>>4471180
The general analogy of the Texas sharpshooter's fallacy is to take a wall of some sort, then shoot at it a bunch. Walk up to the wall and draw a circle around the shot you want to call a bullseye and circle it.
It's a post-hoc (aka after the fact) correlation. You can generally search through series of numbers and find just about any connection you want, even ones that have meaning to you. This goes for events, too. The fallacy is one in which only similarities (correlations, matches, etc.) are picked as proof, but dissimilarities are ignored. There's a Wikipedia page on it.
Q typically refers to "what are the odds" in the context of time stamps and people saying or doing specific things, not the outcomes of a random number generator. The former AREN'T coincidental even though they may seem to be on the surface - that's Q's point. Q has control over those. They can choose when to post and when the actors do what they do.
This case seems compelling to many because the odds really are low-ish, but not fantastically low (the first anon I replied to said 1 in 2 trillion or some such nonsense). I would put it (likely) in the 1 in a few thousand range for a single event called a priori (without my scripting tools, it would be hard to calculate). However, like you said, there were a lot of people on at the same time, many asking the same question, too, which increases the odds of finding matches.
We also don't have any knowledge (BO might) of how the UIDs are generated. For something relatively unimportant like UIDs, I can't imagine there is any heavy duty algorithm generating the values, which typically means successive values are correlated (not always). I don't know how it works, so I can't really say for sure.