The Math Behind the Unluckiest Moments in all of Pokémon 

I actually see it

tysm

lmao

I agree, and not in a insulting way. But those who see it see it and those who don’t are just Yanny/Laurel or whatever 🎉

😂😂

Speedrunners know their ways around calculators. The tools used in making TAS (Tool Assisted Speedruns) and absolutely dissect the coding of games for intended and unintended consequences blows my mind. Aside from how many hours they spent in those grass patches to just kinda know what's normal/not and by how much, they mentioned the approximate odds of an encounter in each tile during the run too-- so I'm sure it wasn't too much of a mystery that it was absurd.

Absolutely, also shen might be the person who has spent the most time playing gens 1-3. If it has happened to anyone it makes sense it would be him

@@jazzisfrommars2476 Bruh. Even a 6 year old playing RBY for the first time would realize something was wrong with that grass lol...

shen is a math teacher, i believe. couple that with literal thousands of hours playing the earlier gens of pokemon, he is definitely the guy to know how absurdly improbable that was lol

@@mhf0gamer Again, even a 10 year old who has played Pokemon once or twice, would suspect something was wrong if no wild Pokemon appeared for that long. The incident may have unintentionally created with the 'Manip technique'. Or, if you want to be really cynical, the could have intentionally used the 'Manip technique' to do this to make an interesting youtube video.

Yeah, also is gen 1 we're talking about

@@mellevala which is why we know this isn't a gen 1 moment, and is just insane luck

yeah i was gonna say the same its almost definitely a weird/rare bug of some sort, i dont think people actually realise how unbelievable that would actually be

@@PrismaAce No, you don't know that. The commenter pointed out that because of how broken gen 1 is, anything involving such astronomical odds is likely a bug and not ''insane luck''. Another thing that confuses me is how seemingly unique this clip is. Has anyone been able to find similar clips of unlucky tall grass? Not 1 in 3 trillion odds, but even something like 1 in 1 million, or 1 in 100k? There should be plenty, no? Gen 1 has been played by tens of millions of people over the span of 2 decades. Tall grass is a very common tile.

Haven't we reverse engineered the gen 1 code already? If there was such a mechanic, we would have known.

oh hey, its ajp! how r u?

Literally just came from watching and completing your 8f shiny mew tutorial

@@Gexthereptile did you get the shiny mew?

@@nept2ned yea, along with a shiny mewtwo and the legendary birds, after that 8f stopped working and just caused my game to crash, still insanely happy tho

@@Gexthereptile nice, did you transfer them or nah you should probably clone them aswell if you want

His breakdown of the phoenix club statistics had some questionable choices but otherwise was good

This is genuinely the nicest comment I've ever received. Thank you!!!!!

Masters student here, completely agree. This guy is amazing at explaining math!

@@tannerboos2268 absolutely - good luck on the masters!

Good luck on the doctorate!

He admitted it was in fact cheated tho, just unintentionally

@@PbnjGuy Yeah

@@PbnjGuy "unintentionally"

​@@PbnjGuyAfter all the time he spent saying it wasn't cheated? Yeah I don't exactly trust the guy

@@PbnjGuyHOW DO YOU ACCIDENTALLY CHEAT

I believe I've gotten it twice, once in Omega Ruby and once in Platinum

I don’t blame you, I believe it’s rarer than shiny pokemon. The one time I encountered it was Pokemon Platinum and Soul Silver, I didn’t even think about the probability of a Pokemon getting it. Not even the probability to find it in both games lol

I wish he'd talked a bit more about how if you see something and then say "Wow that's unlikely", it's much less unlikely than if you predicted it. Like, if you roll a million sided die and get exactly a 69 that's one in a million. But there are so many more values that would have made you say "Wow, that's one in a million" like if it rolled a one, a one million, a 420, a 42069, ect. The odds that the die will roll a number are 100%, even though the odds it will roll whichever particular number it rolled is one in a million.

Emphasis on "almost"

#BatmanWasRight

Well I tried to get a shiny female combee, it took me about 100 eggs and I didn't get any males. My friend did about the same amount of eggs and got 3 shiny Beldums.

I once got a shiny on the 6th egg, but 3 shinies in 100 eggs is really absurd luck, even considering the masuda method odds of 1/512.@@SilentHotdog28

It's roughly 1 in 725 trillion. I counted 79 steps.

@@valkyrie1994 to be honest that is so unlikely, it might have been a glitch

@@Koni.1122 honestly that's pretty likely. Seems like it. Maybe a weird bit flip glitch or something

​@@valkyrie1994alrighty, get someone figuring the odds of that one out

@@jamesringo7070 Wish I could. Don't know nearly enough about programming to even begin.

i found pokerus before my first shiny

I don't know if I caught a Slakoth with pokerus or it got it on the way to box it not knowing it had the virus, but I never had any pokerus event after emerald.

@@MrNigel117 dude me too i got pokerus two or three times on different pokemon games before getting a shiny

I got natural pokerus once or twice as well!

Lilith? Wdym

Same

if my teacher had used it i probably wouldnt have almost failed high school

It's in the book ...

​@@arturocevallossoto5203There's a book?

Tbh, I found it over the top complicated for just talking about a bunch of numbers divided by 100.

Yes exactly, the odds mentioned in this video only apply if you went into the randomiser with the goal of getting those 3 pokemon specifically lol. Otherwise getting the 3 birds is just as unlikely as getting Gloom, Smeargle and Murkrow

@@reio1951 Yes and No. What i meant to compare it too is it beeing as impressive as getting another trio of that kind like Uxie, Azelf and Mesprit or a full evolution chain like Charmander, Charmeleon, Charizard.

@@xXhiXxification false, probability wise it’s the exact same. If you had some special link to Gloom Smeargle and Murkrow itd be no different than it is now where you don’t

​@@reio1951That's not the point though. Randomly getting 3 unrelated pokemon is the expected outcome, because randomly getting three related pokemon is pretty rare. It's not as rare as the number shown though, because there are other trios of related pokemon that would get the same reaction from a player going "omg what are the odds". Of course the calculation is right for the birds especially, but for the general event of getting 3 noteworthy related pokemon to show up happening the probability is somewhat higher.

​​​@@reio1951The probability he calculated applies to any given triad of Pokemon, not just the legendary birds. It doesn't only apply if you have them in mind specifically.

As a software engineer I was looking for this comment

Yes

Yeah, it's just much much more likely that any kind of bug happened, even a "super rare" one than this one in 3 trillion thing happening.

you do realize truly random number generators are also capable of outputting long strings of high or low?

@@nodrance You do realize that "truly" random number generators don't exist? And that's not even getting into the quality of a gen 1 psuedo-random # gen

Underrated comment

Yeah, I feel you, it’s like in Dungeons and Dragons where players will gasp getting two mat 1s in a row and say “there’s a 1 in 400 chance of that happening!!” But technically there’s a 1 in 400 chance of getting any two combinations of numbers so that doesn’t actually mean much. There’s a 1 in 400 chance of getting a 2 then a 13 but nobody freaks out about that lol

@@rwbyab7423 That's because the two natural 1's are a noteworthy combination, much like the bird trio. It means something because it's the worst possible outcome of 2 rolls. Getting a 2 and then a 13 is like getting... idk like phanpy, rhydon and flaafy as your starter selection. Of course no one's gonna freak out about that, it's unremarkable.

@@ZerrieGD fair, perhaps an example more in line with the original comment would be to claim it is equally unlikely to get two 13s in a row or two 7s in a row. While less noteworthy, the odds of getting those specific three Pokemon is equally astronomically low and, now that you've randomly named them off, if it ever occurred I'd be very surprised!

Notable trios (13): evolved starters (3), gen1 eeveelutions (1), legendary birds/dogs/giants/fairies (4), 3/4 pseudos (4), weather trio (1) The chance of getting a notable trio is 3/386 × 2/385 × 1/384 × 13 = 0.0001%.

Not too unlikely, plenty of the tiles in there actually can't generate encounters.

Yeah my go to was some sort of glitch haha. Even with all the hours spent playing Pokemon, 1 in 3 trillion is enough to where there's a very good chance it shouldn't have happened. Insane stuff.

​@@SpookySkeletonGangWith that kind of luck he could have won the lottery jackpot.

In order for that to happen, the seed for that event sequence has to exist. Assuming it does, which was very improbable, it's actually not nearly that unlikely. If it does exist and wasn't a glitch, it's likely that several other people have also encountered that and just not recorded it or realized how unlikely it was.

@@AWanderingSwordsman yeah, I doubt the rng is large enough for those odds to make sense. either the sequence exists or it doesn't and that is what ends up determining the likelihood for some more common events we can assume the rng is truly random I think, but for stuff like this I doubt it

@@SpookySkeletonGang Considering the games have been out for decadess, re-released on the e-shop, have been played by millions, and have tons of footage that hasn't been recorded. I wonder if it's more or less likely to have happened to someone else? With the amount of play times of these game it's got to have happened to someone else or at least be close right?

this is true but you could argue the odds of getting any string of pokemon with exact shininess, ivs, nature, and PID is extremely rare. probability only matters when you have a desired outcome, and having 2 shinies in specific slots was not a desired outcome here, having 2 shinies was the desirable outcome that occurred.

@@hsk5363 Agreed. Odds can be very deceiving - be it on accident or otherwise

According to my probably inaccurate calculations it's 1/16M

@@HelloIAmAnExist i manually worked out my odds, but just confirmed it correct with binomial distribution. Not too sure where you went wrong, happy to check if you send through your maths

@@jnjsorr I believe you, just wanted to share what I got

Speedrunners are already routinely manipulating encounters in gen 1 so they can do it fairly easily if they choose. The world record gets no encounter in Mt Moon which I suspect is a similar number of steps to that safari zone clip

there is no such thing as "random" its just a term applied to an effect to witch we dont know the cause.

It'd be simple but tedious to swap the odds. Nigh impossible to catch a caterpie at 1 hp with an ultra ball but almost guaranteed to catch Mewtwo at 100% with a Poke ball

Not possible. The random number generator in the game isn’t actually a random number generator. It’s a pseudo random number generator. It starts on a set seed and advanced each frame to a different RNG tile. That’s how the speedrun TAS ensures perfect luck on metronome among other things. Given a computer’s temporal resolution it stops becoming a random number generator and starts being a rhythm game.

What does that mean though? The player has the worst possible luck? So... they just lose, then. The player wouldn't be able to reach the first town. Not much of a TAS.

@@NightKev yeah so the TAS could just wait until a frame where every move would miss and ensure every single move ever misses so the game can not be progressed.

Smallant

Werster has gone longer without an encounter in Gold too, although I'm not sure if the probability is the same. See his video "The longest 2 minutes of my life" (actually 1 minute 30 seconds).

I'm surprised Werster's stadium miss wasn't mentioned in the video. It happens just after the 17 hour 37 minute point in his stadium 1 complete the game speedrun.

I just got to the part where Shenanagans doesn't get any encounters. Werster also has a clip titled "the longest 2 minutes of my life" where he doesn't get any encounters for about 90 seconds of surfing. The encounter rate while surfing has got to be a lot less than the safari zone's rate, but he also spends 30 more seconds looking, so now I'm curious what the odds of that happening were.

@@LinkaleeLewisWerster is awesome, but he lowkey got cancelled and he’s kinda ostracized so I imagine some ppl purposely ignore him unfortunately

@@Justpassingby204 How do you lowkey cancel someone? Isn't the point of cancelling to be as highkey as possible about it?

​@@Justpassingby204over what? am new to the community and have been watching his streams, love him

Idk why he didn't explain this. All that build up for a tangent about the social network and then a segue into something completely different. Editing/scripting mistake?

@@guinea_hornCould’ve just overlooked it. Sometimes you think something is obvious and don’t realize its not so clear.

@@Clover298He did explain it though, what are you talking about? From 10:38 - 12:26

@@Zanza05ok? i commented this before watching the full video, why are you getting mad at me and no the other two who pointed this out in the first place?

Odd, the most that 4 EVs, which you need at least TWO knockouts for - in order to possibly get four EVs, is 1 more stat point at LEVEL 100. Story mode doesn't go near level 100. Is the range of tolerance really that razor thin?

I find that hard to believe, but i'll believe you since i got it 5 in a row once.

​@@ninetailedara3802I mean, considering all the crazy stuff that happened in this video, the chance of that happening seems entirely likely compared to everything else

Missing all 8 focus blasts in an online battle scarred me for life. I have not used moves below 90% accuracy ever since.

That reminds me of them time I missed steel wing 7 times in a row idk what those odds were update: it's like 1 in 10k

​@@p4leking fissure, stomping tantrum is still a great strat

More are coming :)

On point!

Exactly. "Where is the shxt of my fxxk"

Personally, I think a cosmic ray bit-flipped the top bit in this counter variable, and that is how it happened. Flipping the top bit would set the counter to 128 instead of the usual 4. So the first 128 steps would be gauranteed no encounters, which makes this a bit more likely. Of course, what are the odds that he would encounter a bit flip like this lol.

@@ZTenski or as someone also said, it may have been a bug with a previous repel, that resetted when he opened the menu. this is gen1 after all, but then again, it could really be unlucky odds

@@ZTenski it's probably more likely than 1 in 3 trillion lol

It doesn't work like that. Opening the menu doesn't have anything to do with repels. ​@@ElysiaWhitemoonOmega

@@lemonke8132 it isn't actually, cosmic rays have been known to fuck with computers and flip bits ever since the computer was first invented

P😊😊😊😊p😊😊p

Man, this comment is so unbelievably kind. I've seen a bunch of your stuff and it is all so killer!!! This is a huge compliment, thank you man!!!!

main character attitude

Larry has a Two segmented tho…

No way!!

@@thekittenwolf I mean, I did explore the entire region before going to school, due to accidently discovering the jump the gap "hack" which resulted in me catching the troublesome Dunsparce on accident as I couldn't run, so I tossed a pokeball while my last pokemon was at 3 HP, expecting it to fail and for me to get a faint screen. I also wasn't expecting it to evolve, as it never had a evolution in the past, so why now.

Sometimes that bad luck just hits different LMAO

Captain obvious reporting for duty🫡

@@thisissostupidqsdfvanah hes saying this cause stupid mfs in here dont believe that lucky events can happen

@@thisissostupidqsdfva We don't need your services today, captain obvious. Dismissed.

I fully well agree that he make have good content but I had a stroke reading this It took my dyslexic ass 10 minutes to understand I need to go back to elementary school or something

​@@bobross547they miss spelled "absolutely" and "insane" so that probably tripped you up.

That's crazy I assumed he had high subs and didn't even notice until I read this.

Wow, that's insane! And here I thought my shiny alpha luxio was rare lol 😂

I got a shiny totodile in Gen 4 with an Adamant nature. It was the starter and nature I was looking for. That's 1 in 6,014,700 The point he was making about thinking critically about probability, is that insane odds happen all the time. In fact, it'd be unlikely for unlikely things to not happen than for them to happen.

I wanna do the math on that one, holy crap Edit: ~0.0000027%

​@@Sion-wn8teis like 27/10.000.000

Ryukahr also did a nuzzlocke where he got like 5 ohko moves to hit I think.

@@thatfuzzypotato1877 The number you got is wrong. i think i know what you did wrong but im not sure. i believe you first went to check the moves that metronome cannot call in generation 3 and came back with 17 moves, then you went to check how many pokemon moves there are (maybe you tried to check specifically in gen 3 but the answer came back with the total amount) and got 900, you then subtracted 17 from 900 to get 883 and divided that number by 4 (there are the same amount of OHKO moves in gen 3 through gen 9) getting 220.75 which you took the inverse of (1/220.75) then cubed (put to the power of 3 for the 3 OHKO moves called in a row) and multiplied by .3 (the listed accuracy of OHKO moves) ONCE to get ~0.000000027 or 0.0000027%. to get the actual odds with the given accuracy of OHKO moves (OHKO moves are weird ill get to it in a bit) you would start with the 354 moves that are in fire red and subtract 17 from it getting 337 moves metronome can call then you would divide that by the 4 OHKO moves to get 84.25 which you would take the inverse of then immediately multiply by .3 before cubing it to get ~0.000000045 or 0.0000045%. however, because OHKO moves hate being simple this is not the actual chance of getting 3ohko moves in a row with metronome and hitting them all. OHKO moves actual chance of hitting is not as simple as the listed 30% as that is only the chance of them hitting if both pokemon are the same level. if they are not the same level then for each level the user is above the target the accuracy increases by 1%, so if the user is 20 levels above the target they will have a 50% chance to hit, and if they are 70 levels above they will hit 100% of the time. because of this, if you dont know the levels of the pokemon in the battle it is impossible to calculate the actual odds. thankfully, since i have no life at all, i sat through 16 minutes and 25 seconds of the video until the first OHKO move is used by a level 56 against a level 37 giving a 49% chance to hit followed by a level 41 giving a 45% chance to hit then followed by a level 35 giving a 51% chance to hit. plugging those numbers in that leaves the odds at ~0.00000019 or 0.000019% or, as expressed as a fraction, ~1/5,177,030

@Aidan-se8yk not quite the method I used I oversimplified it a bit aince when numbers get this insane the practical difference (for something like pokemon stats) is close to nil, but if I am incorrect, I concede

thank you shay!!!!

The best way to think about this is to imagine a lottely with a billion tickets that are bought up by a billion different people. The chance that you personally win is super tiny, but the chance that SOMEONE wins is 100%.

yeah I used to think it was normal that all moves have a chance to miss, mostly because tackle was missing a lot (because it's only 95% accurate) and because of negativity bias I never really noticed when a move was much more accurate than that or when in later gens I ended up not missinng with 100% accurate move. Didn't helpthat gen 1 doesn't let you see the accuracy of move and also that I didn't understand in later gen what the number represented, couldn't be percentages cause obviously all moves have a chance to miss =p Also ended up with me grossly underestimating the accuracy of not 100% accurate move, like, slam misses a lot, but I thought it missed almost all the time because instead of comparing to the chance to hit of a 100% accurate move I was comparing that with the chance to miss of these moves, and while it only hit something like a third of the time these move hits, it misses like 50 times more or something. which feels garbage. I also am honestly wondering if it's a glitch or if it's intended to be like a D&D critical miss and they just realized in gen 2 that it was a garbage idea lol

TBF, nowdays theres a distinction between attacks with 100% accuracy and the ones that cannot miss. The latter, IIRC, just completely skips the calculations to prevent this kind of issues.

@@blasecube I mean it's not just nowadays. swift also did that and it was exempt from the 1/256 glitch. That actually solidified my belief that the baseline for moves was to be able to miss because a move that can't miss was a special tm given to you.

Me when I lie for no reason

Sorry man i think that was in your dreams

@@supremepanta9564 Actually I still have the original save file and the proof! I traded the Zubat up and he is now a Crobat in a newer game but I still have the original Metapod and now a Butterfree. Since its the original save I can see they were both caught on 6/9/2017. I also dont have a completed pokedex so I dont have the shiny charm and they are definitely full odds 1/4096 Im currently in some college math classes so I finally have been able to calculate the probability of finding 3 shiny pokemon in any 4 encounters, about 4P3(1/4096^3) or roughly 1 in 17 billion. which is still more likely than the videos final example of the no encounters in the grass 3 million. However add in the fact that it came before the 10% chance Munchlax encounter which I had been trying to get for over an hour and my odds become a lot worse. im still dumbfounded by it

I meant to tell the truth, I just kinda forgot

I don't know what it is about about Sun/Moon, but I know a few people, including myself, who have encountered 2 full odd shinies in these games relatively close to each other in a playthrough. this makes me more inclined to believe your comment than I otherwise would

What was your thesis about?

​@@enoyna1001convergent beam electron diffraction in epitaxial thin films. It was a real page turner 🤣

No encounters in sprout tower is a good day for me unless I'm going for early gastly

no, it was actually right in the video. The way it was calculated is asking 230 or worse. If you would want the probability for 230 exactly it would be P(fail)^230 * P(Success), and in fact if you want you can do the calculation for the sum of all possible exact probabilities over 230 and come up with the same answer: Sum from k=230 to inf of P(fail)^k * P(success) and it should equal P(fail)^230.

​@@TorchArts This is wrong... that's just not how the geometric sums work out. Do the math with some simple test numbers.