I Should Pick Up the Pace During Pickup Soccer

As chronicled here, I’ve been keeping track of my GPS movements over the last 11 months.  Recently, I decided to look at how fast I tend to move for any one of these outdoor activities.  By calculating my change in distance for the change of time between every two subsequent recorded data points, I determined my instantaneous pace of movement, and plotted distribution of these paces on this density plot:

 

Some interpretations / conclusions:

• I have a unimodal, but rather wide peak for running / jogging (red).  Admittedly, it’s rather slow (7.5 mph corresponds to 8 minutes a mile) but I’m running for consistent exercise, rather than speed anyway.
• I have numerous peaks for walking (blue).  I think this is telling me that rather than taking a continuous range of speeds, I favor taking certain discrete gaits (eg. “fast walk mode”, “medium walk mode”, “slow walk mode”).  Notably, I have a small blip around where my running pace is: that’s probably when I’m running to try to catch the bus during my commute.
• Somewhat expectedly, the distribution of speeds becomes more complex (more peaks) with a more complex activity like pickup soccer (green).  A few of the peaks correspond to the speeds of walking; these must be times I’m walking around the soccer field, and not actively involved in a play.  A few of the paces correspond to speeds I take when I go on runs around the city; these are likely times I’m running around the soccer field, either with the ball or being involved in the play.  A seemingly new peak emerges right around 5 mph, which is right at the transition-point between walking and running; this likely corresponds to a really light jog, when I’m setting myself up to become more involved in the next developing play.  I’m guessing it’s the speed equivalent of a “powerwalk” (which I’d do when I really need to get somewhere while walking, but would feel kind of silly breaking into a jog with my street clothes), and is a bit too slow a pace to really justify taking on an actual run for exercise.  Oh look, wikipedia seems to agree with this interpretation of speed.
• Data is beautiful / ugly: I spend as much time walking when I play pickup soccer as I do actually walking, as the curves at the 2 and 3.25 mph peaks almost completely overlay.  While I do spend less time staying completely still and some more time in the running modes while playing soccer, this finding is still rather disappointing.  But after doing this analysis, now I know: I really should pick up my pace during pickup soccer.

 

Methods: I collected the data using a combination of my iPhone (using the Strava app) and a smaller, more portable GPS capturing device.  I imported all of the data into R, calculated my instantaneous pace using a custom function, and visualized the data using the ggplot2 package in R.