My 0.02$ is pride is preventing a lot of people to make good choices. The allure of having that slightly higher FTP number is too enticing!
If one pays attention to the workout text prior to the workout, and more importantly the text prompt during the workout and act accordingly, a lightly elevated FTP would no be a problem.
It’s all there “if you feel XYZ increase/decrease intensity etc”
I’ll throw in my personal FTP test
I generally hate testing, so I invented my own protocol.
The Bandit Protocol.
There can never be enough protocols 
First, I personally believe this testing is best done outside. On the bike you race, with the power meter you race with. Make it valid to your actual real world performance.
Be very careful with temperature. If it’s hot, as you overheat your performance will suffer. You could use that result as your ‘heat FTP’. Essentially, what it would be at that temperature. Personally, I prefer it cool. The reduced cooling during climbing means the temperature needs to be a lot colder than many think. Best performance would be in the 13-16c degree range. Even colder on a MTB with the reduced speed.
Whatever the temp, your result is relative to that temperature. So, if you live somewhere hot, or only race in the heat. No problem. You’ll just test a little lower than you’re ultimately capable of.
Find a steady low grade hill. Your favorite one. The angle you enjoy the most. It will need to be at least 3mins of steady gradient. 4mins is better, 5mins probably optimal, although for 5mins, less fit riders will likely fatigue too rapidly.
I like 3mins myself, that’s plenty to get me where I need to be.
Wear your HR monitor, if you have one.
Take a guess at your FTP. I’m sure anybody reading this will have a pretty good idea.
Simply ride repeats. Very steady each time. I display 3sec power, 30sec power and HR. Or just do laps with average power. However, if you do use laps on your computer, remember, going 50w harder for 30sec at the end to get the target number will ruin the data. We’re interested in the power and the HR.
Warm up for a considerable time. Maybe, a full hour. Z2, bring the power up very slowly to tempo for a few minutes when you’re close to the hill. Fuel and hydrate correctly.
Simply start your first effort about 20 or 30w under what you think is your FTP is.
Ride the hill, super steady, watch your HR carefully. Take a mental note of what it peaked at. Take a very diagnostic approach. Feel your breathing, feel your pedaling technique. Keep your ideal cadence. Just nice smooth power.
Roll down the hill. No hurry at all. Relax. Take a drink, have a candy. Chill. Give it a minute or two at the bottom to rest, then go again. 10w higher power.
Repeat.
You will soon get to the point where you can feel a kind of sustainable high power. Power that you could maintain for 30-45mins, for real. It’ll likely be 10w less than your ego tells you. The repeats beyond that don’t last long. I usually do two above my FTP. The 2nd, which is 20w over it, is very obviously over it. You’ll feel it rapidly and see it in the HR response.
Pay very specific attention to the HR. Knowing your approximate threshold HR is an incredibly valuable metric.
Take your new found approximate FTP and approximate and threshold HR and experiment with training. Try some longer intervals. You’ll learn rapidly if you need to tweak it up or down a tad.
I suspect most people will get within 10w with this technique. Particularly, if you don’t let your ego have any say in the matter.
The bonus of this technique, you won’t be very tired when you’re finished. So, you can add in a nice long Z2 ride or whatever you want after completion.
I’m a small 60kg rider so 10w increments work for me. I’m also very familiar with my performance and perceived effort. It usually only takes 4 or 5 repeats to get it sorted. I know it’s not 300w and it’s not 200w. So, I’m starting at 230w, to be conservative and feeling it sits somewhere between 255w-265w.
When I do the 270w repeat, I’m starting to feel less in control, I’m certain I’ve gone past it when I do the 280w repeat.
That works for me. It’s almost fun, a billion times less painful than a ramp test and for me, vastly more accurate. For interests sake the ramp test gave me 277w . Well overestimated.
For bigger or very strong riders, it could be 20w steps. To be fair, once you’ve completed this test or have other data, a 5 repeat step covering 50w should be enough to cover every rider.
If you repeat the protocol, do it at the same time. Say, after a rest day. Same amount of caffeine, same fueling etc. Ignore Strava segment times. They are totally irrelevant. Too many variables. Don’t even look at them.
Finally, be conservative. It’ll likely be the step under the one you want it to be. Equally, it might be in between a step. Now we’re talking about 5w. I’m sorry to say that it just isn’t that accurate. Nor, does it need to be.
This is not an all out protocol, like a ramp test, you will feel like you’ve barley gone hard. This is why you’ll think it’s the step above the actual number. If it is, fantastic. You’ll find out soon enough.
Pick the honest repeat. Train off it. Adjust it.
Try it again after a training block. Hopefully, it’s up.
Enjoy 
You approach is similar to what pros do. Once you know what it feels like it’s easy to get it close enough for training. And everything else is probably over-precision.
How pro cyclists test their fitness, with UAE Team Emirates – Rouleur (as posted in the Pro/Elite thread)
Great article.
Totally agree on the over-precision. People think they’re launching a life critical mission to Mars. It really isn’t that complex.
I will say. I don’t know what the ramp test is supposed to be testing, cause it sure ain’t testing FTP. Taking a fixed percentage off the final minute of a different test, seems like a fantastic way to pile error on top of error.
It’s also a fairly brutal test, simply to work out FTP. Particularly, for anaerobic sprinter types.
As many of us have said, the ramp test is designed to estimate map, and it does a pretty good job of it. But any training stimulus designed to change the slope of your power duration curve will then make estimating a single constant percentage of map less valid.
Short power, sustained power are both designed to change your pd curve. Just like kolie mentioned in the podcast, short duration repeats might also be a strong anaerobic stimulus and can allow for an increase in map, but an ftp that did not change or maybe got lower. This has been my experience in the past doing the tr designed build and specialty plans. My 2x8 results never increased and i though i was doing something wrong during the test because they actually got harder and the delta between the two increased. Makes sense now though.
I probably have far more modest goals than many here and don’t like getting too far in the weeds trying to sort out whether my ftp is too high or low by 5-10 watts or more. If my workouts feel sufficiently hard I assume I’m at least in the ballpark with my ftp
I think @Nate_Pearson has said it multiple times here in the forum or on the podcast. The ramp test is designed to set appropriate ranges for the TR workout and plans. Call it ftp, ramp test result, or something else. Based on their data, this test and the result gave the majority (not all!) of people a score that worked well for creating the right zones.
But he also recognized that it wasn’t perfect and some people over tested or under tested. Hence they work on AT. As he says, ramp test + AT = best results. They aren’t in isolation. If you way over or under test it make take AT a little longer to get the right workouts for you.
Now comes predicted FTP which removes the ramp test. It’s all an evolution. And as many people on the forum have said most of us here aren’t professionals. Consistency i and a plan are the most important things. Not getting an exact ftp number.
Are you trying to tell me that if my “true” ftp number is 1 watt off I can still have workouts that help me progress? That doesn’t seem right. lol
TR and cycling in general is heaven for people who love to look at data.
And every day we wonder through the forests to kneel by the lakeside, staring into the water to love our reflections data hoping the reflections data will love us as much as we love them.
Yeah, I think the ramp test protocol can probably get you in the ballpark, and is probably most useful for riders who don’t have a ton of experience with pacing longer efforts. As a coach, I could see some validity in doing something like a ramp test, then prescribing longer “testing” intervals at a lower power to see where the rider is actually at. The ramp test isn’t going to put someone in a big training hole, so if I’m taking a new rider off the couch, it’s not the worst idea in the world to do that kind of thing and then validate it like you said.
But then after a few weeks of training, I have no use for a ramp test.
The only addition to this would be that AT was not in response to the ramp test. AT was in response to all testing protocols and just athletes in general.
I think this is a better way to think of it in non TR branded terms:
That is our approach 
.
Just discussing #1 and saying that is TR’s approach is disingenuous in my opinion.
But that doesn’t mean we can’t improve. Here are some things we need to do to get better:
- Get your MAP
- Find your relationship from MAP to threshold/VO2/Anaerobic/endurance through intervals/workouts you can consistently do.
We could spend a lot more time on #1 in that list and put out many different ways to test your FTP at various complexities, but no matter what we’d still do #2 to get better training.
For those who don’t like their MAP result from us and we put you at 230 when you should really be at 220 you can still put your FTP at 220.
But for us, we might score you at 230 then have you do FTP repeats at 96% and if you’re at 220 FTP we’d have you do them at 100%.
Either way, you’re doing your threshold repeats at 220-221 watts. You end up in the exact same place.
And then to really mess with your brain, your power meter is probably +/- 1.5% accurate and you might have a swing of 6 watts in there…and then there’s heat change in the workout. They have temperature sensors in there to help but they tell you to recalibrate at different temps for a reason.
And then there can be slope issues where how accurate it is changes based on how many watts you put out!
Either way, #2 above helps us overcome those problems.
TLDR; We’re trying to solve the problem of how to make you faster in a much more nuanced way than just a different testing protocol. The adaptive part of AT is where the truly powerful personalization comes into play, and it’s only going to get better.
Enjoyed reading about your protocol. Apologies if I missed one thing in your terse description of the protocol - you may want to consider Dr Coggan’s latest tweet the next time you update the protocol 
Coggan’s latest tweet the next time you update the protocol
Still with the haughty quips. Sigh. Coggan just can’t stop himself.
How about an open letter that two ppl will read (but is true nonetheless).
Dear Andy, I can throw out unhelpfully arrogant sound bites about matters in which I’m well-versed too. I even know the perfect mindless platform upon which to do it. How about 2 watts, then? 67 watts? No? Why? Maybe you can help ppl understand. That’s more difficult and time-consuming than trolling us like it’s 2010 and you’re still trying to sell the idea.
You win. FTP is awesome. It’s valid. Thank you. Seriously. If you hear that enough times will you stop being an asshat about it?
Warms regards,
Tim
He is reliable LOL.
True, lol
I think his point is that threshold is a bounded area (not really bounded, but things start happening quickly within a range of a few(?) watts) and not an inflection point.Theres no one watt where things start happening. That’s not really how nature works, usually. If you watch Sebastian Weber, I believe he says estimating within 5-7 watts is the best you’ll do with lab equipment or field testing. Coggan even recommends rounding to the nearest 5 watts. I buy it.
His personality it kinda off-putting, but his science is pretty good.
To be fair,
He’s absolutely right.
How anybody thinks an FTP test can be 1W accurate is beyond me. You test, it reports you are up by 1W. I hate to be the bearer of bad news, but you could actually be down 7W. It really isn’t that precise, given the way we test and with the equipment most use.
I’ve never set my FTP in anything other than 10W steps. Less than that, you must REALLY want that increase
I know what his point is. I’ve been hearing it for years. I’ve even made the point from time to time myself.
My dad doesn’t keep harping on all the stuff he told me when I was a teenager. I’m 48 yrs old. We’re good now. Maybe Coggan doesn’t have kids 
Someone new to this stuff might NOT know. Who better to enlighten them?
Once again., I can condescend too. And just pipe any push back to dev/null. But what does it prove
(/s, of course 
)
I have this new idea - I’ll call it FCQ, short for ‘Functional Condescension Quotient’, and it measures your ability to be condescending in public discourse.
FCQ = (# of condescending posts in about the previous 30 days)/(total # of posts in about the previous 30 days)
This is tracked over time with CCL and ACL (Chronic and Accute Condescension load).
Most people can generate high ACL values, but it takes true dedication and likely genetics to sustain high CCL values.
I’m clearly not the only one that has noticed this pattern in his public comments…
I appreciate your constructive participation in these forum threads Nate 
This all makes sense and I think is very reasonable. I think what gets a lot of discussion going is how “FTP” is used as part of #2, as FTP has meaning and context outside of TR.
I think it is easily possible for the ramp test to be great for setting workout levels (and moreso with AT), but not a really good “FTP test”. Most of these discussions are about the ramp test as an FTP test, as nothing brings out opinions on the internet like FTP It’s the “FTP==0.75*MAP” bit that gets people going more than anything else I think. (oversimplifying equation for simplicity.)
If there wasn’t this somewhat agreed upon definition of FTP to argue about, I don’t think there would be all that much discussion.
