2 FTP Questions, and 1 non-FTP question

A light bike doesn't replace good fitness.

Moderator: Moderator Team

Post Reply
Posts: 1
Joined: Mon May 18, 2015 6:50 am

by PsiSquared

Hi, I'm new to training with power, and I have a couple of questions re: FTP tests.

1. Flying start or standing start? I would think the standing start would not optimal since the Trainer Road folks say that avoiding surges is best. Of course, according to my wife, I'm rarely right.

2. Do I use average power or normalized power to calculate FTP?

3. Are there any Matlab users here? If so, does anyone have an m-file or similar for reading a .TCX file? I'm still using a Garmin 705.


by Weenie

Posts: 1
Joined: Thu Feb 15, 2018 2:42 pm

by Raia


since you should have been riding before starting your 20-minute-segment that will give you your FTP value (a good warmup with some short, but not exhausting efforts) just roll into it. Ride endurance pace, press the "lap" button on your headunit and then move up to a power you expect to be able to hold for 20 minutes. If there's still alot left in the tank for the last 5 minutes or so, lift a little further. The very first test often isn't exactly great, because you don't have a good idea about your pacing yet.

If you do the test well, normalized and average power for the 20-minute-segment should be pretty much the same. If your normalized power is much higher, this is an indicator of bad pacing and therefore a suboptimal test. Your average power value will be kind of a value for "at the least", while the normalized value is an artificially calculated value, which might work well or not.

This is why it's important to find a place where you can ride without interruptions (traffic lights, intersections) or big changes in gradient. This is the reason that sometimes 2x8 minute blocks are used to test - although there's always been alot of debate whether a good FTP-Test should last 60, 20 or 8 minutes. Shorter testing means easier to do properly (pacing, motivation to suffer) but nobody can guarantee you that with the corrections (95% or 90%) the results will give you your maximum power for 60 minutes accurately.

No matlab for me since university. Luckily. ;)

Posts: 6
Joined: Sat May 27, 2017 12:52 pm

by rawjunk

Just remember to do the 5min all out effort before the actual 20min test.

Post Reply
  • Similar Topics
    Last post