect, and that my last Thursday, Friday, and Sunday rides didn't come over. They were all fairly long and moderately intense rides, especially the Sunday ride, so it would have had a big effect on the graphs.

SethAZ 

See above. Just forgot to quote you so you saw my reply.

Carbonfiberboy 

Cool. Hadn't realized it could do that. I mostly use Dashboard as a training coach. I've built a library of workouts (on the left in Calendar). I set the Dashboard time to past 90 and future 27. Then I plug workouts into the next week or 2 week's Calendar until Dashboard's future looks like I want it to. Then I do the workouts or at least try to, modifying Calendar and library workouts as seems good to me.

SethAZ 

Yeah I still haven't done much with it other than look at the charts, then try to figure out how to recalculate proper threshold and max values, recalculate the graphs, and then figure out that a ton of my activities haven't migrated over. I suppose once I've got usable data in the graphs I can experiment with the workout planning and whatnot. The whole program is brand new to me, though, so I've still got "new user with no feel for this place yet" syndrome.

RChung

The usefulness of the PMC will depend not only on the quality of data that you generate but also the decay constants (the "CTL" and "ATL" values) that you have left at their default values of 42 and 7.

Not only have you been using squirrelly numbers, you've also been using the default decays.

What that means is that until you start getting better data and dialing in your own decays, you really shouldn't be relying on some software product telling you what to do. That's like giving a doctor crappy information about your health status and asking him or her to make a diagnosis of what's wrong and a prescription for you should be doing.

As a final comment, the PMC wasn't exactly designed to precisely tell you when to ride or when to rest. It was originally designed to *help* you predict when you'd perform well for a race based on your recent and long-term training *conditional* on getting the right inputs and decay constants. That doesn't mean that you need to avoid riding on a particular day -- if you aren't expecting to race soon, you may not care that your predicted performance today is going to suck. Even if you *are* expecting to race soon, the PMC doesn't necessarily mean "don't ride today" -- it might just mean go for a ride but take it easy. These things run more on a continuum than like an on-off light switch.

Seattle Forrest

Hmmm. That means Garmin could deny your warranty if anything happens. Not worth saving $45.

Carbonfiberboy 

When I started using the PMC a couple years ago, I simply did my usual workouts and watched what the chart did. I think that's important, as a way to get a feel for what the numbers mean and how the numbers relate to how your body feels and how the workouts go. This winter I'm getting much more numbers and theory oriented, which so far seems to be paying off in terms of fitness. I still don't blindly follow what the chart thinks I should do. If I'm not feeling it, I cut it short. If I feel great, I pile on the TS. Weather and who I ride or ski or etc. with also has an effect on what I do. It's still gotta be fun or what's the point?

What you say about decay rates interests me. I've always used the default rates. I've been assuming that they had been rather carefully chosen and who am I to mess with the decisions of cycling gods? How would one determine if they were right for oneself, and how would one then go about modifying them to make them closer to one's needs? I don't mean how to punch in numbers, I mean how to decide what those numbers should be? What are the symptoms of incorrect decay rates?

SethAZ 

Yeah I haven't touched the decay numbers, and I will only do so after informing myself a lot more so that there's a chance I'll be doing it in a meaningful and helpful way. I already know that I don't recover as quickly today, at just under 49 years old (right after New Years), as I did 20 or 30 years ago, so it's likely that making the ATL decay somewhat slower would be better reflective of how my body recovers nowadays.

It's less about me wanting a program to tell me when to ride, and more about me wanting some intelligible data to help me understand when my fatigue level is such that I'll benefit more from holding back, resting, taking a day off for recovery, than I'll gain by riding hard or long in the current state. Or knowing when I'm in an optimal training zone where I'll get the most benefit from pushing hard, riding fast and long, etc. I don't believe in "Embrace the Pain" just for pain's sake. If I've already embraced it, and my muscles haven't recovered yet, then embracing it again is just going to short-circuit the recovery process, where the benefits actually come to fruition.

RChung

I agree. There's a period of time when you just watch and observe, and learn how to "calibrate" what you're seeing with what you're feeling.

I had a long and discursive reply that bikeforums just ate but the short version is, as I've gotten older I recover less quickly so I've lengthened the short term constant a bit, and may do so again. Phil Skiba has a way to estimate the decay constants but that may be overkill.

[Edited to add:] Part of my long discursion was about the history of the PMC; part of it was about the time constants. "42" and "7" actually show up in the model as something related to exp( -1/42) and exp( -1/7). -1/42 = -.0238 and -1/7 = -.1429, which are harder to remember than 42 and 7. If you're familiar with calculations about radioactive decay, you'll recognize that the half-life of an isotope with a decay rate of -.0238 is about 29 days, and the half-life of an isotope with a decay rate of -.1429 is about 4.8 days; so another way to describe the constants is that the CTL constant implies a half-life of training stress of about a month, and the ATL constant implies a half-life of training stress of about 5 days.

SethAZ 

I got tired of waiting for a response from the TrainingPeaks support over why a large number of my activities had failed to migrate in from Garmin Connect. Luckily I still had the .fit files on my Garmin going back to Oct. 31st, so I uploaded those files manually. This still leaves some number of activities missing from prior to that date, but since the ATL decays over such a short period, at least the TP dataset will be much closer to reasonable. It now reflects all of the rides I've done this month, and a lot of the rides I did prior to this month. Stravistix has all of the rides going all the way back to August.

So I brought up the Stravistix training load graph, and put it side by side with the TrainingPeaks graph. At least for the period covering this month they are remarkably similar. This isn't that surprising, since the calculations are supposed to be based on the same theories. Some differences also exist because Strava will weed out things like non-moving time and such, while TrainingPeaks will consider every recorded second, moving or not. Also, this Stravistix graph makes use of the estimated power outputs that Strava comes up with, while TrainingPeaks is using only heart rate (I don't have a power meter yet).

What's interesting is that both graphs show remarkable agreement on the current Form (TSB). Stravistix has me at -25.5 as of today's ride, while TP has me at -19.

Anyhow, if this interests any of you I did a screen grab of the two graphs side by side. As time passes and more data comes in to both, it will be interesting to see how closely they track. This graph is probably the thing that would most interest me in doing TP premium (currently in the 7-day free trial), but if it doesn't offer me intel I can't already get for free from Stravistix, it may not be worth it. As I learn how to do the workout planning and such I might discover other benefits that might be worth it, but that remains to be seen.

Carbonfiberboy 

I tried changing the ATL constant to 9 days and recalculated. Didn't notice a better correlation with being ready-to-ride, maybe worse, so I went back to 7 days.

Carbonfiberboy 

The gross shape of the graphs are similar but the details are not the same. If you think your Garmin is recording your HR constantly, under Bike Settings change Auto Pause to When Stopped. That won't change your Time-Elapsed, only your riding time (Time) and it'll improve your data.

I have premium accounts for Strava, Training Peaks, and RWGPS. I recommend supporting the effort that these people are making to improve our cycling lives, just like I use Patreon to support artists. It's a tiny amount of money.

gregf83 

How do you know what the ‘optimal training’ zone is? Obviously you want to be fresh when you’re peaking for a race but it’s not possible to build fitness if you’re always fresh and rested. Plenty of gains to be had by riding hard when you’re tired.

SethAZ 

If you look at these charts and the three values that are calculated, and how they are represented, at least in Stravistix they label the value of "form", ie: the chronic training stress level minus the acute training stress level, into some zones. One zone is "freshness," where fatigue has been recovered from, and training has built up. If one reaches this state too soon before a race, presumably one's fitness will slowly decline. If one doesn't reach this zone before a race, then presumably one is entering the race with some carried fatigue which will hamper their peek performance. Below this are other zones. The "optimal" zone is the zone where one is supposed to be in the best position to gain from further training. Fatigue is there, but not to so great an extent that it indicates that one would not benefit from further training. Add in some more fatigue and one enters the "overload" zone, or overtraining. This is where one is so fatigued that presumably the muscles have not recovered enough from recent hard training, and further training will just fatigue them even more without really reaping the benefit of the earlier training.

So it's not expected that to get the best bang for the buck one must always let all fatigue fade away through resting and recovery. Due to the way fitness is calculated in these graphs, as soon as one stop training one's fitness starts to degrade. But fatigue degrades faster. So the trick is to time one's training so that one is unfatigued in time for a race, but without much time having elapsed for one's fitness to degrade. If an athlete times it right they can be as fit as possible while being unfatigued.

Of course all of this depends on the formulas and math being correct. Since these have been developed over decades by sports physiologists and experts in the field of training, recovery, etc. I can only trust that they've done their homework.

I am new to using these charts, but if I understand the concepts well enough, the trick is to balance training and recovery so that one is in the "optimal" training zone as much as possible. That means not resting so long that one's fitness drops a lot, and not overtraining so much that the training value disappears and one just gets weaker.

SethAZ 

For Trainingpeaks, don't they want people to keep recording? I read somewhere that they specifically include every second of recorded data whether moving or not, because while not moving one is resting and recovering to some degree from what one was just doing. I've never used auto pause, but I just turned it on at your suggestion to see what difference it makes. For my Strava activities in the past I just looked at "moving time" rather than "elapsed time" to see things like average speed. I do think that the heart rate always included all the data, even when one's heart rate was falling while waiting at a stoplight or whatever.

Carbonfiberboy 

The Garmin doesn't record when it's auto paused. I think it's assumed that one does that, just like any bike computer would do. It seems to me that one wants totals and averages for active work. At least on my Garmin there's also a built-in 2 mph speed cutoff for recording, which assumes you don't want it to record when walking your bike or the few feet surrounding a stop when your HR or power won't add to your information about the workout. I think there's a plausible argument for that. Your hrTSS is calculated by multiplying the HR in each database entry in the device by some number and adding them all up. IOW if your Garmin doesn't pause, you'd increase your hrTSS by just standing around, not doing anything. I don't think that's very helpful when trying to understanding training load.

SethAZ 

Ah cool. Well thanks for the replies. I set it to auto pause, and will watch the data to see what subtle difference it makes.

gregf83 

I've used Strava for years but admit I've never really looked much at the Stravastix data. I had a quick look and I think you're overthinking this.

Stravastix are software developers not physiologists. The zones they're using are arbitrary. Basically, if you're in a build phase you'll naturally be in the optimal zone while your CTL grows. If you go on a cycling vacation or camp you'll be in the overload zone and when you're riding in a key event you should be fresh.

I think it's reasonable to observe this data while you're training but I wouldn't use it to guide your workouts without a few years of riding under your belt.

RChung

I'm not sure that it will take a few years but I concur that at this stage of his riding career the OP should view these things as descriptive, not prescriptive.

SethAZ 

I appreciate the comments. For the record, I actually have a few years of riding under my belt. I'm new to this mode of training and attempting to look more closely at the numbers to assist in my efforts at improvement. I wouldn't say that I'm new to cycling.

I'm in the awkward position of not having ridden nearly the miles that some have, but having ridden way more miles than most people who own a bike.

I'm still overweight. I'm still a superclyde. I weighed 380lbs in mid 2009, and had gotten to a low point of around 245 or so by mid 2010, and then after that went up to around 265 and have fluctuated up and down within a 265-290 range for most of the last six years or so. I've laid down something like 13-14k miles on my various bikes in that time. I was 285 this morning, and am slowly seeing the weight drop. I had hit a sickening high of 305-310 before I started cycling again in August after a 2-year break. I've done a little over 1700 miles since then, and have been averaging around 150 miles/week for the last couple of months or so, which is very similar to the mileage I was getting prior to my deployment in 2014. I think it's a mileage average that I can sustain.

What's got me focused on this stuff is that I'm trying to re-orient my thoughts on cycling. Rather than just focusing on how many calories will I burn this week on the bike that will help me lose weight this week, I've been thinking about what kind of cyclist do I want to be over the next few years, with the assumption that if I ride like the cyclist I want to be, and lay down the miles, and get my endurance up, my cardiovascular endurance, efficiency, and so forth, and continue to eat sensibly, the weight will take care of itself. One of my goals is to be able to easily do century rides, rather than have a century be something that I can do, but feel wrecked afterwards. I'll continue doing 30-50 mile solo rides during the week, 30-50 mile fast group rides over weekends, etc. But I also will push out more and do some 50-70 mile rides on weekends, then push it out to a full century some weekends, until when I do that I'm fine afterwards and that's now part of my normal capabilities. My goals for 2018 include doing my first 200km brevet with the rando folks, and then seeing where I go from there.

So the shift in thinking here is one from how can I use cycling to help me lose weight or keep it off, to how can I become the cyclist I am interested in being, under the assumption that my weight will naturally converge on some much lower, healthier weight simply because I'm doing what cyclists who weigh less than me do, eating healthily, etc.

Right now I'm focused on improving my performance as a cyclist. Improved performance as a cyclist will have knock-on effects that will include more efficient fat burning, hopefully a faster metabolism, etc. The weight loss that will come with those improvements will further improve my performance, etc.

rubiksoval

What you'll learn is that your body is not 1s and 0s and that a chart and how you feel are not always going to mesh. There's an art to it, as well as the science.

rubiksoval

You lose weight in the kitchen. Very difficult to run a caloric deficit and do higher intensity training at the same time for any significant duration and expect performance to continue to grow.

In any case, you got to focus on what goes in, rather than what goes out.

SethAZ 

Yeah, tell me about it. I have been focused in the kitchen on creating a calorie deficit, since I recognize that that's where the largest percentage of weight loss will come from. I also am very familiar with the effects on lower calorie intake have on the training I can do.

For the last few weeks I've been focused on rides using a heart rate that I calculated according to Dr. Maffetone's 180 formula. I believe that keeping to this heart rate for most of my rides for the time being has helped improve my aerobic endurance and efficiency.

I've also been doing more "empty tank" rides where I just drink water and unsweetened coffee in the morning prior to a ride. Since the heart rate target I'm using is down near the zone 2/zone 3 border it's been possible for me to do these targeted rides on an empty stomach in the morning, forcing my body to get better at burning fat for energy. On my 32.5 mile route a couple of months ago I would have to have eaten something before the ride, and then also take a water bottle with a little gatorade powder in it in order to make sure my energy didn't flag too much during the ride. I'm now doing those rides on an empty stomach in the morning, ie: fasted state, with just unsweetened coffee, with plain water in my water bottles, and riding the route just fine, at a higher heart rate than I maintained a couple of months ago, and also at higher average speeds. It seems my body is in fact getting better at deriving energy for the ride from my waistline.

Then, after the fasted state morning ride, I'll eat my breakfast to concentrate on foods that will help in recovery the most. So, I'll get back from the ride and have some good protein and a small to moderate amount of carbs. Later in the day I'll eat mostly fats and protein and lower carbs. The carbs that I do eat help replenish the glycogen, the protein helps the recovery, and protein and fats help keep my body in a fat-burning state more throughout the day.

I do think that it's possible to be losing weight through calorie restriction while also training on the bike if the training improves the fat burning efficiency, and the food eaten is used mostly to recover from training rather to provide the caloric fuel for the rides and to get through the day.

I do realize that once my aerobic base is built up as much as it should be, and I start adding back in more high-intensity things like intervals, tempo rides, over/under threshold kinds of things, the fat burning probably won't be enough, and I'll have to modify my food and energy plan accordingly. Hopefully by then I'll be lighter, and my metabolism will be a lot better at burning fat.

I'll never be a racer, so the highest of high intensity stuff isn't really something I'll probably ever focus on. I'm primarily interested in being able to do things like centuries in respectably fast times, and to go ride with groups of mostly younger and thinner riders and do more than just hang on. I want to be pulling my share of the time, avoid getting dropped on the climbs, etc. Fortunately groups I've been riding with so far have gone fast, hammered some of the time, etc. and I've been right there with them, other than falling behind somewhat on the longer climbs. For these group rides I always eat before the rides, and bring some calories with me. These group rides are about performance, not relying on fat burning, etc. so I treat them accordingly. But my solo rides constitute most of my riding, so I get to follow my plan most of the time.

RChung

As I mentioned earlier, I had a longer response to a question that got eaten by bikeforums giving the history of the PMC. The PMC was developed over maybe a couple or three years by one guy.