111895 2013-03-08 15:39:55 -0800 run-perf-tests should make the "mean" stand out in its output 2013-03-08 15:55:50 -0800 1 1 1 Unclassified WebKit New Bugs 528+ (Nightly build) Unspecified Unspecified NEW P2 Normal --- 1 eric webkit-unassigned abarth dpranke rniwa oldest_to_newest 851569 0 eric 2013-03-08 15:39:55 -0800 run-perf-tests should report the mean instead of (in addition to?) the median I'm told I should be reporting the mean in my bugs: https://bugs.webkit.org/show_bug.cgi?id=107337#c40 I may have been a Math major, but I plead ignorance for stats questions. :) 851575 1 rniwa 2013-03-08 15:44:29 -0800 (In reply to comment #0) > run-perf-tests should report the mean instead of (in addition to?) the median > > I'm told I should be reporting the mean in my bugs: > https://bugs.webkit.org/show_bug.cgi?id=107337#c40 > > I may have been a Math major, but I plead ignorance for stats questions. :) We do report them. e.g. Running Bindings/document-implementation.html (4 of 114) DESCRIPTION: This benchmark repeatedly accesses document.implementation. RESULT Bindings: document-implementation: Runs= 337.387461737 runs/s median= 336.553080486 runs/s, stdev= 8.51712865431 runs/s, min= 321.659870522 runs/s, max= 347.246989122 runs/s The mean is 337.387461737 runs/s. 851585 2 eric 2013-03-08 15:55:09 -0800 I see. That was not at all obvious to me in reading the output, but perhaps it should have been. 851587 3 rniwa 2013-03-08 15:55:50 -0800 (In reply to comment #2) > I see. That was not at all obvious to me in reading the output, but perhaps it should have been. If it was't obvious to you, it clearly is a bug. Let's fix that.