This is a key misunderstanding. I think your FTP is set too low because your NP was 1.2 for 30 minutes, not because it was higher than AP.

You're right. NP is not always higher... it's just always higher in my history of tracking power. And apparently it has always been higher in your experience too. But it can be less than AP in unusual conditions such as a steady power level with sudden/short spikes in power... your AP will include these power spikes while NP will smooth them out. For all intents & purposes I'd say it's safe to say that NP > AP 95+% of the time and in this case the OP seems to not understand this aspect of power tracking and thinks that NP being larger than AP is somehow an error that must be taken into account for the high IF.

If you're going to think about heart rate you need to define zones based on more relevant physiological states, almost always lactate threshold heart rate at the high end and perhaps aerobic threshold at the bottom. Both are independent of resting and maximum heart rates.

LTHR varies across populations from about 80 to 95% of maximum. It's influenced some by training - mine was 163-164 before I did a lot of riding below my aerobic threshold and 168 after out of 184 maximum, not changing with resting heart rates varying from 45 through about 80.

My aerobic threshold has varied from somewhere in Friel's Z1 through Z2 although I haven't tried to measure precisely. It increased with plenty of riding below my aerobic threshold but not at any sort of fixed relationship with my other heart rates.

FTP is approximated as 95% of the average over 20 minutes to exhaustion, usually with a preceding 5-minute maximum effort to deplete your anaerobic reserves. You can also use NP over a 100% 60 minute effort, although that's unlikely to happen outside a racing environment.

For me, NP is always higher than average power. I live in the first residential neighborhood outside of downtown. Lots of traffic, lights, etc.

Problem: My IF is too high, I know because my NP > FTP by too much. Solution: Set FTP from max 20-min avg power. Result: My IF will be high, because my NP > FTP.

I'm confused because the fix here is just going to lead me back to square one. If I can recognize a problem based on my normalized power and then I fix it by applying average power, then I test my fix using normalized power ... it's going to lead me back to where I am now, just with slightly lower IF scores. I'm kind of having a hard time believing that you don't pick out the highest NP (instead of avg P) to substitute for an FTP test, that would be apples to apples. But I don't have software that can do that, or if Golden Cheetah can I haven't figured it out yet.

Not at all. What I don't understand is how/why you'd use AP to estimate FTP, if the reason you think you need to re-estimate FTP is because your IF is too high (which means NP, not AP, is higher than FTP).

Yes, IF is based on NP so I can understand why you're confused when you're told that you should retake an FTP test (which is based on AP). But my posts are based on the sparse data that you've produced. On the ride where your FTP was 1.2 what was your AP, and what is your FTP? I'd be willing to bet that your AP is greater than FTP/0.95.

Right, for you and for most people on most rides, NP is typically higher than AP. But that is not the same thing as saying universally that NP is greater than average power, because its not. Minor point, though, no need to get caught up in this.

No one is really saying your FTP is too low because your NP is greater than your FTP in a ride. People have been saying that your IF of 1.2 is implausible for your perceived effort, it does not match up with your HR data on that ride. The most likely explanation is that your FTP is too low.

In the TT example I mentioned previously, my NP for the race was higher than my FTP. That's expected/ok, because its an almost-max-effort for under an hour. But still, even though my NP > FTP, my IF was only 1.028. My FTP is as correct as you can get these things.

My max 20 min power hovers within 5 watts of my FTP pretty consistenly.

???

It really shouldn't matter if you use NP or AP for an FTP test. You're going to ride it like a TT anyway, the numbers should be pretty identical. For example, I (newby) am told to ride a 40 min FTP test, something I've never done before. I'm told to ride it at a max/constant effort. My NP comes out to be 1 watt higher than AP. The numbers will be effectively the same for an FTP or a TT or any steady/smooth effort.

Experience. What it feels like, perhaps watching your heart rate farther into the test because LTHR doesn't vary much and takes a long time to change.

You can also set zones off critical power calculated off long (10-45 minutes) and short (2-5 minutes) efforts to exhaustion. Try for 4 minutes, under-estimate so you fail at 3, or over-estimate so you fail at 5 and you still have valid results.

As RChung notes that over-estimates maximum sustainable power for longer durations, although an hour or few out that doesn't matter for training purposes because those power levels are too low to improve your lactate threshold/VO2max and too high to benefit your aerobic base.

If the FTP will be effectively the same whether you use AP or NP then why argue the point? Find me a source where it suggests that you use NP for your FTP test. It's also much simpler to calculate your best 20 minute AP within a FTP test ride, which is longer than 20 minutes typically, then NP. I don't understand your confusion. Are you seriously suggesting that you don't think that people typically use their AP for an FTP test?

I'm saying it effectively doesn't matter whether you use AP or NP for an FTP test and therefore its not confusing. Because to ride an FTP test properly (or at least how I've always done it/been instructed to do it), you are riding at a max steady effort.

Your comment to OP was that you could see why it was confusing that something to do with NP would make people say he should ride another test if an FTP test is based on AP.

I'm just saying that's not confusing at all, because NP and AP for an FTP test should effectively be the same number.

You seem intent on arguing semantics & rhetoric. The bottom line is that every source I've read states that you use AP to calculate FTP. Period. When I did my last FTP test it was a 40 minute ride. Explain to me how I would use the data from that ride to derive the 20 minute NP interval. Your point is that it shouldn't matter because NP should be the same, or close to AP. Explain how I would calculate that in order to know that is correct for that ride.

This is also part of what confused me. I'm not going to ride an FTP test, I'm going to try to find my FTP using recent data. So it won't be a TT effort and my AP and NP will differ.

I understand that NP can equal AP, even if for my riding it never will. I'm trying to ignore the software developer in me and only pay attention to how this stuff applies to me.

I also understand that it's entirely possible for my NP to be higher than my FTP, this can be a routine thing. I can do a short but really intense ride, or I can do a medium length ride with a lot of hills or some sprints, where I work hard but also have time to recover.

I'm not sure exactly what IF tells me that I need a new test, but people seem to agree that my numbers are suspect. Maybe I'm making too many assumptions about what a red flag is.

I didn't ride last night, I did a short run (need to have my head checked) and then a walk with my girlfriend. But here are some numbers from my last handful of rides.

7/15: IF = 0.972, TSS = 134.8, TE = 2.5
7/14: IF = 0.866, TSS = 76.5, TE = 2.3
7/13: IF = 1.069, TSS = 84.8, TE = 3.8
7/12: IF = 1.043, TSS = 136.4, TE = 3.7
7/12: IF = 0.577, TSS = 1.7 (errands, didn't wear HRM)
7/9: IF = 1.205, TSS = 56, TE = 4.2
7/8: IF = 1.002, TSS = 79.4, TE = 3.5
7/7: IF = 1.064, TSS = 73.2, TE = 3.1

What about the AP on those rides?

I'll have to go back and collect that, will post shortly. I've got a spreadsheet but it isn't tracking average power. I remember from looking at the details in each ride though that average power is always much lower than normalized for me.

[interruption] It's really not that hard, you know. [/interruption]

Is it as easy as looking at my data in Garmin Connect where it says in the Power section "Max Avg Power (20 min)"? I'm sure it can be calculated in Golden Cheetah, but I don't know how, so it was an honest question.

btw RChung, do you suggest a cyclist use NP or AP to calculate FTP?

All people are saying is that from experience a ride with an IF of 1.2 is a very challenging ride- not really an everyday workout kind of challenging but an extraordinary type of effort. You report that your ride was not that challenging and that your HR was not that high. Therefore your IF probably wasn't really 1.2, it was probably something less.

IF = NP/FTP

If people are saying your IF was likely too high, then either NP was too high (unlikely since this is being calculated by your software) or FTP is too low.

That is all. You are bringing all kinds of things into this that don't pertain, its a very simple concept.

Golden Cheetah automatically calculates xP (which is a variant of NP). I don't remember if it automatically calculates max 20 minute anything -- it doesn't usually think there's anything special about 20 minutes as opposed to 19 or 21 -- but you can configure it to do so if you want. I suppose you can also configure it to calculate NP rather than xP but I've never done that. Before Golden Cheetah existed, before WKO+ existed, I used to calculate NP over arbitrary intervals so I just continue to use the stuff I wrote back then.

When friends absolutely need to know their FTP I tell them to use their AP over a one-hour TT-like effort. When they don't absolutely need to know it, I tell them to ballpark it with Golden Cheetah's CP.

However, if you did ride a one-hour TT-like effort your AP would be within a watt or two of your NP, which is well within the margin of error for an FTP test, so really it wouldn't matter which you used.

Right? ;)

Either I misspoke, or I provided too much irrelevant data and confused things.

Only one ride since I got my PM set up properly had an IF around 1.2, that was a hill repeats ride last week. That was a pretty hard workout. I'd put it around 9/10 for RPE. I was asking about a different ride that had a high IF (but not nearly so high) and a surprisingly low (for the IF) TE, and whether that meant it was time to re-assess my FTP. But that wasn't for the IF = 1.2 ride. The IF = 1.2 ride was almost a half hour, several sprints up the hill, after each one I'd wait for a break in traffic, do a U-turn, ride down the hill, hit lap at the bottom, and sprint back up.

Looking at your list of rides, you have a lot of IFs greater than 1. My personal experience (as I've posted upthread) is that tough 1-2 hr workouts/rides have an IF in the 0.8 - 0.9 range. I train regularly and very hard (albeit possibly with less intensity than some people because of the high volume/endurance component of the type of cycling I do) and its very rare that I have an IF >0.95 for a data file. This is over about 8 months of data.

So its not just that one ride that make me wonder if your FTP is too low, its the entirety of what you've been posting.

Well, I'm not sure I'd say "a watt or two." I'd say a few watts. It depends on the course. The last time I did a one hour TT, it was on a rolling course with a little bit of head, tail, and cross wind. The only time a constant power output is time-minimizing is when the conditions are constant. When conditions aren't constant then the fastest way around a course is not constant power but varying power. Not too varying though. A friend recently did an hour on the velodrome in an attempt to break the old hour record. Those conditions were pretty constant so power output was pretty constant (extremely constant) and NP was almost exactly the same as AP (especially when you ignore the first couple of laps at the start).

For the IF = 1.2 ride, AP was 85 % of NP. For the most recent one I did, AP was also 85 % of NP and it was 82 % for the ride before that.