Analysis of Real-World Marathon Power Meter Data from a Top 10 New York City Marathon Finisher

Patrick Smyth NYC Marathon Running Power Dashboard

Adapted with written permission of from the article “Understanding a Marathoner’s Running Power via Stryd Data”.

Patrick Smyth is an elite American marathoner who placed 10th in the New York City Marathon on Nov. 6, 2016 with a finish time of 2:16:34. During the race, he wore a Stryd power meter foot pod which provided data that measured his power, pacing, and fatigue.

In this article, Smyth’s coach Ryan Bolton—the founder and head coach of Bolton Endurance Sports Training (B.E.S.T.) and The Harambee Project elite training group in Santa Fe, N.M.—offers his analysis of Smyth’s power data.

Patrick Smyth NYC Marathon Running Power Dashboard

Patrick Smyth’s Power Data

Pace (blue), Power (yellow), and Elevation (purple) are shown above in Stryd’s analytic platform Power Center. Overall, Smyth ran a very steady pace and tried to avoid the mass accelerations that are often seen on New York’s rolling course profile, especially from miles 16-19.

Patrick Smyth NYC Marathon Running Power Dashboard

Time in Power Zones

The chart above shows how much time Smyth spend at different power ranges as measured in watts. Nearly all of Smyth’s race was between 280 and 350 watts.

Smyth’s average power for the entire race was 306.58 watts. The first half average was 319 watts, which is 8% higher than the second half average of 294 watts. Total Power is a measure of how intensely he races. In the Stryd file, the most notable drops in total power came early in the race on the downhill section of the Verrazano-Narrows Bridge (4:50 through 10:20) and then later on the downhill of the Queensboro Bridge (1:17:10 through 1:21). The first drop in power was caused both by a slowing of his pace and because the downhill required less power to maintain the pace. The drop in power on the Queensboro Bridge came only from the easier downhill running; his pace was steady.

This suggests that Smyth could have run both sections faster and maintained power with not much more additional effort.

As will be seen in the next image, Smyth’s Form Power actually increased during both of those sections due to the increase in the downward force and his less efficient running form on these sections. (He might consider some downhill running form drills to improve his technique and efficiency.)

As the race progressed—particularly from 1:20 and later—Smyth’s power decreased in direct proportion to a decrease in running pace. This is expected since running slower on flat terrain requires less energy.

If Smyth has increased his pace in the second half of the race, his total power would have actually increased as the race wore on.

Patrick Smyth NYC Marathon Form Power and Total Power

Form Power in Relation to Total Power

 Less efficient runners have a higher Form Power and a higher ratio of Form Power to Total Power. Basically, as run form falls apart, your Form Power rises. Smyth has a very efficient running technique, so his Form Power numbers are relatively good, as you can see in the chart above. His average Form Power was 65 watts, but it rose 2.2% throughout the race; Smyth got less efficient as he fatigued. For amateurs, this increase will be much higher. What we see in Smyth’s power data is that he did slow slightly in the second half of the race, yet he also didn’t let his form fall apart.

It’s most useful to express Form Power as a fraction of Total Power. In the early parts of the race, Smyth is efficient and his Form Power is about 20% of Total Power. As he fatigues, his Form Power nears 30% of Total Power, which shows a decrease in efficiency in the second half of the race.

Patrick Smyth NYC Marathon Ground Time Vertical Oscillation

Vertical Oscillation

Vertical oscillation is a measurement of how much an athlete bounces vertically while running. Normal values range from 8-14 cm. Smyth’s average was 8.1 cm during the marathon, which means he’s a smooth runner who doesn’t bounce much. Being this efficient as a marathoner is very helpful. Some oscillation is, of course, a necessary part of running. Vertical oscillation over 15 cm is too much bouncing and under 7 cm is not good, either.

Ground time measures how long the foot is in contact with the ground from foot strike to toe off. Typical values range from 150 to 300 milliseconds. This number varies with running speed and also running efficiency. Shorter ground contact times are associated with a more efficient running stride and the energy savings can really add up over the course of a marathon or half-marathon.

Take a look at the chart above and you’ll see air time (yellow), ground time (orange), and both (blue). What typically happens with more fatigue is that ground time goes up while air time goes down.

Watching the yellow and orange lines diverge is a great indicator of the levels of fatigue in most runners. Smyth’s average ground time  for the race was 175 ms, which is at the low and more efficient end of the range—just what we’d expect from a runner of his high caliber. As the race wore on, Smyth’s ground time increased with fatigue. Yet his longest ground time was about 240 ms, which is still well within the normal ranges. Smyth is an efficient runner! His ground time rose 8.4 % during the second half and this is likely due mostly to the decreasing speeds and pace of the race.

Patrick Smyth NYC Marathon Running Power Chart

Power Data Comparison

 Without knowing Smyth’s race finish time, a lot can be concluded from looking at this chart above.

  • Power fell during the race, mostly because the race slowed down.
  • Form Power was consistent, showing that Smyth maintained his running form despite fatigue.
  • Ground Time rose during the race. A normal range is 150-300 ms. Smyth’s race was well within the normal range, despite fatigue, which reveals that he is an efficient runner with good endurance that helps him maintain form throughout a race.
  • Leg Stiffness is a measure of how much energy a runner can recycle with each stride. Higher Leg Stiffness is generally better than lower Leg Stiffness. Smyth’s Leg Stiffness fell during the race, revealing fatigue.
  • His Cadence fell slightly from 184 steps per minute to 181.5 spm. A normal range is 180-200 spm.
  • Smyth’s Vertical Oscillation fell slightly during the race, meaning he became less bouncy. A normal range is 8-14 cm and Smyth was on the more efficient end around 8 cm.
  • His pace rose from 4:57 minutes per mile to 5:30 pace, an 10% decrease in speed. In this case, the overall race pace slowed.

The Big Question

 How can we use Smyth’s data to guide his training and racing going forward?

The most notable change in the data is what happened to his form during the second half of the race, especially from mile 16-26. His Form Power increased and rose in relation to his Total Power, which indicates a loss of efficiency. His Leg Stiffness, Ground Time, and Vertical Oscillation values agree with this assessment. The good news is that all these parameters are related. When we work on one, we can improve the others.

We can use Stryd data from future workouts to see if Smyth is improving. One key way that Smyth might improve is to monitor Ground Time in real time on his watch. Thinking about having “quick feet” should lower ground time and also improve Leg Stiffness and Vertical Oscillation.

He might consider attacking the downhills of future races since upping the pace during a downhill is relatively easy. He might also add some downhill running form drills to smooth out his stride on downhill sections.

Find out your own key power numbers in Jim Vance’s complete guide, RUN WITH POWER.

Run with Power by Jim Vance

RUN WITH POWER is the groundbreaking guide you need to tap the true potential of your running power meter like SHFTStryd, or RPM2. From 5K to ultramarathon, a power meter can make you faster—but only if you know how to use it. Just viewing your numbers is not enough; you can only become a faster, stronger, more efficient runner when you know what your key numbers mean for your workouts, races, and your season-long training. In Run with Power, TrainingBible coach Jim Vance offers the comprehensive guide you need to find the speed you want.


1 thought on “Analysis of Real-World Marathon Power Meter Data from a Top 10 New York City Marathon Finisher

  1. I think Ground Contact Time on Garmin watches refers to the amount of time in contact with the ground per stride. So increasing cadence may reduce ground contact time per stride but not ground contact time per distance covered, as it normally results in shorter strides for the same effort. So more strides are needed to get the same distance. A better measure might be ground contact time divided by the total time per stride. A similar correction is already being made to Vertical Oscillation per stride by dividing by the stride length, giving the Vertical Ratio. Ground Contact Time Ratio would be a useful addition.

Comments are closed.