Re:Uber's experiment in behavioral science to subtly entice ...
by jjcm on Sunday April 02, 2017 @09:59PM
(#14020151)
I've noticed this so much that I've recently started a spreadsheet of estimated arrival times, actual arrival times, estimated trip duration, and actual trip duration. I currently only have 4 entries so the results don't really hold weight (though right now the average wait is ~4 minutes estimated with ~12 minutes actual). Uber's time estimates have been simply atrocious lately.
[ Parent
| on HN ]
Re:Uber's experiment in behavioral science to subtly entice ...
by javiramos on Monday April 03, 2017 @09:45AM
(#14022678)
Me too! I have roughly ~35 entries and I observe, on average, that actual times are ~2.5 times the estimate (roughly what you have seen as well). I am sure that there is some strategy here, estimates that are consistently 2.5x off are just ridiculous... With the amount of trip data that they have, they could easily build a fairly accurate ETA predictor. Even google maps would do MUCH better...