OUGrad05

I am a Strava "Summit" Subscriber which is their paid service. I use Golden Cheetah to look at various power levels over time, it's pretty good.

Getting a power meter might help you improve your power by allowing you to properly train, especially on back to back days. I'm not sure how old you are but I suspect you'll make some gains in the coming months regardless of your age.

OUGrad05

Ok, I found this page. It says my estimated FTP is 249 watts, which is funny because I updated this past weekend to 250 watts after I sustained 269.7 watts for 20 minutes. I don't believe the 95% rule applies to me and believe it's more like 92 or 93% so I use that number.

Rides4Beer

I don't have Strava premium, which I guess you need for the FTP estimate? How about the average power estimate, is that at all accurate? I recently did a 5 loops on a 7 mile TT course at a hard effort, was pretty steady the whole time as far as avg speed goes, and Strava estimates 280w avg power. It was an hour and a half, instead of an hour, which is what I guess a true FTP test is supposed to be, it was a pretty hard effort, but not sure I'd call it "all out". So would my 1 hour FTP be higher? Or am I misunderstanding the whole thing?

Just curious, still fairly new to cycling, and there probably won't be a power meter in my near future. Right now I'm just doing like I do with running, go hard when I'm feeling good, and take it easy when I'm not feeling it.

Psimet2001 

Sweet geebus. Anyone care to guess why the fact this thread exists really cranks my gears?

Before you hand wring your way through another post on here answer a couple of basic things for me:

1. What is FTP?
2. Why is it beneficial to know what your FTP is?

Now if you answered those two questions correctly you might start to be able to understand that FTP is not:

1. It's not a genital measuring device.
2. It is useless when estimated after the fact for ever single reason ever except for 1 and even that has little value if your entire regime of training is not 100% power based.
3. Strava is not and never will be the source for FTP data and information. It will be a nice place to file it but leave FTP to the proven devices and methods for training with power.

hsuehhwa

It's only comparable to power meters when climbing steep grades (aero resistance is minimum).

It tends to be very inaccurate when you're riding with a big group (over-estimate) or in an extremely windy condition.

base2 

In my experience it is pretty accurate. I have direct measurement power meters on all of my bikes and have done many all out 1 hour efforts. Everything jives and I can't walk the next day or 2, I know it was truly "all-out" effort.

I then plug that number into "settings--->my performance" to feed the fitness/freshness graph.

Stravas recovery time (fatigue & form) estimate...That is optimistic, I think. But maybe that's my being a crybaby lil "b" about tired legs.

RShantz

How are you so certain of this? Why wouldn't someone want Strava to continually improve their data reporting capacities? I'd think they should be able to pretty similar to the other software out there. Personally expanded, accurate data analysis from Strava would be great - I'd stop using any other data acquisition/interpretation software.

I'm really confused by this comment. What other magical proven devices do other softwares use that Strava doesn't have access to. Silly me thought the proven device was essentially a power meter and Strava certainly has access to that data.

wphamilton

That's just what you'd expect, because on climbs the variable parameters contributing to drag are less prominent and the invariant parameters of weight are more prominent. Because of lower speeds, and more effort to lifting the mass.

By the same token you'd also expect the long downhills to be the least accurate.

As for the article's claim that Strava is least accurate "at low efforts", it's probably because his low efforts were (almost) coasting downhill. Strava power is likely to be most accurate at low efforts riding uphill.

wphamilton

Just to crank your gears a little more:

1. How hard can you go for an hour - everyone knows that
2. Who cares why? Maybe you want to know if you're better now than you were 2 months ago. You say it's only for devising training regimes, but that doesn't make it so.

3. Strava COULD be a source for FTP under the right conditions. Especially if you're only concerned about the ratio and not absolute numbers.

Abe_Froman

Because power cannot be measured with software. There are exactly 2 components to power: force and speed (cadence). So without measuring how hard you push on the pedals, and how quickly you are turning the pedals around, any software is only going to be giving a best guess estimate. Which COULD, I suppose, be useful in an abstract way. But only for very rough estimates. If you're new to cycling and are on a known course, and test yourself month to month. If the first run is 150 watts all out, and the second one is 250 according to strava...you could probably say with a reasonable level of confidence that you got stronger. But not necessarily.

Because, all of the things that strava does not account for are meaningful enough to render an FTP estimate useless for what it is only used for (exluding e-peen measurement, which I acknowledge may be the most predominant and important aspect of FTP), which is training. Trainer with power requires that you hit numbers pretty close to spot on. +/- 5% can make a very big difference for training. As in...being able to complete it, or not being able to complete the training at the prescribed power. The clothes you wear, choice of tires, wind, whether you are drafting in a group, position on the bike - each of those things, ON THEIR OWN, can create a large enough discrepency (over 5%..) to render training meaningless, or at least sabotage your effort to train to a plan.

redlude97

You have no idea what you are talking about...strava records and reports your power meter data the same way as your garmin/wahoo do. In fact in most cases its just pushing the garmin data directly to strava. We aren't talking about strava power estimates, we are talking about how strava uses powermeter data to estimate FTP, in ways like wk04, golden cheetah, xert etc.

redlude97

The biggest issue with strava right now is that changing your FTP value as the season fluctuates changes your fitness/freshness profile for all previous activities. This alone makes it currently worthless for training and tracking workouts. Strava with the elevate plug in replaces training peaks for the purpose and is free.

Abe_Froman

Oh got it. My mistake.

In that case I adjust my opinion of Strava Power - downward. I judge it utterly worthless.

Further...the mere existence of strava estimated power is ludicrous. It's like 2 guys at the gym going for max bench press numbers. One guy looks at the bench press. Instead of seeing how much he can lift...he asks the other guy how much he thinks he can lift...and writes that down as his max bench press number.

Gym Rat #1 : "Bro!!! Let's see what our max bench press lift weights are! I'm SOOOO jacked!"

Gym Rat #2 : "Awesome! I love it. You go first. But hurry, Dawsons Creek is on in an hour."

Gym Rat #1 : "Will do brosky! [looks at bench press with trepidation] Bro, what do you think I can bench? [flexes gargantuan guns in front of a mirror]

Gym Rat #2 : "Mmmm. 650. Easy. Just LOOK at you, you're HUGE!"

Gym Rat #1 : [writes down 650 as personal best in log book] "You mind if I make that kilograms?"

RShantz

How are you so certain that the calculation Strava uses to estimate FTP isn't as accurate as how you estimate FTP (using the same exact power meter data)? I just don't know how someone can say it's useless if you haven't determined it's methodology compared to what the expert say is the best way to estimate FTP.

Again, to me, having the data analyzed in one software is a good idea.

redlude97

So I see you've decided to double down...

Abe_Froman

Definitely. Go big or go home

Abe_Froman

If you only ever ride at 60% of your FTP or below, how is Strava coming up with your estimated FTP? Or if you only ever ride on group rides? Or if you very often do short interval training?

I can't possibly imagine how a simulated FTP number serves any purpose other than to be a cute toy to look at online with the purpose of increasing and maintaining subscriptions.

redlude97

I don't even think you know what FTP is at this point, do you even have a power meter?

Abe_Froman

I have a smart trainer with power. I'm fully aware of what FTP is, thank you.

base2 

Because no one who is interested in FTP only rides at 60% of FTP.

It has all the data from all the powermeter equipped rides. At some point you did a max effort somewhere & it uses that as your "best effort" and goes from there.

As conjecture: It also has your heart rate data if you are so equipped and knows if you've exerted yourself, calorie usage, respiration rate, & thus metabolic efficiency (VO2 max from Garmin or another source) may even factor accordingly.

As with anything, it's accuracy depends on the data you feed it. So what bike you ride, the type of bike, the weight of the wheels, the weight of the tires, your weight, all the miscellaneous components, the frequency/accuracy of GPS data points, your heart rate, the wheel mounted speed sensor to suplement GPS drift, the terrain grade,the frequency of your rides, the length of time you've been a cyclist, (meaning your adaption to the activity) all of that feeds into it's estimate if you do not have a direct measurement power meter. So it's estimate accuracy depends on you, the user.

People who don't care about FTP get innacurate estimations. That's not really news.

redlude97

So you don't actually know how to train with power, got it.

Abe_Froman

And how exactly would the software know if what you did was a max effort? Rather than a sweet spot interval? Further, if you have a power meter, and actually DID do a max effort for a significant length of time, like, say, 20 minutes or an hour...why on earth would one take Stravas best guesstimate for your FTP instead of using the common and accepted formulas used to derive your FTP which you JUST TESTED IN THE TRADITIONAL METHOD?

base2 

It doesn't know if what you just did was a max effort. It knows what your efforts were within the last 6 weeks. The only meaningfully accurate number is derived from direct measurement. Everything else in all other methods including the traditional method is estimation. Is your heart rate 110bpm & your power 130 watts? That ride goes in the large gray area under the power/time line & is not used to determin FTP. It was just a ride & nothing more. Time to adjust the form/fatigue graph according to the settings in the "my performance" tab.

Your actual FTP on a given day is remarkably transient in nature anyway depending on form & fatigue. That's why the term "over trained" exists. Your peak efforts are used instead. Doing it this way discounts the lesser effort variations that would tend to pull the line down & provides a more accurate representation of athlete potential because by definition it pulls from the most strenuous data points. Even the traditional method has you start in a fresh/good form state in order to achieve the same result. Strava just has more data resolution to pull from.

redlude97

If you haven't been riding at your max intensity how would you know you are riding at sweetspot if you haven't formally tested recently?

Abe_Froman

So, you're saying strava ftp estimates only work if you already know your ftp through proper testing??