When I do my FTP testing during the summer I have a route that's not completely optimal. I guess the best thing would be a relatively flat course with no corners. Well, this is a course with rolling hills and some corners where I need to slow down quite a bit. So I was thinking whether or not it would be better to look at normalized power instead of average power for FTP?

_________________
-----------------------------
I like Spanish bikes and cars. Riding an Orbea Orca and a Seat Ibiza.

if the time at zero cadence is minimal, I'd use the average power without zeros. Other method would be using the trainer for a consistant method with your (20min test ?) or do a ramp test on the trainer, its shorter and a consistant protocol will give you a good idea of your progression.

_________________
2013 rocky mountain vertex 29er
­Twitter
www.cyclingtrainingnuts.blogspot.ca

Depends on the duration of your test and what software you use. You can always smooth the file a bit, but I think a better thing is to see how much NP and AP vary. Its supposed to be a quasi semi state test so if its like 3-5w I wouldn't worry a ton, but if its a big amount I'd find a new way to test.

i tested outside from march til about august (missed 1 or two months), but like you it wasn't super ideal in a sense where somewhat rolling, and 2 acute roundabouts where i would have to coast through them....1-3 seconds. You can see the elevation change in the strava link i posted.

but from those 20minute test , and using the adjusted FTP for future workouts, it seemed to work well.

i just had to remember to put on my 11t to pedal (and keep power up) downhills, and go into small enough gears while going up hill to maintain the same cadence but somewhat steady power. someone will probably come in here and tell me my protocol sucks, but ehhhhh i am not bothered to change it for outside and will probably continue on 2013 the same when testing outdoor =)

segment highlighted below...
http://app.strava.com/activities/23458419#z2386|3594" onclick="window.open(this.href);return false;" onclick="window.open(this.href);return false;

In theory you should use normalised power, since it calculates the equivalent constant power output you would have achieved if you had ridden on a perfect course.

But in reality I wouldn't get too hung up on using average power. In my view it's a useful measure since if you can do 300W round the test circuit with a few gaps in pedalling, then you can also achieve 300W on a typical ride on a similar circuit.

I would say use average power; FTP is a training metric first and you want it to reflect your ability over your typical training and racing terrain. Its nice to see the bigger number that np will likely provide but it will also likely make your workouts too difficult (assuming you are training on ap and on similar terrain as your test circuit).

_________________
"But when you hit that asphalt down below nothing else matters, your pulse is skyrocketing toward 200"
Speed Concept
Colnago Mapei
S-5 VWD

Thanks for the replies. But what if we look at it differently? Let's just asume that I've done an FTP test, and the NP and AP are different.

AP: 300 watts
NP: 315 watts

Now, I'm setting off on a 40 km TT, but if I aim to hit my AP level, my AP at the finish line will probably be lower than 300 watts when accounting for turns, coasting, braking etc. Whereas if I aim to hit the NP level, I'll probably cross the line with an AP of 300 watts. Situation 2 will get me quicker across the line?

Would this be correct to assume?

_________________
-----------------------------
I like Spanish bikes and cars. Riding an Orbea Orca and a Seat Ibiza.

just try one. see what happens...

If you're seeing that big of a gap during a FTP test - you're pacing is bad. Most FTP test/ITT should have AP and NP within a few watts of eachother.

Like Ghost said.

There's no point in a 20 min test where you should not be pedalling.

Thus your AP & NP should be pretty much the same.

_________________
Baum Ristretto, Baum Mondrian Corretto & the '89 Merckx! | "If you could eat it in one sitting, hold your line." | twitter.com/Tinea_Pedis

Or as the OP said, the course is less than ideal.
I simply dont have a course that is really suitable either and my VI is often 1.05-1.08 on a FTP test.
But in the end it doesnt matter, so long as the test is repeatable and you are getting a similar VI in each test, I wouldnt worry about it at all and would use NP.