I believe it’s a mistake. I’m probably using incorrect terms to describe this part(!), but the 2 Event Key value appears to be linked to a Non-Random Biased result…
My epidemiology professor (i.e. medical statistician) proposed that when examining raw data, one should let the data speak for itself when it comes to revealing anything of interest. I went and compared Mean Machine’s results to those from @Ashasekayi Black Dragon, @Saluki INTRIM, and @Aelthwyn Anonymous.
In all cases, whenever a pull from Table 2 resulted in 2 Event Keys, it was ALWAYS positively linked to receiving either 6 Gems Keys (Table 3) OR 2 Gem Keys (Table 1) in that same Legendary Task.
I wouldn’t mind if the number crunchers could take a look at these findings, please.