Seattle Forrest

I did a long ride last night, at an easy pace, but with a lot of hills. The stats from the ride show a training effect of 2.5 (halfway between "maintaining" and "improving" my cardiovascular fitness) and an intensity factor of 0.972 (almost as much as I'm capable of).

Does this mean it's time to take an FTP test? If so, can I pull the numbers out of my ride data? I'm ok with a close approximation, and I only get so much bike time lately.

DaveWC

How long was the ride? How does your 20 minute power within the ride compare with your FTP? I don't see how it's possible to have an easy paced ride with an IF of 0.972 and not have your FTP be too low.

RJM

Time for another FTP test, I would think.

Seattle Forrest

About an hour and a half. Average HR for the ride was 138, which is zone 2.2 based on the difference between my max and resting HR.

When people talk about the best 20 minute power within a ride, is that average or normalized? And if it's NP how do you calculate that? I've got Golden Cheetah but not Training Peaks.

RChung

How much accuracy in FTP do you need?

I hardly ever do formal FTP tests any more because I don't need much accuracy. I do eyeball the running evolution of my CP (or, one estimate of it) and if I start to get a few exceptional performances in a row then I'll adjust my estimated FTP a bit.

If my needs were more critical (sic) than that, I'd do a formal FTP test.

DaveWC

Average. Garmin Connect gives it to you as a standard statistic of every ride.

Seattle Forrest

It doesn't have to be perfect, but I'd like pretty good. I'm using it for training purposes and trying to figure out how to get a handle on my training load.

Seattle Forrest

Ok, good. They give you average and normalized power for the whole ride, and then in Connect you can "zoom in" on the graph to see your best average for different time periods, like 20 minutes. I know there's software that can calculate NP for portions of a ride but I don't have it ... and it sounds like I don't need it at this point.

RChung

That's not a particularly demanding usage, so you can probably either do a formal test or you can eyeball it. If you were racing ITTs you'd probably want a better estimate but tracking training load is a "low fi" application. Start paying attention to Golden Cheetah's CP estimates over several rides where you're making consistent big efforts (that is, high session RPE). I wouldn't base a new estimate on a single exceptional effort -- you're looking for a pattern, not an exception from a pattern. Then I'd bump my FTP up in accordance with that. Note that GC's CP tends to overestimate your "true" threshold by a little (that's why if you were racing and needed a good estimate you wouldn't do it this way).

Seattle Forrest

Thanks RChung! I need to spend some time learning more about the critical power chart now...

PepeM

On the topic of FTP. Say I wanted to do an FTP test. I would be doing it on a trainer so I can keep a constant effort for one hour (or twenty minutes, or eight, or whatever the method you're using calls for). How would I know what power output to aim for? Say I feel good at a certain output after 10 minutes, how would I know that is the max effort I can sustain, or that I would not blow up too soon?

valygrl

What is TE in trainigpeaks lingo?

Seattle Forrest

TE = Training Effect. It's not a Peaks thing, it's Firstbeat, a research company Garmin (and Polar and Suunto) buy algorithms from. It uses heart info to determine how much stress a workout put on your body and what effect it had on your fitness. Here's their white paper.

It's only one stat among many but generally tracks pretty well against RPE, including yesterday. Earlier this week I got TE = 4.2 and IF = 1.205 for about half an hour of hill repeats day.

valygrl

Thanks.

IF 1.205 for half hour would say to me it is definitely time to test/increase your FTP.

TrojanHorse

FTP on a trainer will probably be higher than your FTP if you measure it on a road somewhere, but if that's the best you got go with it.

Part of the process of developing an FTP is familiarizing yourself with the test, so odds are you will not do it right, or well enough, the first couple times. Don't let that get you down, it's like everything else and a little practice will improve your results. Next time you'll know to go a little harder right from the start. (or easier). It should get to where you are nearly dying at the 20 min. mark if you're doing one of the 20 min. FTP tests.

Make sure your legs are nice and warmed up first.

Seattle Forrest

I'm pretty new to riding with a power meter, so I'm probably asking some dumb questions. Sorry everyone has to bear with me. I'm trying to learn everything I need to know, I think I'm at the point where I can't absorb much more until I spend more time using it.

But to make sure I understand:

My IF was unreasonably high, because my NP was > my FTP. So instead of taking the time to do a test, I could go through my data. I've set three personal records for max average power in the last two weeks (the most recent was 1w higher than the last), and it sounds like I can take one of those for my FTP. But if I do that, my normalized power will still be higher than my average power.

I live in the city and ride after work almost every day. Lots of lights, some traffic to contend with, etc. NP will always be > avg P.

cthenn

Good stuff in this thread, thanks!

DaveWC

NP is always higher than average power, they're closer if you ride indoors at a steady speed/power. It's an estimate of what your average power would be with the same overall effort you put out for the ride. So a 1.20 IF for a 30 minute ride means that your FTP is likely too low. It should not be an easy ride to get your IF>1.0 for rides of >20 minutes.

Athens80

Wouldn't the critical power (CP) curve in Golden Cheetah give the OP a starting estimate for their 60 minute power, just using the historic data already entered into it?

RChung

Yes, though it will depend on how well the curve is filled out with efforts that were truly maximal for each duration (or, at least, the durations that the GC algorithm uses for estimation). It's a clever algorithm but it's not clever enough to know when you're dogging it.

The three-parameter fit is often a little bit better than the two-parameter fit (you still have fill out the curve with maximal efforts). Either will tend to overestimate the power you can sustain for an hour-long TT but the three-parameter fit usually overestimates less. If you really really need to know, if you're training for something very specific, you'll probably want to do an actual hour-long maximal effort rather than use these shortcuts. However, for training purposes these estimates are usually okay -- that's because general training isn't a terribly demanding application for a power meter.

svtmike

A local coach told me to use maximum 45 minute NP as a proxy for testing during the parts of the year that I refuse to get on the trainer. I have a custom chart set up in TP for that and it works well enough for me. YMMV.

Heathpack 

NP is not always higher than average power. In the perfect TT, your NP would be equal to your average power, in which case your VI (variability index, or NP/AP) is 1.0. I'm a TT newby, have ridden exactly one TT in which my VI was 1.005. So riding a NP = AP is obviously not that difficult a task.

Im aware that this is not the major point of this thread but OP is trying to understand these power metrics, so it does make sense to try to explain it correctly.

Heathpack 

I'm confused by your confusion, I'm not exactly sure what you don't understand.

Yes, on a "typical" road ride when you're intermittently coasting, stopping at lights, etc., your NP is higher than AP. But this doesn't tell you anything or indicate anything, it's just the nature of the metric. Basically, every little surge you have on a ride has an exponential physiologic impact, adding 5% more watts does not increase physiologic impact by 5%, rather it increases physiologic impact by some % which is exponentially higher. NP is just a metric that accounts for this, it puts more emphasis on watts produced when you are at a higher power output. So any ride with surges has an NP > AP. Smooth rides without surges have NP = AP.

If it helps you, I'm pretty confident my FTP is correct. It's based on a 40 min road FTP test and is supported by my on-going power data. I have a structured training plan put together by a coach.

I have no rides that last 30 min. My closest thing was a 20k TT, which took me 35 minutes. I rode it at an IF of 1.028. That was a pretty max effort, although in retrospect I think I could have gone a bit harder. My HR was supermax for the entire TT.

My typical difficult trainer workouts might last 1-2 hours (including 20 min warm up and 10 min cool down) and have an IF in the 0.85 to 0.90 range. Strong road rides of that duration are about the same.

A max effort ride of 1 hour would mean you were riding at your FTP for 1 hour and IF would by definition be 1.0. Any 1 hour ride with an IF > 1.0 therefore means your FTP is too low. A ride shorter than 60 minutes could in theory have an IF > 1.0. But even then rides with an IF more than 1.02 or 1.03 would probably seem like you were putting out a really strong effort, the kind of ride that feels really memorable. So an IF of 1.2 for a ride that did not feel extraordinary is implausibly high, and it suggests your FTP is too low.

RChung

You guys are getting pretty far down into the weeds.

Because of details in the way it's constructed, over short periods of time (like, maybe under 10 minutes or so) NP can be less than average power. This is an artifact of the way it's defined so Coggan doesn't recommend that you pay much attention to NP over intervals shorter than maybe 20 minutes.

NP is also kind of an empirical rule of thumb to account for varying physiological demands. So it's possible to have an NP > FTP over an hour and not mean that your FTP (which is itself kind of an empirical rule of thumb) is off. But it shouldn't be way over. Andy used to have a standing request for people to send him "NP busters" which were data files where NP over an hour was greater than 5% higher than a well- and reliably-estimated FTP. To put it another way, he was saying that NP for a "maximal" one-hour effort should be within 95-105% of FTP.

Heathpack 

Lol, as if you are someone who believes understanding the details is unimportant.