I was wondering if anyone else has run into a situation where the FindBugs statistics, specifically, the deltas that are calculated between two builds, appear as being incorrect. From what I've observed, if the delta is fairly small, these stats appear to be calculating correctly; however, if the (absolute value of the) actual delta is much bigger than ~3-5, then the calculated deltas just seem to be plain wrong.
Here's an example:
The delta is correct: 125 - 2 = 123.
is just wrong! Obviously, 123 - 17 != 29. The errors are also wrong.
Is there anything that might be potentially misconfigured? As far as I can tell, the absolute numbers are correct as compared to running FindBugs in IDEA (or at least in the right ballpark; not sure if the checkers I have in my IDE match the ones we're running in TeamCity).