Dream Controversy, Did he cheat?

[Cal]

Both sides are out now.
Speedrun.com did an analysis that claims over the course of 6 of Dreamm 's streams, he had 1/7.5 trillion odds of getting the enderpearl and blaze drop rates he got.

Dream commissioned a report from https://www.photoexcitation.com/, which is supposedly run by someone with a PhD specializing in statistics in astrophysics. My interpretation of this report was that Dream had 1 in 7*10^9 luck for messing with the probability at the inflection point when he had the 6 stream lucky streak he was called out for. Or 1/10 million for any point during the 1.16 streams. He probably didn't mess with the odds before he started streaming 1.16 speedruns.

I think the most interesting argument from Dream's report was the Bayesian analysis of his potential boost rate. On page 11, there is a graph showing that for the selected 6 particularly lucky streams, it was most likely that the enderpearl drop rates would occur with the odds increased at a rate of 3x enderpearl drops.

My overall interpretation is both sides handled this poorly. I don't think it's fair to of expected speedrun.com to produce a report as strong as the one Dream produced, but it still did read unpolished. It certainly did not contain the academic polish it felt like it needed.

Dream's report feels accurate to me, it seems the most generous to him it could of been, it certainly doesn't look as good for him as he would like you to think. It is very strange the person who wrote it won't put their name on it. I do not think Dream was explicitly lying about who wrote the report, but their website is sketchy to say the least. I would not be shocked if their qualifications were misleading.

It seems the most reasonable conclusion is that he boosted the pearl drop rates by 3x, it's a suspiciously clean number, it's lines up with the data from the speedrun.com report, and it allowed for the kind of subtlety I would expect from someone trying to make an interesting live viewing experience.

If you have a better grip on math than I do, please point out anything in the Dream report that seems odd. Also share your opinion in general.

 

[BOOP]

I'm gonna go with yes, he did modify drop rates, and definitely not because I have an intense hatred towards Dream.

 

[Spenser]

He did, however dream stans are a cancer. Just to have that noted.

 

[roy]

Both sides are out now.
Speedrun.com did an analysis that claims over the course of 6 of Dreamm 's streams, he had 1/7.5 trillion odds of getting the enderpearl and blaze drop rates he got.

Dream commissioned a report from https://www.photoexcitation.com/, which is supposedly run by someone with a PhD specializing in statistics in astrophysics. My interpretation of this report was that Dream had 1 in 7*10^9 luck for messing with the probability at the inflection point when he had the 6 stream lucky streak he was called out for. Or 1/10 million for any point during the 1.16 streams. He probably didn't mess with the odds before he started streaming 1.16 speedruns.

I think the most interesting argument from Dream's report was the Bayesian analysis of his potential boost rate. On page 11, there is a graph showing that for the selected 6 particularly lucky streams, it was most likely that the enderpearl drop rates would occur with the odds increased at a rate of 3x enderpearl drops.

My overall interpretation is both sides handled this poorly. I don't think it's fair to of expected speedrun.com to produce a report as strong as the one Dream produced, but it still did read unpolished. It certainly did not contain the academic polish it felt like it needed.

Dream's report feels accurate to me, it seems the most generous to him it could of been, it certainly doesn't look as good for him as he would like you to think. It is very strange the person who wrote it won't put their name on it.

It seems the most reasonable conclusion is that he boosted the pearl drop rates by 3x, it's a suspiciously clean number, it's lines up with the data from the speedrun.com report, and it allowed for the kind of subtlety I would expect from someone trying to make an interesting live viewing experience.

If you have a better grip on math than I do, please point out anything in the Dream report that seems odd. Also share your opinion in general.

Quite honestly I have absolutely zero knowledge of the situation though I can point out a few things that appear and sound fabricated. First of all, the website Dream claims to have commissioned a report from is an anonymous grant proposal review site. They review grant proposals, submissions, panels, and provide consulting and publishing advice to individuals whom request it1, all of which is done within the fields of "planetary science, astrophysics, and exoplanet communities"2 so I seriously doubt the fact that someone with a PhD in statistics in astrophysics actually went out of their way to work with someone trying to attain a Minecraft record and this does not even mention the following facts taken directly from their website:
a). this website charges extremely high rates for their services, a full review from a field expert runs at $1600 while a review from even a non-expert "savvy-reviewer" runs at $300.
b). this website also bluntly states all reviews and reviewers are anonymous and you will receive no information about the individual who reviews whatever your submission/request is, not to mention their FAQ also states all requests come with an implicit Non-Disclosure Agreement, so I truly do not understand how Dream received information about the individual who reviewed his request AND released the information publicly.

Nonetheless, a final thing to point out falls on the probability side. Not only did Dream have consistently lucky enderpearl drop rates on six occasions, but on all six occasions he had exactly a drop rate 3 times higher. This is where probability comes in, we would need to assume that based on the circumstances he- A. had lucky drop rates on six streams, and B- these drops rates always came out to the same percentage higher than average. I won't bother with doing the actual math as it does not mean enough to me, but for anyone who wants to the probability formula is P=(A|B ), which is read as the probability of A event happening given that B event also happens and considering event B was 1 in 10 million, I presume the probability is quite low.

--

To avoid looking biased, you could also say that an 11 page report with academic language accompanied with the proper application of Bayesian statistics would be very difficult to fake and its unlikely Dream would've gone through that much effort without actually being innocent. Though this is an assumption no one can actually prove the legitimacy of without comment from the anonymous author, while we are capable of proving the probability of his lucky drop rates are unfathomably small.

 

[Taiga]

Either Dream thought no moderator would be nerd enough to start bringing out the calculator for a speedrun on a block game, or this is a massively successful publicity stunt. All PR is good PR method.

I will bet you someone will write a long-ass post as a rep- yep someone's already dunnit.

I'm with dream on this, and he doesn't care much about the leaderboards. I just like his content and his response video was on point.

If he didnt care about leaderboards, he wouldn't submit his run has a world record.

 

[EncryptedSystem]

I think yes

 

[Brandon]

I personally think he cheated, and I don't really care - I understand that he would just to make his content more enjoyable (it's not fun to just watch bad rng over and over).

 

[Rhos]

Clearly not legit

You can watch and enjoy his content but you gotta take the L on this one lol

 

[Cal]

i'm not talking about him, i'm talking about the kid who made this forum post and those who actually care enough to have an opinion

he appeals to 6-10 year olds probably, i dont know his demographic but i know i'm not part of it

get over yourselves, stop whining because he cheated, who gives a damn, i cannot believe people have went to the extent to bring out a complete analysis complete with statistical models to demonstrate how he's supposedly cheating

Awfully bold of someone who thinks people care about their wax poetic profile posts to judge what people want to talk about it.

Don't contribute to a conversation you don't care about. People are interested in what their interested in. Don't be the arsehole who talks down to them for it ¯\_(ツ)_/¯

 

[thmihnea]

Even if he did or did not cheat, his image wasn't affect by a large margin, as shown by his like-dislike ratio:

I personally think he did modify drop chances, though.

 

[solo]

Well guys, you got me. I modified the drop rates

for myself, don't know if dream did tho

=/

 

[PolarGFX]

I don't think he did honestly I've watched both sides and dreams vid seems pretty legit and he did say the mods put in some false info so no I don't think he did and why would he.