It seems that a test result that is skipped is included in the failure count. For example I have a test that has 15 failures and 3 skips, the stability status say: "Failed 17 times in the last 22 runs.". That does not seem correct to me since a skip is not really the same as a failure.

I think it should either be excluded from the count or at least make it configurable how the skips should be counted.