I wonder how long the bball thing took XD

@BlacksmithTWD These are all of the streams Dream did in 1 week, they are all uploaded on twitch

@TheAtheistPaladin I think 100 million absolute best case. Most likely, more like 1 trillion

@Douglas Morrison Probably did but he realized it would be good for the video so he re did it until he didn't smile.

I love minecraft, and find it too be highly mathematical, I am sure you found this to be true as well, I came for the math though and thanks. My fahter and I both actually love this game, we use the mathematics embedded in the game to create massive structures in Survival. And in doing so, we learned to count at incredibly fast rates, although, only usually based on 2 and 4 though, or sometimes 64 if we must. It has created an interesting view for myself, as I began to notice real mathematical patterns in the real world, which was really wild.

@thefizzydude I'm saying because he did the record calculation given an estimate of the set of all runs instead of the set of all runs completed the probability of the record will be way higher.

They're such an anomaly though, even including them the odds are negligible at best

One in 2187. Ish. People don't select their hands entirely at random, after all. Might be vaguely of interest that it's the same probability as winning 7 times in a row and losing 7 times in a row.

@Athenri yeah i know it would just be funny if it happened

nah, it's pretty straightforward to do the math. That's the great thing about math when done right - people can verify that your calculations are: - the right thing to calculate - done right.

Like crash mentioned they were his luckiest streams but they also were in a row. He took a break after 5 as he was getting frustrated with his luck then a while later starts again and his luck is through the roof.

I believe they were 6 streams in a row. There were a total of 11. The first 5 happened before he took a break from speedrunning.

@Jannik M. Meissner And I understand that! It feels like it *should* be natural. It's totally coincidental, but it feels so right.

@Athenri good point, that is where my thinking went wrong… it just feels so right 😂

@Athenri Can you tell me more or link me to a video you think is good about the 24 hour clock being a historical accident? Sound like a fun and interesting thing. Maybe even a subject for minutephysics or numberphile.

The earth could also rotate slower or faster and have the same orbit period.

@Jannik M. Meissner yes, but we didn't *have* to make days 86400 seconds. That's basically an arbitrary number that comes from a 24 hour clock (historical accident) and dividing each hour into 3600 seconds. We could just as well have had a 30 hour clock. Then pi billion seconds wouldn't have been anywhere close to a century.

You seem to be talking about biogenesis and not evolution. Evolution is as much a scientific fact as gravity. As for biogenesis, how do you know that it’s impossible? Have you crunched the numbers and compared the chance of proteins forming against the number of opportunities? You also need to account for the fact that there are potentially other ways life formed in a different way to us, otherwise your paining your target around your dart. We also have no way of knowing if our universe is the first or only in which case biogenesis is not only likely but in inevitable to occur.

The only thing it disproves is that creationists know what they're doing.

Evolution doesn't state how life began. Simply how species, well... evolve. But I'd love to take a look at the paper in question.

He's using math to come to a conclusion. He made a video about something that many people have requested. What about this is something you don't agree with?

He wasnt complaining, he was using the example as a tool to educate people on math and statistics.

@Athenri well then my chances were improved, i mean they already are 100% since u read my comment lol

@zoiuduu This video is a bit unique since there's a bunch of mathy guys in here trawling "newest" to beat up Dream fans with stats and take their lunch money.

@Athenri 1 month after the video was posted, i didnt even imagined the other option lol, but thanks

Depends, do you mean in one month from now or since it's been one month since the video was posted

The amount of white noise created by world rng makes this impossible and even if it were there’s no way you’d achieve it outside of a set seed run, and even then it would take impossibly high precision that only a tas would be capable of, and even then Dream would need to provide an explanation for how he was doing it for the run to be verified.

It isn't. Also, if Dream had found a way to game the RNG, he should have used that as his defense for why he was getting so many drops and not "nah, this luck is not that rare"

The amount of white noise created by world rng makes this impossible and even if it were there’s no way you’d achieve it outside of a set seed run, and even then it would take impossibly high precision that only a tas would be capable of, and even then Dream would need to provide an explanation for how he was doing it for the run to be verified.

A little more complex. It defines a seed based on time. That seed progresses every time it's triggered by something to a new value that you can't predict. In the nether, where you're going to be doing this, the lava will be triggering the seed thousands of times each second, making it completely impossible to aim at any value.

Let's say I create a mod that's embedded within one of the allowed mods. This mod makes pearl drops 3 times more likely and gives blaze rod drops a boost too. I stream this all, doing everything else normally. Random world, etc. How would be streaming this verify me being legit?

That's why it was easy to see that he was cheating, yes.

The mods' original paper did do this. I believe they compared it to some runs of Illumina.

@crash yeah I saw that which is why I first got the book

He did a great google talk promoting this book and telling stories he couldn't get in the book.

@Athenri ohh i see

@Burn Stick Part of the issue with Dream's expert's math is that he often doesn't show exactly what he did - but you can tell from the result that something went bonkers, but unless you go through what he *says* he should be doing, you can't actually see that it's wrong.

​@Athenri okay i didn't know dreams expert did his math wrong (which I would have expected to be criticized in a video about this topic). Now I actually fully understood what Matt did (I thought he only corrected partially (due to the assessment from the "expert")) and this now sounds way worse for dream. thanks for the explanation

@Burn Stick No problem. I'll try to give you a really succinct brief on it. They both (roughly) agree on the massive number 1 in 20 sextillion. Dream's expert doesn't agree that the binomial distribution is accurate for pearl trades (it is - it is in fact about equivalent to Dream's expert's Barter Stopping Simulation over the 22 runs or so where he actually does pearl trades) But you always have to correct for biases when doing statistics. It's rare for a *particular* person to win the lottery, but not rare that *someone* wins the lottery. So maybe Dream is just the lucky person among the many trying, right? And maybe there are also multiple ways to get lucky, and not just Ender Pearls and Blaze Rods. Painstakingly correcting for these biases (honestly, overcorrecting), the mod team pushes that probability down from 1 in 20 sextillion to 1 in 7.5 trillion that *any* speedrunner in the top 1000 would *ever* experience this kind of luck in *any* two variables. Dream's expert then does his own math. Some of it is wrong, some of it is ludicrous, and a lot of it is certainly overcorrecting to an even more absurd extent than the mod team did. He ends up with a final probablity of one in ten million that *any speedrunner would ever have this kind of luck in any two variables*. That's still, in my book, indicative of cheating, and if I only had Dream's expert's report, I'd have thought Dream was cheating too. Along comes Matt. Matt says "Well, instead of digging into whether we should correct for top thousand or top ten thousand players, whether we should include the old runs or not include the old runs, whether we should correct for 10, 20, 40, or 100 different ways to get lucky - how about we roll all the way back to the one in 20 sextillion? That's the raw probability of Dream's luck or better occurring to a specific person in one trial. That's before they start disagreeing on how to correct for biases etc." And that's when he points out that *even if we take some event that every single human on earth can do every single second for a century*, we still get nowhere close to enough events to expect even a single person to ever experience this kind of luck.

@Athenri ohhh i didn't knew there were 2 big numbers. My mistake

@Athenri thanks that seems pretty obvious now thanks a lot.

Well, the probability is the same-ish whether or not you yell "I'ma get six sixes". The difference is that there are some other results that would normally be notable when you consider biases. Like let's say you're playing some game where *any* six of a kind is very valuable. Then clearly there are *six* ways to get that, and we need to adjust for that, because six sixes, six fives, six fours, six threes, six twos and six ones would all have been equally amazing results. In this case, you're locked specifically into that one, particular result, six sixes, so all we need to ask is "what's the probability of getting that particular result" There are 46656 different results when you roll six dice. You're interested in exactly *one* of them, so it would happen, on average, once in 46656 sets of six dice being thrown.

RNG manipulation is not possible in this version of minecraft. The mods actually went into this in their original paper, taking a look in the game code. And even if that was the case Dream could easily prove his innocence.

Nice, a cuber

@piratestu bad take, boo

Bias describes having a preference for or prejudice against a particular party. Him getting paid by IRmys does not tangibly put him in favour of one party or another. He isn’t a fan of dreams, doesn’t really play Minecraft and isn’t a part of the community so he doesn’t have a dog in this fight.

@Athenri friend, I want you to think about something... you are white-knighting a guy who spent at least a third of his video talking about a bias in mathematicians but you can’t begin to think that person bringing it up could be biased themselves because they just said so at the beginning? I hope you enjoy the 3+ copies of his book he sold you and, if you’re interested, have I got some snake oil that you’d definitely be keen to know about.

@piratestu man, that's the stupidest argument I've seen today, and I've seen a lot. More than matt is interested in scoring big on a video about potential cheating in Minecraft, he's interested in having integrity as a math person on the internet and among his mathematician friends.

@JuukaRock of course he can, but it doesn’t make him unbiased... now if he said he’s disinterested and all proceeds of this video will go to charity, then he can talk... until then he has a financial interest for the video to do well... and ask Fox News the best way to get eyes is to take the most outlandish option.

@morbideddie he is the hypocrite imagine having a name REXY THE TREX FROM JURRASIC PARK

Explain.

@OmegaAudio Don't want anything "from him". This junk popped up in my suggested videos. He needs to rethink doing viewer requests if his audience is this stupid

Oh, I see what your saying, making a video on a topic that is both highly requested and still highly misunderstood by a large portion of people is not worthwhile for an education channel because it happened a little while ago. I'll go tell my history teacher to pack up things and find a new career.

@RoomerJ He literally said he got a ton of requests for this subject, so he's answering the requests. The hell do you want from him?

Is someone forcing you to watch it?

@morbideddie3 weeks and this issue has been solved months before that...

Possibly - but then he might have had to play more before he had a lucky run good enough to make it, and then it would have been clear once more

@Robert Lukotka And this is why we don't need to take you seriously, because you don't know the situation you're talking about. It's six consecutive live streams (i.e. there is no cherry-picking of lucky runs or bad runs involved, on either Dream's side or the mod team's side). I recommend that next time before you make a bombastic declaration of "your reasoning is completely incorrect" - you make sure that you actually know what you're talking about, because in this case you did not. Embarrassing.

@Robert Lukotka He was live-streaming. So there was no real bias for picking lucky runs. The video even went into trying to account for a bias of picking just the lucky runs. And this wasn't 6 runs either. They were 6 streams. about 24 hours of it. So I feel like with all considered, there is a large enough sample to base a conclusion out of.

@crash If you do not know how many runs were in-between that were not published (maybe due to bad result because of bad luck) you must consider them separately at least with respect to some aspects of the problem, otherwise you may end up in my coin flipping example evaluating an ordinary event as unrealistic. My claim is that if you make 100000 runs (probably less, there are more aspects that needs to be accounted for) and you pick 6 best ones, this is the expected outcome. This is far less tries than 10^-22 suggests.

Why do you need to calculate each run separately? The game doesn't care whether you log out, start a new world, or whatever. It's not like with the dice where getting 7 many times in a row is considered good. Getting a 7 50% of the time for instance with hundreds of rolls will be suspicious though. No matter if they happen all after one another or not.

Yes, if someone is deceptively passing off non-consecutive runs as consecutive, you obviously can't use it as a measure for how likely a particular thing is to happen. That's a silly complaint.

@morbideddie just adding to this - if he *did* have some kind of wild glitch that meant he wasn't playing the same kind of game as everyone else, that'd still not qualify for glitchless.

Geosquare adressed this on his video

The RNG in 1.16 minecraft is generated at irregular intervals and is dictated by various factors in the game world which is itself randomly selected. The white noise present due to this makes the RNG effectively random in that it cannot be reasonably manipulated or predicted for either blaze rods or ender pearls. In effect this means it’s random. This is backed up by other runners and simulations showing drop rates in line with the numbers we would expect from random numbers. I’m no computer expert so it maybe possible that dream had a specific processor error that allowed his computer to run normally but happened to give him above average luck in the two items he needed. That said I think the chances of that happening are about as unlikely as him getting the luck he did in the runs.

Though I do realize that the things like the ender pearl and blaze rod chances are unaffected by the worlds terrain so maybe it doesnt matter..... thats why I asked haha

These things are corrected for in the original mod paper. Correcting for them, and for the fact that it could have been any speedrunner, yields roughly one in 7.5 trillion.

unrealistic results in 2 categories doesn’t affect a players skill or results in other categories, for instance, fortress generation, stronghold generation, eye of enders breaking, etc. it’s ignorant to assume that these two specific instances (blaze rod drops and eye of ended trades) would account for the entirety of the run

he only got fourth place because one of the eyes of ender broke he wouldave gotten first otherwise

so.....screw the entire community of people who DO genuinely care about minecraft speedrunning? That seems a little harsh just because that's not your interest.

@Derploop Research on the whole Dream dilemma, thats what I meant. I agree with you that the luck factor can be pretty tedious in Minecraft speedrunning, but in my opinion that doesn't take away the fact that it's still a great game to speedrun. There are many brilliant strategies that can be used and new ones are always popping up. Minecraft really is a beautiful game where almost anything can be done, and some players like to choose the destination of speedrunning. I myself don't speedrun that often but I'm still intrigued by it. Props to you for being an old OG player too, that's pretty epic.

@Fluffy White Llama research on what? I've played minecraft since about a week after rails released if that's what you mean? That's pre-alpha. So, I'd say I'm pretty experienced in the game and more than qualified to provide my opinion, which is that there are better features of minecraft than its questionable suitability for speedrunning. Minecraft has always been a journey before destination game for me, so while I respect that PvP minigames or speedrunning are a totally valid way to play an open ended game, they also make it closed ended, and that misses what makes the game so good in the first place, IMO.

How much research have you done? This sounds like an echo of what Game Theory said

All I hear is “I don’t understand anything about the subject but here’s what I think”

1.12 allowed RNG manipulation but 1.16 certainly doesn't. But if Dream actively did that, then he would be able to prove it.

That applies in normal games however in minecraft it’s not really achievable. It’s a random game seed and the rng has so much white noise it’s effectively impossible to manipulate.

This is like saying to pick any random number, sure, it is a one in an infinite chance that you picked that number, but you were going to pick one.

If I throw a dart out a plane and hit a bulls eye on the ground that’s a low probability. If I throw a dart out of a plane, find it and then paint a target around it that’s a 100% probability. You being born is the second scenario. If it weren’t you it would be some other combination of your parents DNA.

Irrelevant. There are trillions of us and the concept of "you" is too abstract. Sure, the chance of someone exactly like you being born again are ridiculous, but there was no "you" measure before you were born, so it's not a relevant standard.

Since when was ant venom a respected mathematician?

Antvenom basically bends over backwards to fudge the numbers in favor of dream. His bias corrections are silly. Also, the one in 600 million he gets is "any livestreamer ever having been as lucky as Dream when you include the non sus set of runs"

We’re u in 1.16.3-1.16.5 because they changed pearl rates

Yeah, that's pretty unlucky. Around once in 1400 attempts you're going to get at least that screwed.

Hilarious and original