But the data point is just wrong or poorly described. Did you really just test one side (the "funny elbow") and compare it to the possibly very different thing

But the data point is just wrong or poorly described. Did you really just test one side (the "funny elbow") and compare it to the possibly very different thing from the article. A search for "fake British accents" brings up multiple videos with >100000 likes each and significantly more such than for funny elbow.