*5.1. General Statistics*

The six subject enrolled in this study used the app to collect data of a total of 22 runs amounting to more than 22.5 h of data. In Figure 6a, we present the durations of the runs for the different subjects. In addition, Figure 6b depicts the total distances traveled during each run for the different subjects. Finally, Figure 7a presents the deviation of the runner from his or her target cadence for the run in question. Note that in some runs, the subjects elected to disable GPS data collection, and those runs were omitted from Figure 6b, which is why the number of runs per runner is different in the different plots.

We note that for the majority of runs, runners were running at a cadence lower than the recommended cadence for the run, signaling the need for more personalized training regimens as stated in our long-term goals. Moreover, we note that the auditory and haptic feedback is provided whenever the cadence is more than 10% off the target value for the run, for a period of 30 s or longer. Therefore, runs that had a deviation in cadence within a 10% window of the target cadence level should be considered as ones that met the target cadence. Moreover, runner s28ikk reported running with the phone in his hand. As will be detailed in Section 5.3.2, this violates the design assumptions of the cadence estimation algorithm, which explains the large deviation from the target cadence for that runner.

In order to provide contextual perspective, we present the self-reported levels of fatigue after each run by the different subjects in Figure 7b.

**Figure 6.** (**a**) The durations of the different runs as observed in each subject (*N* = 22); and (**b**) the total distances of the different runs as observed in each subject (*N* = 20).

**Figure 7.** (**a**) The deviation of the average cadence from the target cadence for each run, expressed in percentage points (*N* = 22); and (**b**) subject-reported level of fatigue after each run. A value of three represents the level of fatigue reported after an average run; a value of one is least fatigued; and value five is most fatigued after an average run (*N* = 22).
