The goal to capture a weighted health, that reflects the likely hood of frustration and overall desirability

Which means it needs to track events (late, early, cancelled, missing full - frustration) but thoes events should be scored/weighted by the situation. Below is the brainstormed idea for metrics:

Tracking:

• Time of day weighted (rush hour is more important)
• Last or only of the day, cancelation punishment
• Frequency (how often, and if this is a enduring problem) weighted
• Daily and weekly consecutive weight (pos/neg - back to backs, and every day same hour missing) but moderated by total wait time (1+ waits are punishing)
• Fullness weight (Overcrowding or fine)
• Bonus points if the reader is down (free rides)

Not going to track:

• Modifier for if there is "other options" (ie. Many busses go downtown/to a station), because that biases the formula and its not true for all patrons
• Quality of the ride, as buses change
  • Could track on a "trip/time" level when most often the anti-social ride the buses, vs battle hardened regulars.
  • Adding a "quaity modifyer"