## 3.26.2012

### Power File Terms Part III

In Part I, I defined AP, NP, IF, VI, KJ and TSS.  In Part II, I showed how they change from ride to ride.  In Part III, it's time to dig a little deeper.

What I gave in Part I were simplified definitions.  My goal in Part III and Part IV is to expand on what I wrote in Part I about Normalized Power and TSS, but not get carried away and make your head spin...or my head spin.

So let's talk Normalized Power. It's a little tricky.

You've probably read definitions very similar to one I wrote in Part I before.  So you know NP is a better indicator of intensity and tells you the ride you just did was the equivalent of riding at ___ watts if you had ridden steady on flat terrain.  Let's dig a little into the why and how.

Average Power (AP) is pretty straightforward.  For AP, you take every data point from your ride, add them up, and divide by the number of data points. I think most people understand the concept of AP.  As I mentioned in my earlier posts, AP can be deceiving because two rides that were very different can have the same average power.

The problem is that the relationship between power and physiological stress is not linear, meaning 400w is not twice as stressful or difficult as 200w.  It's much, much more difficult and creates a lot more stress on the body.

Say you're riding at threshold, your one-hour power.  If you increase your effort by 10%, the physiological stress increases by a lot more than 10% and you will only be able to hold that effort for a fraction of the time you can hold threshold. Normalized Power gives those harder efforts more weight in the calculation to account for that stress and gives you a better idea of the intensity of your ride. Average Power gives all efforts equal weight and that's why it can be deceiving and isn't the best indicator of the intensity of the ride.

Another thing that's helpful to know about NP is that it's calculated off of a rolling 30-second average.  The reason for this is because the physiological reactions to higher intensities aren't instant.  They take a little time, so a 30s average was used.  Using 30 seconds wasn't random, it has to do with half-lives of physiological responses.  I don't know all the details; I just know it's not a random number and it's based on science and our physiological responses to exercise intensities.

So if you've ever done some short intervals and uploaded your data and got n/a for NP power for that effort, it most likely has to do with NP using a rolling 30-second average and your interval being too short for the calculation to work.

Sidenote:  Does your bike computer have the option to display a rolling 30-second average?  My Garmin does, and I can't say for sure they went with 30 seconds over other options because NP is calculated on a 30s average, but I bet it is.

So Normalized Power takes a rolling 30 second average and then everything is raised to the 4th power.  Again, not random. The 4th power was based on studies that compared blood lactate levels to power output.  Raising everything to the 4th power is how the larger numbers get more weight in the calculation.

Then everything is averaged, and then they take the 4th root to bring it back down to normal numbers.

So it looks like this:

- take a rolling 30-second average
- raise to the 4th power
- average everything
- take the 4th root

See?  It's a little tricky.  The main takeaway is that it's a weighted average giving harder efforts more weight, so in the end you have a number that gives you a better indicator of the intensity of your workout.

And because NP is a better indicator of intensity than AP, it's used in the calculation for Intensity Factor (IF).  That calculation is pretty simple.  It's just NP/FTP.  Notice it uses your FTP.  That's because it's taking your current fitness into account whereas NP is not.

NP is also used in the calculation for Training Stress Score, which I'll get into in my next post.