"He cheated" "No he didn't" "He is the worst" "Just admit it" blablabla...

Give the guy a fucking break will you? So what if he did cheat? What if he didn't? I have been in this fandom for 6 months and there hasn't been a single day people weren't out to get Dream's head.

I am not a stan, neither am I a Dream apologist. I just think all of you need to calm down for 5 minutes, for both his sake and yours. This much stress is not good for anyone and you are stressing out the rest of us that see Dream as their comfort Youtuber.

so binomial distribution is basically the probability of success or failure. only two outcomes. already, i find it strange they’re using that for the trade system, where it isn’t a two outcome situation. i know the two posibilites they’re comparing is a “yes you got ender pearls” or “no you didn’t get ender pearls” but either way there are other probabilistic models u can use that better match the circumstances. it’s almost like they chose this model specifically because it gave them the suspicious probability they wanted

anyway, there is criteria that you have to meet in order to use this model. one requirement is that the trials ur collecting data from have to be independent of each other. in the pdf the mods say they’ve accounted for this independence by accounting for their sampling bias, but they haven’t. all they say is defense that they haven’t cherry picked their data is that the possibility is “extremely negligible”, which is hilarious considering what they’re trying to prove. i personally am not of the opinion that the trials are independent given the sequence of streams, which then makes any calculations of the binomial distribution void.

the pdf also states that they’re going off a bernoulli distribution, which is closely related to the binormal distribution but has the added criteria that the probability of failure (not getting the pearls) is equal to 1 - the probability of success (getting the pearls). again, i don’t think that applies to this situation - judging from the probabilities and the weight from the pdf, the math doesn’t add up. i also took the values off of the minecraft wiki to double check, and anyone who is reading please feel free to correct me, but the the added probability of failure of pearls and success of pearls doesn’t equal one. so their data is either incorrect or they can’t be using this model.

these are just the issues i’ve picked up on skimming the pdf, but i’ve seen other dissenting opinions from people im sure are way more qualified than i am to be speaking about. sorry for rambling!!! but yeah here are my thoughts

don’t apologize at all, thank you so much for this! again i probably am too invested in something that i can’t have a say in, but this is a very interesting insight. i’m not sure why you’re saying that each trial is not independent, is it because of the data being cherry picked?