Each number shows how many different users used each pair of words/phrases. 2 means it's just Quickseller and PrimeNumber7. So out of 28 possible combinations 21 are unique to those two accounts. I couldn't find a quick and easy way to make a 3-dimensional one but there is only one triple combination that has more than those two users. Beyond that (4 or more out of 8) it's only QS and PN7.
can you qualify that data? if there are 5 word combinations on bitcointalk that only 2 people share, does that somehow prove they are alts? how do you account for coincidences? what are the actual scientific standards for this kinda shit (if they exist)?
I think I know the potential rebuttal to that. Those words and phrases could have been picked in such a way that only those accounts are matched. But since I know for a fact that's not what happened this is sufficiently convincing for me.
yes, i believe that's true. i'm glad you're willing to concede that. i can't really verify the data either, so i'll have to take your word for it.
First of all, we're talking about "common expression" as in common on the forum; not what is yielded as a Google search result.
that's a point of contention for me. if some combination of expressions is more common in general internet usage but less common on bitcointalk in relative terms, anomalies will appear more severe in your data than they should and the likelihood of coincidence increases.
take this extreme example: two red blooded murrrican hicks from north carolina (go panthers!) join an internet forum mostly comprised of brits/euros. these two share several linguistic similarities/phrases and worldviews in common because of their southern hick upbringing. to the rest of the forum, they both stick out like a sore thumb. are they the same person?
this is the logic of your claim: "so few people on the forum use these terms that anyone who does use them must be the same person". you see how that's fallacious/circular reasoning, right? it's not self-evident.
"to sow discord" is also a common english expression:
https://www.google.com/search?client=firefox-b-1-d&q=%22sow+discord%22
there are tens of thousands of examples of people on news sites, twitter, etc saying "maximum benefit of the doubt"
Yet on the forum, there's only 13 accounts that have ever used that phrase, and 2 of them are QS/PN7. 13 users out of thousands.https://www.google.com/search?client=firefox-b-1-d&q=%22sow+discord%22
there are tens of thousands of examples of people on news sites, twitter, etc saying "maximum benefit of the doubt"
common general usage makes it difficult to rule out coincidences in your narrow data set. see above.
expressions/idioms are also tricky because you should be capturing all forms, otherwise you're not truly representing their prevalence. eg for "sow discord" you should be including users like:
I'm beginning to think that whoever wants to see Bitcoin fail - his easiest strategy might probably not be buying gazillions of GPUs but to just sow the seeds of discord within the community 
And what a wonderful job they are doing carrying out Putin's wishes as puppets engaging in sewing division and discord.
It's obvious tactic of BTU shills to sow some discord and doubt and escalate FUD
etc etc
anyway, good luck PrimeNumber7. some people around here are obviously itching to burn you at the stake.....