The worst-performing devices were the two Garmins, which ranked last on most performance metrics and with biases that were often more extreme than actigraphy. Based on their high EBE NPV values, it is possible that the poor performance of the Garmin devices is in part due to having a relatively higher threshold for wake (and a lower threshold for sleep) than the other devices, which resulted in the Garmins performing poorly for tracking both sleep and wake because so much actual PSG wake was missed – but the wake epochs it did detect were more likely to indeed be PSG wake than the other devices.
https://academic.oup.com/sleep/advance-article/doi/10.1093/sleep/zsaa291/6055610
Why Garmin don't you fix this. You have the capacity to rule the market. To be the best. But you refuse to fix the issue.
