I'm not sure you're looking at this correctly, if you're doing it automatically your process is not working as you think. Names with different spacing or slight variation will not match up algorithmically (usually). I manually looked at the lists for the last three events and matched them up by hand.
It's a blend of function and manual checking. It's mostly accurate.
There are 35 people that got in for the last two events. The Be Our Guest event had 98 people on the list and the Splitsville event has 125. That means that 35%(!!) of the people that made it into BoG are also going to Splitsville. Of those 35 people 21 also went to the Wreck-It-Ralph event.
The difference is you are comparing a smaller subset - while I compared against all events. The events you compared only represent 223 slots - wreck it ralph alone was a bigger event then those two combined. that's why your numbers are so different.
You compared two events and found 35 people in common between those two events. But when you extend that.. you find that commonality drops to 13 when you go beyond two events. And of those 13, 9 were at both BoG and Splitsville. So if you want to talk habitual repeaters, you need to look beyond just 'one' offense - and at BoG that is 9 out of 98, (~10%) and 9 out of 125 (~7%) at Splitsville. Again - a very small minority.
If you want to look at it event by event...
BoG had 44 out of 98 'first timers' - 44.8%
Splitsville had 67 out of 125 'first timers' - 53.6%
Wreckit had 227 out of 229 'first timers' - 77.2%
Marathon had 39 out of 40 'first timers' - (lifestylers don't run!) - 97.5%
Holloween had 107 out of 150 'first timers' - 71.3%
Now not all is equal.. first, the more events you have, the more potential for overlap you have of course. So later events are more 'prone' to it if you haven't explictly done anything about it. Second, not all events were equally desirable (Marathon for instance).
I didn't do the match up between Wreck-It-Ralph and Splitsville but I'd wager that it's at least 1/3 of them were going to Splitsville.
So 21 people went to the last three events.
Nope - only 13 have made it to more than 2 events. And the drop off between 2 events, and three is huge (70 vs 13)
Looking at the guest list for Splitsville --
28% also attended BoG
17% also attended BoG and Wreck-It
If you just look at BoG --
21% also attended Wreck-It
35% are going to Splitsville
I see 33 names in common between BoG and Splitsville. But if you say BoG, Splitsville, AND WreckIt.. that number drops to 13.
I think it makes it clear - it is not a small minority of people hogging these events. There are a significant percentage of people that are going to repeat events.
You conclusion is weak and flawed. It's not looking at the whole picture and relies too much on extrapolation. It's akin to seeing people working Monday and Tuesday.. so you conclude people work every day of the week because you saw them working both days you sampled.
The last two events had a much higher percentage of repeats compared to previous ones - but by looking beyond those two events, you can see it's not the same people hogging all the slots. And the ones that are only two time repeaters... if they are blogsphere people.. their presence as an individual SUCKS. That doesn't bode well for the theory that the events are dominated by hand-picked people for their social media presence.