So... Player 1 is teleported into another dimension where he fights against OP garbage. And Player 2 is teleported into another dimension where Player 1 is possessed by a bot. ... NOICE! :DDDDD
It's possible to emulate the Mobile System GB somewhat and, if the Virtual Console patch is anything to go by, then you won't need a Mobile Adapter GB if Japanese Crystal does come out with the mobile features still available.
Archiwa telewizyjne (rezerwowy) They could just fix the ROM that they use for the digital VC version of Crystal, however, this would lead to possible incompatibilities with the original cartridge, so some glitches would need to stay active when dealing with these versions (Crystal uses some of the Gold/Silver battle glitches to keep compatibility with the cartridge versions of those games).
Nintendo would probably shut off the online features anyway if anyone injects some save/their own patch with names/stats to break the trade server or something
I have looked for this information since around the time the games came out, and never found anything until now. it was always hard to word when searching. all i remember was battling my brother through a link cable until eventually our battle gets "interrupted" and we each battle a pokemon with the color of a ditto, in my case a Tangela in Ditto color. so much satisfaction after all these years
If I may, I might recommend a visual identifier of some kind on the text box so readers will known when you're about to change the text. I think a simple countdown from 5 in the bottom right corner would suffice.
Just wondering: would it be possible to at least partially resynchronize games if right player would change pokemon when left player knocked out a glitch pokemon?
This is incredible! My best friend and I had the glitch with the level 205 Butterfree years ago (among other things)! I never expected to see video of it being recreated. This brings back so much nostalgia!
Interesting how wonky Beat Up is. BTW, I've been curious about something in the first 2 generations for a while now. Why is it that enemy Pokémon can sometimes miss with 100%-accurate debuff moves such as Growl, Leer, and Tail Whip, even if nothing has happened to affect accuracy or evasion? (If that's something you'd already covered in a video in the past, can you link me to that video?)
No, those debuffs miss far more often than 1/256 of the time when used by opponents. It's more like about half the time or so (not sure exactly). And, they can miss like this in Gen. 2, which doesn't have the 1/256 miss glitch.
From a generic scenario, definitely not. You might be able to revert an specific desync process if you are very careful and the rng streams have not gotten very out of sync, but in general it's just not possible to resync it completely.
The best way I can think of is for both players to switch pokémon. That will at least get rid of the nameless glitchy monstrosities generated by desyncing.
Simply put, the one Gameboy thinks the Blissey fainted, while the other thinks that it didn't, causing the two linked Gameboys to have different results, and memory corruption.
Dude, so I totally remember seeing a Tangela like that when I was link battling my sister years ago. It's crazy that such a precise environment happened by accident to us. Does this only happen with Beat Up? I don't remember using that move.
No, it's not specific to Beat Up. It's a common pattern to micro communication errors in these games. Don't know the specifics but it's probably related to what the receiver interprets as the received data (or lack of it). Games sent commands to each other and when the receiver interprets one that is out of bounds, something unexpected (but usually consistent) will happen.
Nice exploit! But what is it that causes the desynchronisation? Does one Gameboy execute the move and tell the other what happened (and got this wrong)? Or do both execute separately and use a slightly different algorithm?
What happened was that the Beat Up on the left side had 3 hits (due to data being read incorrectly because of an error in the coding), while the Beat Up on the right side read the data correctly and had the correct 2 hits. This causes a discrepancy in the remaining HP of the attacked pokemon; because Blissey fainted on the left side but not the right, the left game tried to interpret what was happening and sent out a garbage mon because the right side didn't actually switch into anything. From my understanding, during a link battle the games share situational information. Like, if 1 game says that the pokemon in slot 0 used the move in slot 2, the 2nd game knows what that is based on the situation. But if they get desynchronized then what slot 0 move 2 is for game 1 is something different for game 2. They're still communicating, but the data no longer lines up.
CoopersCrazy Ah ok. I was still wondering what caused the difference in running beat up in the first place? Why does one game have a correctly-programmed version of the move, while the other does not? I assume they'd have to share more than just which move was used, so for example if double slap is used, it has to be seen to hit the same number of times, and critical hits etc on both sides
I'm not sure, I imagine the difference is that the sent information is correct, but the error is in what happens in the user's game. Maybe Crystal_ can provide a more in-depth answer
I could guess that the data for your opponents team is probably stored in a different maybe "cheaper" way (you can't read your opponents team stats, attack moves, pokemon names etc. all the time) and it could be stored in a different place in memory as well. Which means the method used to read the data of your opponents team could be different to the one used to read your own teams data since data which is stored in a different way (and in a different place) might need a different handling. This is especially true because a link battle is a special case that probably uses a entirely different way of reading and storing your opponents data as the rest of the game.
A link battle's integrity is based on the fact that same action + same rng = same outcome (link battles use a shared prng stream that is previously generated in one side and sent to the other). Of course, this implies that player side code and enemy side code *performs* identically, or link battles would desync easily. The *implementation* of each side has to be strictly different however, if only because player variables and enemy variables are stored in different places. Therefore, each move effect is implemented in the game twice, once for each side (even though most move effects can reuse most of their code for both sides). This beat up glitch is caused by the player side implementation of the move not performing 100% identical to the enemy side implementation.
why couldnt they just sync on each animation? the data isnt that large so even 9600 bauds could probably be more than enough to send the overhead plus the current battle state plus pokemon states.
the gb link protocol works like, now I send a byte, then you send a byte. It's not meant to send big payloads of data to each other. plus even if you could, that's not how pokemon battles work, you execute an action that leads to things happening. sending the battle state to the other side won't be of much help if you don't tell him what led to that state, because then the only thing the other side can do is abruptly changing the displayed information.
"There is a possibility that you haven't understood a thing of the explanation I provided."
Truer words have never been spoken.
Lol me all the time on this channel
So... Player 1 is teleported into another dimension where he fights against OP garbage. And Player 2 is teleported into another dimension where Player 1 is possessed by a bot.
...
NOICE! :DDDDD
This is what happens when gen 2 lacks the Synchronize ability
Oh, you.
ua-cam.com/video/ObpcGNCU944/v-deo.html
Ugh I hate it when I make my pets fight to the death but one attacks when they weren't supposed to and then the universe breaks down
....A Butterfree. With a shitton of Stati. On lvl 205 with the cry of a Magicarp.
Don't see anything wrong.
I think that's a Bellsprout's cry. I guess the evil lurking in Pokemon Crystal really likes Grass types.
Or they must love sprout tower haha
I always forget how different Sneasel looked in gen 2. Also probably due to no one ever seeing one in he games
Emmett H your rival has one in the games.
4:54 at least here, they are """synchronized"""
Imagine desyncing a mobile battle in Japanese Crystal, especially considering there is a limit of 10 minutes.
It's possible to emulate the Mobile System GB somewhat and, if the Virtual Console patch is anything to go by, then you won't need a Mobile Adapter GB if Japanese Crystal does come out with the mobile features still available.
Archiwa telewizyjne (rezerwowy) They could just fix the ROM that they use for the digital VC version of Crystal, however, this would lead to possible incompatibilities with the original cartridge, so some glitches would need to stay active when dealing with these versions (Crystal uses some of the Gold/Silver battle glitches to keep compatibility with the cartridge versions of those games).
They don't actually touch the ROMs for VC, they simply instruct the emulator to apply patch files during gameplay.
Nintendo would probably shut off the online features anyway if anyone injects some save/their own patch with names/stats to break the trade server or something
@@Flametix except you can easily generate ant Pokémon with any OT in Gen 1.
I have looked for this information since around the time the games came out, and never found anything until now. it was always hard to word when searching. all i remember was battling my brother through a link cable until eventually our battle gets "interrupted" and we each battle a pokemon with the color of a ditto, in my case a Tangela in Ditto color. so much satisfaction after all these years
1:30 you really know your audience.
Hey Crystal_, what happens if you use Beat Up with more than 6 pokemon in the party? I think that would make for an interesting video!
If I may, I might recommend a visual identifier of some kind on the text box so readers will known when you're about to change the text. I think a simple countdown from 5 in the bottom right corner would suffice.
I wonder why more people don't already know about this...?
Oh, right, because Beat Up is an absolute shit move and nobody ever uses it.
Actually, people do use beat up, to hit justified pokemon on their side in a double battle to get a massive attack boost
Except in the TCG.
Beat Up deserves a buff, possibly a 20 BP move instead, 120 BP with full team is a better option.
It's cool to see this glitch explained and in action. Who cares about the viability of the move?
@@8Kazuja8 Actually someone made a gimmick team of a choice band weavile with beat up + a bunch of high attack mons and it was pretty damn successful.
congrats for reaching 6.000 subs, keep it up with such amazing content that you provide us along those years, i love your channel
Thank you! You noticed it before I did xD
Just wondering: would it be possible to at least partially resynchronize games if right player would change pokemon when left player knocked out a glitch pokemon?
This is incredible! My best friend and I had the glitch with the level 205 Butterfree years ago (among other things)! I never expected to see video of it being recreated. This brings back so much nostalgia!
Shit like this is exactly what creepypastas are made of.
Interesting how wonky Beat Up is.
BTW, I've been curious about something in the first 2 generations for a while now. Why is it that enemy Pokémon can sometimes miss with 100%-accurate debuff moves such as Growl, Leer, and Tail Whip, even if nothing has happened to affect accuracy or evasion? (If that's something you'd already covered in a video in the past, can you link me to that video?)
SomeGuy712x It's a mistake, 1 in 256 chance to miss.
No, those debuffs miss far more often than 1/256 of the time when used by opponents. It's more like about half the time or so (not sure exactly). And, they can miss like this in Gen. 2, which doesn't have the 1/256 miss glitch.
SomeGuy712x Oh, how odd then.
Yes, stat lowering and status moves have an independent 25% chance of missing when used by a non-link non-battle tower opponent.
Huh. Alright, thanks. I'm guessing it was programmed that way to help the player out during the story, but it does seem rather odd.
Is there a video is document listing the glitches of desyncronizing a link battle in GSC?
Would there be a way to resync the battle?
From a generic scenario, definitely not. You might be able to revert an specific desync process if you are very careful and the rng streams have not gotten very out of sync, but in general it's just not possible to resync it completely.
The best way I can think of is for both players to switch pokémon. That will at least get rid of the nameless glitchy monstrosities generated by desyncing.
I've never seen what the old link battles used to look like, wow! Thanks, dude! This is super interesting😃
i'm playing pokemon (EV grinding for the elite four) while watching this :)
I was wondering how much damage would it do with maxed out attack, but too bad it ignores temporary boosts entirely.
Simply put, the one Gameboy thinks the Blissey fainted, while the other thinks that it didn't, causing the two linked Gameboys to have different results, and memory corruption.
Dude, so I totally remember seeing a Tangela like that when I was link battling my sister years ago. It's crazy that such a precise environment happened by accident to us. Does this only happen with Beat Up? I don't remember using that move.
No, it's not specific to Beat Up. It's a common pattern to micro communication errors in these games. Don't know the specifics but it's probably related to what the receiver interprets as the received data (or lack of it). Games sent commands to each other and when the receiver interprets one that is out of bounds, something unexpected (but usually consistent) will happen.
See kids?
Thats why structs were invented in C.
Because messing up your offset is not a good thing...
bugs are possible in C too, and it's possible to fix bugs in assembly
Metronome sometimes desynchronize link battle too.
Using "Beat up"
@@clonefighter1996 Probably, yea
@@clonefighter1996 I remeber it happening to me a long time ago, so I'm unsure whether it desyncs due to beat up or other sloppy code
Hey! I remember seeing this glitch. Not the Beat Up thing but the one where you face a level 0 Tangela.
Can you make a video on how the Pokemon that has data read from an errant value is composed?
I did that for RBY if you're interested: ua-cam.com/video/XYIUaPKlA7s/v-deo.html
Irrelevant statement. Snake game seems almost compatible with English Crystal.
Also, how am I supposed to write to DA10/FA10? I just wanna walk through walls for fun.
oh yeah I forgot you could enter a link battle pre-damaged in GS
My brother and I had this happen once, was freaky at the time.
Nice exploit!
But what is it that causes the desynchronisation?
Does one Gameboy execute the move and tell the other what happened (and got this wrong)? Or do both execute separately and use a slightly different algorithm?
What happened was that the Beat Up on the left side had 3 hits (due to data being read incorrectly because of an error in the coding), while the Beat Up on the right side read the data correctly and had the correct 2 hits. This causes a discrepancy in the remaining HP of the attacked pokemon; because Blissey fainted on the left side but not the right, the left game tried to interpret what was happening and sent out a garbage mon because the right side didn't actually switch into anything.
From my understanding, during a link battle the games share situational information. Like, if 1 game says that the pokemon in slot 0 used the move in slot 2, the 2nd game knows what that is based on the situation. But if they get desynchronized then what slot 0 move 2 is for game 1 is something different for game 2. They're still communicating, but the data no longer lines up.
CoopersCrazy Ah ok. I was still wondering what caused the difference in running beat up in the first place? Why does one game have a correctly-programmed version of the move, while the other does not?
I assume they'd have to share more than just which move was used, so for example if double slap is used, it has to be seen to hit the same number of times, and critical hits etc on both sides
I'm not sure, I imagine the difference is that the sent information is correct, but the error is in what happens in the user's game. Maybe Crystal_ can provide a more in-depth answer
I could guess that the data for your opponents team is probably stored in a different maybe "cheaper" way (you can't read your opponents team stats, attack moves, pokemon names etc. all the time) and it could be stored in a different place in memory as well.
Which means the method used to read the data of your opponents team could be different to the one used to read your own teams data since data which is stored in a different way (and in a different place) might need a different handling.
This is especially true because a link battle is a special case that probably uses a entirely different way of reading and storing your opponents data as the rest of the game.
A link battle's integrity is based on the fact that same action + same rng = same outcome (link battles use a shared prng stream that is previously generated in one side and sent to the other). Of course, this implies that player side code and enemy side code *performs* identically, or link battles would desync easily. The *implementation* of each side has to be strictly different however, if only because player variables and enemy variables are stored in different places. Therefore, each move effect is implemented in the game twice, once for each side (even though most move effects can reuse most of their code for both sides). This beat up glitch is caused by the player side implementation of the move not performing 100% identical to the enemy side implementation.
why couldnt they just sync on each animation? the data isnt that large so even 9600 bauds could probably be more than enough to send the overhead plus the current battle state plus pokemon states.
the gb link protocol works like, now I send a byte, then you send a byte. It's not meant to send big payloads of data to each other. plus even if you could, that's not how pokemon battles work, you execute an action that leads to things happening. sending the battle state to the other side won't be of much help if you don't tell him what led to that state, because then the only thing the other side can do is abruptly changing the displayed information.
Who needs a good protocol, when you have a roll of cheinese tape laying around(aka the "protocol" Gen 1, 2 games use)?
-GaaeFreak
4:36 Gigamax Butterfree
A video without dislikes!!!
And it still is!
@@limecyanizer4394 Not anymore (Not because of me)
Wow this is preatty cool
Is this another case of the wrong branch instruction being used or something similar?
Devs just screwed up the variable-reading.
So instead of reading the field with pokemon id, Beat Up's code reads the first byte of health variable.
wrong comparison
You're amazing
Hah thats awesome
This video is super loud lol
Vengó de parte de *_Nigueru Willian_*
X2