@jamesthomson Yes, I’ve recently changed my own policy on this - I’m finished with them. Once you go through more than two when you know you’ve gotten them all correct, it’s time to pull the plug.

@sommer No snow???

@jsnell Could not agree more - I will not buy into another platform which uses this model. This approach turns a computer into a TV and is in the context of the iphone is hilarious to me, Apple’s original pitch was essentially to allow innovation by breaking the carriers control over the devices, but what we actually ended up with is the exact same thing but just with Apple being in control.

@drdrang I've thought about this before. My best guess is that they're searching for a split where via a 2-sample t-test there would be a statistically significant difference in means between the two populations. Mathematically this would be valid although logically not IMO. There might be other more specific tests to identify subpopulations in sequential data but I’m not personally familiar with them.

@sommer I was once in a meeting where we all three had the same name but with different spellings. Without exaggeration a career highlight.