Link takes me to the top forum page. Executive summary?

Pert near.

That Garmin Forums article said that you can use either elapsed time or moving time as the denominator, and that you have to be sure which one you're using. In addition, the "moving time" calculation from Strava can differ by a couple of seconds from Garmin. This was about speed, and it said nothing about whether the same thing happens with average watts.

I don't recall ever looking at "moving time:" I always look at elapsed time, so I wouldn't have noticed any disagreement.

My son stayed at the house and used my trainer with my zwift account and raised my ftp by 50watts! He wasn’t even trying.

My son a few years ago when he was in college used my Garmin 935 watch for a run and all of a sudden everything was off. I had set a record at every distance I had ever run and my max VO2 went through the roof. He was a 1:57 800-meter runner and ran cross country. His training pace was my 5k PR.

Just did the ramp test today and the good news is that in 4 years my FTP is down 2 points (Just turned 70). My legs had far more to give at 300 Watts for 1 minute- not even remotely entering a pain zone, but I just couldn’t get enough air - ran out of breath at 182 BPM and it was game over. In mid-October I managed 376 Watts for a minute, but was before my 6 week hiatus - should have measured it then but it certainly wouldn’t reflect today.

What is depressing is that I was far from making it even to the half-way point on the ascending ramp (regular one). Will take what I can get. Understand HIIT training should improve oxygen uptake, so will work on that.

Much appreciated but your numbers are very impressive especially with 268 hours. Well done sir.

No advice about the disparity in your indoor and outdoor numbers.

Strava does not use normalized power but plagiarizes isopower. It's the difference in using a 25-second moving average versus a 30-second moving average. Are you counting zeros? https://support.garmin.com/en-US/?fa...iH6PwvWWSGx8q5

Unless things have changed recently, Zwift's automatic FTP detection (like every other app) and default tests, test high. Hunter Allen's original protocol is a 5-minute VO2 max test followed by a 20-minute time trial. Without the 5-minute blowout, your FTP test result will be high. FTP, like all psychological thresholds, is not tied to an artificial time constraint. It happens when it happens. That is to say, it's NOT one hour. Your time to exhaustion is however long you can hold it. TSS is set to an hour. I prefer Kolie Moore's FTP protocol, https://www.trainingpeaks.com/blog/t...ing-protocols/.

Hope you meant “physiological” thresholds. The one-hour thing is a pernicious and pervasive misconception.

Strava doesn't use normalized power, nor is the difference the difference between a 25-second and 30-second moving average; and I'm pretty sure that "plagiarism" shouldn't be the charge. If anything, Strava avoided using NP because TrainingPeaks had trademarked the name (though not the concept), which is odd since they didn't come up with that name -- I did. I was pretty surprised when I found out that they trademarked a name I came up with.

Interesting, all mine recordings match up in Golden Cheetah and I'm not the only one making that observation.

I always inferred it was Dr. Coggan. Thanks for setting it straight.

Yup. I've noticed that too. Golden Cheetah uses Phil Skiba's xPower, so it's pretty clear Strava is using Skiba's xPower. XPower uses a fixed "time constant" of 25 seconds which means a fixed 1/25 = .04 = 4% decay rate rather than Andy's fixed 30-second "uniform kernel" moving average.

You almost surely know that Andy uses as defaults "time constants" of 42 and 7 for CTL and ATL. Many people mistakenly think that means CTL is based on a moving 42-day average of TSS. Just like Phil, Andy's time constant of 42 means a fixed decay rate of 1/42 = .0238 = 2.38% per day, and a time constant of 7 means a fixed decay rate for ATL of 1/7 = 0.143 = 14.3% per day. Fixed decay models are all exponential, so Andy called his CTL and ATL "exponentially weighted moving average" models but in other fields we just call them fixed or constant decay.

There's some (medium but not overwhelming) evidence in the literature that Phil's fixed 4% decay model for xPower matches HR a *tiny* *little* bit better than Andy's fixed 30-second moving average, but I don't think the difference between NP and xP is all that huge. GC and Strava went to xP not because of the difference but because TrainingPeaks originally tried to strong-arm them on NP while Phil said "no problem, you can use xP if you want." It was during the strong-arm that I found out that TrainingPeaks had trademarked the name they hadn't come up with.

It's a nice model, but it seems to me that If individual variability in those decay rates across the athlete population is comparable to that of nearly every other measure, a lot of us should be tweaking those constants.

CTL & ATL? Alex Simmions has a write-up titled "A time for a bit of sensitivity (analysis)" at Alex's Cycle Blog: March 2013. Also see . I seem to remember 10 ATL for older & 5 for younger athletes 5 for ATL as a more recent recommendation for older athletes (Wattage Google group?) but here's an old thread https://wattage.topica.narkive.com/I...-8-days-thread. If you are referring to 25-seconds vs 30, isn't that something back to the adaptation of Banister's Impulse-Response model?

Yup. In fact, Phil gives some guidance on tweaking them. When Andy came up with them, he said they were defaults and thought that over time people would start to "tune" them individually, though he didn't give any guidance on how to do that. But other than Phil (and the athletes he advises) I don't know of anyone who really does that. I think ATL/CTL/TSB/PMC are good as conceptual models to help you think about things but not for prediction of performance or prescription for training.

I was referring to the decay constants for ATL and CTL.

If I remember correctly, I set my ATL from 7 to 10 to reflect my great age and decrepitude, but (in the intervals.icu implementation at least) doing that seems to low-pass filter the whole response, which makes the fatigue hole look longer but shallower. Shallow is not how it feels the day ofter a hearty helping of overreach.

Thanks. I'll look at those links.

As I said above, I think the NP/TSS/ATL/CTL/TSB (i.e.,PMC) ecosystem is a good conceptual model because it makes us think about how things are related but I don't unreservedly rely on them. Andy had one input variable, TSS/day, and used the same analytical form (a fixed decay with impulse) with two different decay parameters (1/42 and 1/7) to construct fitness and fatigue measures, and their difference is freshness. So that's a lot for a single input and two parameters that no one ever tweaks to do.

That said, it sometimes helps some people in tweaking if they think not in terms of the 1/42 or 1/7 but in terms of decay half-lives. Those of us born in the atomic age are pretty familiar with decay half-lives. Anyway, multiply the "time constant" by ln(2) = .6931. Or, more simply, multiply the time constant by 0.7. So, a time constant of 42 => half-life of 42*0.7 = 29 days, or about a month. a time constant of 7 => half-life of 7*0.7 = 5 days. So the default parameters say that your fitness decreases by about half after a month of rest, and your fatigue decreases by about half after 5 days of rest. For that recommendation of increasing the "fatigue" constant from 7 to 10? It means that your fatigue decreases by half after a week of rest, i.e., it takes longer to recover from a big ride. But, as you noticed, the hole you dig (i.e., TSB) is shallower. As you may have forgotten or, even better, never knew, one of my day job areas of research is renewal models. My renewal models are generally more complex, but they almost always allow for a difference in digging holes and recovering. But that requires more than one input and two parameters. Andy made his one input and two parameters work pretty hard, and getting as much as he does out of them is pretty damn clever.

Are you sure Strava estimates FTP? I’ve never seem that feature and I have the full paid Strava sub.
All I can see are my best power averages over various time frames.

Hmmm. Strava has an estimate of my FTP but I'm not quite sure how it knows. I rarely look at it so I don't know how it got there. It also has a value for my weight, and I must have provided it.

I just had a look at my Strava and FTP is just a manual, static, user defined value under my profile. I see no mention of Strava making an estimate for you based on any of your ride data.

In the Strava user guide it suggests how to estimate your FTP (0.95 x 20 min power). I wondered if it would do this automatically off your best 20 min power, but it doesn’t for me. The value it has for my FTP just looks like a value I put in myself that is now well out of date!

I guess I must've entered that.

There's a Strava user guide?

On your Strava account page, go to Settings / Training / Power Curve for a graph of your recent effort's output. Mine shows 6 week (default I guess) as 275W at the start, ramping down to 125W after about 50 minutes. There's a check box there that if enabled suggests Strava will derive an estimated FTP; in my instance that's 176 right now.

I've been struggling with arriving at a preliminary FTP # in Rouvy for a few days m'self.

Doing a RAMP as suggested worked the first time fairly well, gave me something like 135 for the 20-minute test. I attempted a longer RAMP last Sunday after I'd set my FTP # to 170 & just about died getting thru stage 2 after fifteen minutes. Could barely move those cranks so I quit, figuring that manual-set 170 was where it'd started, rather than the default 150 it'd suggested was the starting value in the description.

Nothing hurt but my ego, but hey at close to 76 years I'm well & truly used to that by now.

I suspect I can do better but actually getting to a # I can both work with as well as trust over time will take more work... getting back to that in a few minutes, now that I've cleaned up my bedroom for the year

Ah thanks I see it now. For some reason it doesn’t show that estimate on the mobile version. I have to say the estimate it gives me is massively optimistic! I would be happy to get within 70W of it’s estimate 😂

Pete your help a few days ago gave me the input I needed to get more comfy with using Rouvy so my post is just payback! Thanks!

I just did about an hour's Just Ride route, "Flatlandia" (virtual in Oman, halfway around the world and a LOT more conducive to biking outside today than where I live!) 13.5 "miles" in ~ 46 minutes & not much elevation compared to what I'm used to encountering out on the roads around where I live. Strava's estimated FTP didn't change at all but I'm a lot more comfortable using it than before you talked me through some of the subtleties about setting it up.

Stats are encouraging: avg. 17.5 mph; 29% Z3, 18% Z4, all of 5 SECONDS @ Z7 w/ 264+ W to show for it.

I'm scheduled to work tomorrow so Friday will see my next trainer op. Will be looking to stay in Z2 & 3 next, push the duration up past an hour or more. Save Z4+ for the coming w/e.