Original Paper: Hi! I feel like this wasn't communicated strongly enough but this video is NOT my content. Please please DO NOT interact with my channel after watching this, this is just a PSA from the mod team. As mentioned in my comment, before releasing this I disabled monetization on my videos because I didn't want to gain anything, because principally I find that abhorrent. In retrospect I should have made a burner account solely to release this video but unfortunately did not have that foresight, so here we are. My deepest apologies.
*3:20 Binomial distributions can't be applied to any scenario without the possible need for corrections. In our case it is a really good estimate that we can improve by examining biases in the data; nonetheless the intuition of a binomial distribution still holds.
UPDATE 1: Illumina is actually in the 82.5 percentile (not 91st) as pointed out by Do_The_Upgrade on reddit. This does not change the ultimate conclusions, although the error should have been caught. Unfortunately we didn't place as much time on the "meaningless" example verification as we should have.
UPDATE 2: 10:40 I misinterpreted what Dream said while writing the script. Apparently he does not delete mod folders regularly. What he had were multiple mod profiles, and during the investigation he ended up deleting his specific 1.16 speedrun profile because he was frustrated with the investigation. (source: Dream)
Karl Jobst's Video:
Music:
Lemmino - Cipher:
Epidemic Sound - Miscellaneous audio
Willz
Credits:
0:00 Introduction
1:13 Background
1:50 Binomial Distributions
3:34 Speedrunners' Odds
4:45 Dream's Odds
5:39 Examining Bias
7:02 Incorporating Blaze Data
8:09 The "Game Is Glitched" Argument
9:49 How?
10:49 Why?
12:35 Conclusion
0 Comments