If the system of numbers they designed is so robust and according to you his overall sample size is so small why did they not take ALL his streams?
I didn't say the sample size was "so small." This is what I said:
Those streams include 262 piglin trades and 305 blaze kills, each of which is a statistically independent event with a known probability. That's more than enough data to come to a valid conclusion.
They went and looked at his run when he submitted it, went "hmm his trades seem lucky" then went and found 6 other streams and took the BEST RUNS FROM EACH OF THEM NOT ALL OF THEM
No, they used literally
every piglin trade and blaze kill in every run from Dream's October speedrun livestreams. I don't know why you still think they cherry picked the best runs from the October streams when I already explained to you a few pages back that they didn't. Their data includes
every single run in all six of those streams. I don't know how to state this any more clearly.
So now they have is world record run and 6 more runs that were personal bests of his for each night, add that together and that's 7 runs that were all personal bests. How do you look at that and go "yes an even sample of his runs were gathered"
Again, I have no idea where you're coming up with the idea that they cherry picked the best run from each stream. Also, there was no WR run and Dream only set a single PB on his very last run.
Even if they are 100% correct and he cheated like a black man on an IQ test why not put his other streams through the system of equations to give contrast of how much his luck changed
Ok, I can do that. Outside of the six October livestreams Dream did, the only other 1.16 speedrun streams he did were five streams in July. Let's compare the luck he had in July to the "luck" he had in October.
Here is the paper from Dream's mystery Harvard astrophysicist. This portion has information about Dream's ender pearl trades and blaze rod drops from his five streams in July:
If you compared Dream's results from his July streams to 100 other Minecraft speedrunners performing the same number of piglin trades and blaze kills, then based on a simple binomial distribution, you'd expect roughly:
- 86 out of 100 runners would have better pearl luck than Dream
- 5 out of 100 runners would have equal pearl luck
- 9 out of 100 runners would have worse pearl luck
- 13 out of 100 runners would have better blaze rod luck
- 4 out of 100 runners would have equal blaze rod luck
- 83 out of 100 runners would have worse blaze rod luck
In other words, Dream was rather unlucky when it came to pearls and rather lucky when it came to blaze rods, but neither value is far enough off the expected values to be suspicious.
Now let's compare that to Dream's luck in October: 42/262 ender pearl trades and 211/305 blaze rod drops. That would mean if you compared Dream to
one trillion other runners, or approximately 128x the total human population on Earth, you'd expect roughly:
- 6 out of the one trillion runners would have better pearl luck than Dream
- 4 out of the one trillion runners would have equal pearl luck
- 999,999,999,990 of the one trillion runners would have worse pearl luck
- 9 out of the one trillion runners would have better blaze rod luck
- 5 out of the one trillion runners would have equal blaze rod luck
- 999,999,999,986 of the one trillion runners would have worse blaze rod luck
Is that enough of a contrast in luck from July to October for you? If you still think it's all just cherry picking by vindictive mods, I honestly don't know what to tell you. You might be straight up retarded.
