How can I fix the discrepancies between Apple crash reporting and my own?

Question

I use a third-party crash reporter for my app, and when I look at the crashes in the web console for it, I see hundreds. However, when I look at the crashes that Apple reports via Xcode, I only see about 30. I realize that I'm getting far fewer crashes from Apple than from the third-party reporter because Apple's system is opt-in, but this seems like far too few crashes. For example, the top crash in Apple has around 6,000 users, which is not that far off from that of the top crash in the third-party reporter, so 30 for the total number of crashes in Apple seems way too low.

Answer 1

My intuition is that the Apple crash reporter is probably doing a better job of coalescing multiple related crashes. This is something you could probably verify by looking at some of the crashes in the third-party tool, and seeing whether they're very similar to one of the 30 that Apple gathered.

The "long tail" effect is probably also in play here. Typically, you will have a few "common" crashes, and a lot of "uncommon" crashes. If only 10% of users click the "report" button in the Apple crash dialog, this means your "common" cases will have about 1/10 as many reports in the Apple system. For the "uncommon" cases, you might get ZERO reports, if there were only going to be ten instances total of that crash over a particular period.