Bah, seems like those 170m and 165m links are broken: here are better links.

165m: http://build.chromium.org/p/chromium.perf/builders/Win%208%20Perf/builds/958

170m: http://build.chromium.org/p/chromium.perf/builders/Win%208%20Perf/builds/1056

Here's a quick Google sheet I threw together with the runtimes of each of the individual stories: https://docs.google.com/spreadsheets/d/1Geh-ssYjJYty8z8dua232JsXf-GKh7oD1_hEgKrJd9A/edit#gid=0

It doesn't seem like there's any particular story that's a huge outlier, which makes me think that there's some environmental noise on the system that's causing it to sometimes run slower.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/6c651646d2231aaf7f4f608f075b5d02f8dc5e2a

commit 6c651646d2231aaf7f4f608f075b5d02f8dc5e2a
Author: Charlie Andrews <charliea@chromium.org>
Date: Wed Aug 09 15:51:09 2017

Increase swarming timeout of loading.desktop from 3 hours to 4 hours

TBR=eyaich@chromium.org

Bug: 753798
Change-Id: Ibef3870b8abd3175aabb1d587082d82f5a951580
Reviewed-on: https://chromium-review.googlesource.com/608469
Reviewed-by: Charlie Andrews <charliea@chromium.org>
Commit-Queue: Charlie Andrews <charliea@chromium.org>
Cr-Commit-Position: refs/heads/master@{#493000}
[modify] https://crrev.com/6c651646d2231aaf7f4f608f075b5d02f8dc5e2a/testing/buildbot/chromium.perf.fyi.json
[modify] https://crrev.com/6c651646d2231aaf7f4f608f075b5d02f8dc5e2a/testing/buildbot/chromium.perf.json
[modify] https://crrev.com/6c651646d2231aaf7f4f608f075b5d02f8dc5e2a/tools/perf/core/perf_data_generator.py

This seems to be passing after that CL, should we mark fixed?

#no: that is just a bandaid fix to increase the total timeout. I think we should reduce the cycle time by 20% to have room for upcoming SW tests.

Assign this bug to Sakamoto@ for culling stories in loading.desktop to reduce its cycle time by 20%.