Touch rugby GPS data
Our high resolution high bandwidth prototype unit allowed us to extract some interesting info for Touch Rugby. First you can see from the speed profile of a player below that the game involves a lot of acceleration and deceleration with top speeds here of 25 km/h. Players typically play a set or two then sub off, then come back on while their buddy subs off, etc. The data underlined in red shows time "in the sub-box". The idea is to always be able to play with high intensity.
From this data, we can draw some basic statistics and answers the key question: what distance was covered? The table left shows you 2.56 km in this game, 15:49 of “on field time” (full game is 40 min), an average speed just below 10 km/h.
But as we said earlier, the game is a lot about intensity. And that average speed of 10 km/h could really be achieved with 2 profiles:
- Profile 1: someone jogging at 10 km/h all the time (no intensity)
- Profile 2: someone stationary half the time and sprinting at 20 km/h for the other half.
I’m simplifying here but you get the idea. One way to differentiate these 2 profiles is to further classify the speed by speed range as per the graph below. There you can see that the larger the red & amber zones, the closer you are to profile 2.
Then if you want to compare games relative to each other, a visual representation can be the one below. In this graph we compare games of the same duration, for the same player and in the same position. It provides a good relative effort "signature" of each game. The main factors affecting the position of the marker here will typically be game level and defensive strategy.
How did our prototype compare to other existing GPS - Garmin 230, Suunto Ambit 3 Peak and Fitbit Surge
The next 3 pictures show you the result of comparative tests between our prototype unit and three GPS watches. Each GPS watch was worn at the same time as our prototype unit and each picture shows:
- The speed vs time profile for proto (in blue) and GPS watch
- An extract of the GPS track for GPS watch (bottom left) and our prototype (bottom right).
To the right shows the result for the Suunto Ambit 3 Peak watch, you can see how it has fewer samples than the our prototype and the effect that it has on the speed signal missing some of the accelerations and instead averaging to a lower speed.
The fitbit surge displays similar characteristics: a reasonable number of samples but again not enough to get the real speed signal which results in lower speed records.
The Garmin 230 is the worst performing with terribly slow sampling rate and subsequently the worst speed data.
Measuring heart rate with Fitbit charge HR during touch game
This paragraph is not connected to the GPS project but relevant to the description of touch rugby above.
Both Tell and I had been very interested in knowing what our heart rate data looks like during a game of touch. But wearing a chest sensor is too risky for touch as it could hurt ribs if unlucky in a dive. Fitbit were the first to make heart rate measurement possible for us when they released the charge HR, so we jumped on the opportunity.
We were on and off simultaneously, while playing at the same position in a full game and we very much enjoyed seeing the 2 trends being quite similar below. (from min 20 to 25 is half time).