On a long run earlier this week my brain decided to overthink speed error on my treadmill. So I did some rough video analysis (painted an approximately inch square white on the black belt, measured the belt length, had a program count frames between seeing the white square at the same spot) to calculate speed. This probably isn't the kind of thing people are super interested in, but if anyone were going to be, I figured it'd be here on a Sunday.
I got my Pro Form 995-SEL treadmill in a broken condition from one of the Habitat for Humanity REStores (which area great!) for $10. so I assumed it'd be on the worse end for this test.
Data were collected while not running (I imagine running would make it slightly worse), after the belt had come to speed, averaged over 30 seconds. My treadmill claims to max out at 10mph which probably explains the spike in error at the top end.
All error was too slow.
Requested speed (mph) | min / mile requested | measured speed (mph) | measured min / mile | extra time / mile | absolute error % |
3 | 20:00 | 2.86 | 20:59 | 0:59 | 4.67 |
4 | 15:00 | 3.86 | 15:33 | 0:33 | 3.50 |
5 | 12:00 | 4.8 | 12:30 | 0:30 | 4.00 |
6 | 10:00 | 5.72 | 10:29 | 0:29 | 4.67 |
7 | 8:34 | 6.76 | 8:53 | 0:19 | 3.43 |
8 | 7:30 | 7.83 | 7:40 | 0:10 | 2.13 |
9 | 6:40 | 8.76 | 6:51 | 0:11 | 2.67 |
10 | 6:00 | 9.3 | 6:27 | 0:27 | 7.00 |
View attachment 981055