But really, when something feels random, it usually means that there’s something we don’t understand You'd improve that understanding by showing standard devia

But really, when something feels random, it usually means that there’s something we don’t understand You'd improve that understanding by showing standard deviations on that graph. Are the fluctuations bigger than expected by chance? How much bigger? You'd increase it even more by mentioning that subsequent research wasn't nearly as definitive on the "lunch break effect": http://journal.sjdm.org/16/16823/jdm16823.html http://journal.sjdm.org/16/16823/jdm16823.html