It might not be the test that's flaky

I have a feeling this was glaringly obvious to everyone but me, but late last week I had a bit of an "a-ha" moment.

When we say "flaky test" we imply that the test itself, or the test environment, is somehow at fault. Although this may often be the case, it's just possible that the test itself is perfectly sound and the implementation is flaky.

Nothing earth shattering, but I just thought I'd mention it...