Tips‎ > ‎


Thoughts and explanations will appear here on the subjects of:
  • Why and how to use power - Coming soon
  • Power and pacing strategies - Coming soon
  • Effect of uneven power strategies  - Coming soon
  • Normalised power
  • Training load and stress scores - Coming soon
  • Tools to interpret data
  • Ironman power as a function of FTP

Normalised Power
Your power meter or Power Analysis software will often supply both Average Power (AP) and Normalised Power (NP).
Each may be sub-divided into "with zeros" and "without zeros", indicating whether time spent coasting contributes to the calculation.
Normalised power is a measure of the equivalent effort, had the rider applied the same power for the whole race, ie the power that the rider ought to have been capable of sustaining for the given period of time.

Normalized power is typically taken as the 4th root of the average of the 4th power of the power.
As an example, let's look at one of my typical Computrainer interval workouts:
    2 minutes @ 400W
    1 minute recovery @ 200W
    Repeat 8-12 times.

The Average Power over the 3 minutes (and multiples thereof) is (400+400+200)/3 = 333W.
The Normalised Power is  ( ( 400^4 + 400^4 + 200^4) / 3 ) ^1/4 = 364W.

Thus, although the average is 333W, the effort required is equivalent to maintaining 364W the whole time.

What if I were to stop completely for each of the 1 minute recoveries?
    Average Power becomes 267W (drops by 66W).
    Normalised Power becomes 361W (
drops by 3W).
Interestingly, this demonstrates that the workout wouldn't actually be much easier despite the much lower average power.
Furthermore, it shows how the effective workload will be highly influenced by peak power.
If there were no hills then even pacing is a good strategy, the necessary power adjustments for wind are almost negligible, but for hills there are measurable and worthwhile.
Understanding these concepts is key to effective pacing strategy..

Below is my Power and heart rate data from the NJ State 40k Time-Trial held in May 2011. Average speed was 27.5mph.

Of note is that the Normalized Power (367W) was only 1 watt more than the Average Power (366W). This demonstrates that the power was distributed fairly evenly. If the applied power was constant then NP and AP would be the same, but otherwise NP will always exceed AP. I consider this to have been a well-paced race (note also the almost constant increase in HR, it doesn't drop off, which would typically be accompanied by a loss in power late in the race).

Ride Analysis Software Tools
  • WKO+  from Training Peaks. The best tool, typically $129.
  • Golden Cheetah - The best free tool. Doug is evaluating the aerolab feature with a view to helping clients with powermeters to optimize their aerodynamics without having to visit a wind tunnel.
  • Sporttracks - Powerful shareware tool with Power analysis tools available for extra $$$. This is what I use.
  • Garmin Training Center - Free, not at all powerful. Best avoided except for uploading certain file formats to your Garmin.

Suggested Ironman Power as a function of FTP

Joe Friel suggests that a Total Stress Score (TSS, ie [T_hrs*(%FTP)^2]) of 300 is an absolute maximum for the Ironman bike if you're going to run well, and that we should stay slightly below this.
I created the following graph showing suggested FTP for various values of TSS based on bike split. From this, if we know the anticipated bike time, and accurate FTP (ideally based on the same calibrated power meter used in the race) then we can set a goal for Normalised power within the race.
That goal will be:
Target NP=FTP * sqrt ( Target_TSS / (Time_in_hrs * 100))

Note that the numbers 268 to 300 are TSS, not Watts.