Interesting discussion from Kaiser Fung. I don’t have anything to add here; it’s just a good statistics topic.
Scroll through Kaiser’s blog for more:
Dispute over analysis of school quality and home prices shows social science is hard
My pre-existing United boycott, and some musing on randomness and fairness
etc.
He seems to have no idea what Uber has been accused of by Apple and how any of it works.
If you read my post first, you can see I make no claims on knowing the details of the implementation. I cited what the NY Times said Uber did. The press so far has failed to explain how Uber broke the rules, how this scheme faciliated fraud detection, and whether the data sleaze was used for any other purpose. As Jonathan pointed out below, people are speculating as to what they did. I spent most of my career as a business statistician so I understand how data are collected and used by businesses.
https://techcrunch.com/2017/04/23/uber-responds-to-report-that-it-tracked-users-who-deleted-its-app/
I believe Uber was assigning an ID to a phone when it registered the app. When the app was then deleted, Uber did not track the phone at all; it was just a registry of phones kept by Uber. When the app was installed again, the phone was ID’d by Uber. The unknown question is how this was done: did they access a UDID? Did they use their own algorithm? Speculation is they actually accessed the phone serial number, which violates Apple policy in a bunch of ways. And this is the proof offered:
https://mobile.twitter.com/chronic/status/856250223777206273