Lower Variability Index for Greater Speed 3

How variability index can help you get more

speed out of your cycling power

Your variability index (VI) is a measure of how consistent one’s power output is during cycling.  VI is simply calculated by dividing normalized power (NP) by average power (AP).  Average power is proportional to speed while normalized power is a close representation of how physiological fatiguing the ride was. NP is a weighted average calculation that follows the 4th power, similar in shape to a lactate curve.  If one performs a lot surges during their rides as opposed to producing a steady power output, this will be more fatiguing and will be represented by a high VI due to a high NP in relation to the AP. 

In general a high variability index is an indication of poor pacing and power application. However, there may be instances where performing a number of surges may be necessary. Tactically, it may be necessary to cover breaks, bridge gaps, launch attacks in bike races and draft legal triathlons. The pros in Kona employ a significant amount of tactical racing that may not be optimal for pacing but necessary to gain and maintain position.  However, the human body likes steady efforts.  As the intensity increases, the physiologic stress increases exponentially.  

This is best illustrated with the following example.   Let’s take an athlete with a FTP of 250W: 

  • ~ Workout #1, the athlete rides for an hour at 200W, a slightly difficult ride for this athlete. 

  • ~ Workout #2, the same athlete alternates their power output every 5 minutes between pedaling easy at 125W and hammering at 275W . 

Both rides have an average power (AP) or 200W (80% of their FTP).  The first ride also has a normalized power (NP) of 200W with a VI of 1.00 (200NP/200AP). The second ride has an NP of 234W with a VI of 1.17 (234NP/200AP).  A NP of 234W is 94% of the athlete’s FTP making it a much, much more difficult ride – if they could even complete it! 

Had the second ride been a race scenario, the athlete would have effectively given away 34W due to the average output of this ride being only 200W with the physiological cost of riding at 234W. Had the athlete produced the 234W at a low variability index, they would have ridden faster without any additional fatigue. A savings of 34W is more than a carbon TT frame, race wheels and aerohelmet combined!

 So how does one determine what their variability index was for a given ride? First, you need to access your uploaded workout within Training Peaks. Then click the “Map & Graph” button.


Summary Page


Regardless of which view is being displayed, upon accessing the data, the variability index for the ride can be found on the right-hand panel. If you wanted to determine the VI (or any of your metrics) for a specific portion of the ride, in the Graph view you can simply highlight the section of interest, and the metrics in the right-hand panel will display the results for only the highlight section.


low variability index

 In the graph above we are showing a ride with a variability index of 1.00. Here the athlete rode with an extremely consistent power. This is reflected by a steady cadence, power and heart rate profile. Contrast this with the ride below.

 high variability index

Here, the same athlete was out on a group ride with friends exhibiting much less control. At a quick glance it is obvious this ride is not as clean, and the variability index for this ride turns out to be 1.17! The difference in the two rides is easy to observe by looking at how often the athlete’s power and heart rate spikes.


Here are some tips for learing to ride with a low variability index:

  •  On your cycling computer, enable 3 second data averaging. This will help eliminate some of the variability in the data streaming from the power meter.

  • Try to keep the 3 second average within 10 watts of your target power at all times.

  • Pay attention to the amount of pressure you are applying to the pedals at your target power. As soon as this pressure begins to increase or decrease, it is time to either shift or adjust your cadence. With practice and focus, you will soon be able to determine when your power changes before it registers on your power meter.

  • Shift often and don’t be afraid of using your small ring… We want smart riders with no big egos!

  • If you find your power often spikes or dips when climbing or descending, it might be time to reevaluate the gearing on your bike.

Each individual athlete has an ideal variability index on a particular race course.  For most athletes racing long course triathlon, this equates to a VI of 1.00 – 1.03 on a flat course and 1.04 – 1.07 on rolling/hilly course.  Your variability index should be closely watched, managed, and employed both in your training and race day pacing strategy. 

Dave Sek and Jeffrey Capobianco

23 October 2013


Leave a comment

3 thoughts on “Lower Variability Index for Greater Speed

  • Steve Seide

    Small ring question: when I ride in the smallest cassette 2-3 gears [biggest gears] I am rubbing on something/hearing a noise. Seems to be the case with both of my bikes. Isn’t riding in the large ring but lowest gears the same outcome? If you were experiencing what I described, what would you do? Thanks

    • Dave Sek
      Dave Sek Post author

      Riding in the big chainring with and in the easiest gears on the cassette you can generate the same or similar gear ratio as riding in the small chainring in the middle of the cassette. Since this is happening on both bicycles, I don’t believe it’s a mechanical issue, but what you’re likely hearing is the chain running along the inside of front derailer as a result of “cross chaining”. If you had the chain crossed the other way – small chainring with small cog on your cassette – you can also run into issues with the chain catching the teeth of the big ring.

      Either way, crossing your chain places an increased amount of strain on your components, resulting in faster wear. You should avoid any prolonged (>30 sec) riding like this. Your drivetrain will be most efficient when the chain runs parallel to the bike. This means using the smaller half of the cassette in your big ring and larger half of your cassette in the small ring.