Get any Pokémon in Emerald description.

This is me using the data substructure change glitch aka 'Glitzer Popping' glitch; a derivative of the access Pokémon beyond the sixth slot glitch. I force the EVs & Condition substructure data to be read as Growth data. The Growth substructure determines things like your Pokémon's species and species is the first word here; i.e. two bytes. In the EV & Condition substructure the first two bytes are the HP EV and the Attack EV, so what this glitch is about is converting HP and Attack EVs into a Pokémon index number.

In this video, I obtain Jirachi, through the use of 153 HP EVs and 1 Attack EV. This reads "9901" in hexadecimal, but the game reads the species bytes in reverse, so it's "0199", Jirachi's index number. If you wanted another Pokémon, let's say Mew, you need to know its index number. You can find the list here: http://bulbapedia.bulbagarden.net/wiki/List_of_Pok%C3%A9mon_by_index_number_(GBA)

Mew is hex:0097, so you need 151 HP EVs and 0 Attack EVs. Because if you convert 97 from hexadecimal to decimal using e.g. that Bulbapedia table or Windows Calculator on programmer mode via 'view', you get 151.

On to some technical details; Each Pokémon has a personality value. This is usually random. The personality value determines the order a Pokémon's data is stored. The data is split into four groups called substructures, and they are 'Growth', 'Attacks', 'EVs & Condition' and 'Miscellaneous'. The substructure order is calculated by the personality value modulo 24; see here: http://bulbapedia.bulbagarden.net/wiki/Pok%C3%A9mon_data_substructures_in_Generation_III

When you do voltage's access Pokémon beyond the sixth slot glitch; it's in this video but see also this https://www.youtube.com/watch?v=KME8eusvRAc, having your cursor on position FF up to D7 will randomly corrupt the personality values of box Pokémon. When you do the scrolling to those areas you're managing with data that have addresses that follow your party Pokémon addresses, and FF to D7 happen to manage box data with FF being position 23 in box 2 and D7 being position 1.

When you corrupt the personality value, you're likely to corrupt the substructure order due to a different new modulus, but the corruption is not too random. Highlighting the relevant position will always add a bit or two bits in the personality value if the bits aren't set. The bits are bit 0 and bit 2; i.e. +0x05 and bit 6; i.e. +0x40.

Now: In-game trade Pokémon always have the same personality values, unlike wild Pokémon. This means that you can have an idea about what the new personality value might be. It eliminates the problem of you using random Pokémon and hoping that you force the game to interpret something like the EV substructure or move substructure as Growth. When you use random Pokémon, you may have Pokémon that will never be able to have certain 'mutations' like 'EVs & Condition' into 'Growth'. You could catch a lot of Spinda and place their spots on the program A-Save to find out whether they have a good personality value, but that is not recommended. It takes a lot of time.

There is just one in-game trade Pokémon that can have its EVs read as species and it always has the potential to work! It's the Horsea called SEASOR from the house south of the Pacifidlog Town Pokémon Center where a lady is looking out at the sea. This is better than moves->species because moves only go up to 354, Psycho Boost and there are Pokémon with index numbers higher than that; 354 corresponds to Minun.

SEASOR's personality value is 0x0000007F. Take the modulus and you get a substructure order of 07; i.e. AGME. Three of eight of the 'change bit' operations give a modulo for the new personality value of 23, i.e. MEAG. Notice that 'Growth' or 'G' is now the fourth substructure in the order, instead of 'EVs & Condition' or 'E', so there's always a potential chance EVs may be read as Pokémon bytes. Furthermore, if any of the first two least significant bytes of the personality value gets changed, a Pokémon turns into a Bad EGG, but you can have the game set the bits on the upper two bytes, get the above modulus, and have it not corrupt.

For some reason, one or more of the corrupted PC Pokémon may turn into a regular EGG. If you put it into your party with all fainted Pokémon you can see its contents. If it's what you were trying to get, e.g. Jirachi, from there you can simply use your Bicycle to move around and hatch it. Only problem is that it will likely have glitch moves that freeze the game when you view their names, but you can work around this by placing the newly hatched Pokémon in the Daycare Center and moving around until all its move are replaced. Just don't use something like a Beldum that can never learn any moves for this if you want to make it stable.

Illusory Pokémon EV values:

0 Attack EVs and 151 HP EVs; Mew
0 Attack EVs and 251 HP EVs; Celebi
1 Attack EV and 153 HP EVs; Jirachi
1 Attack EV and 154 HP EVs; Deoxys

Please note that Mew and Deoxys will never obey with just this glitch due to anti cheating measures. Maybe you'd have to trade them to Ruby or migrate them?

To remove your Bad EGGs after the glitch, this altered version of the cloning glitch apparently works: http://forums.glitchcity.info/index.php/topic,6868.msg195251.html#msg195251


See more information about the cause of the glitch here. I recommend reading it: http://forums.glitchcity.info/index.php/topic,6868.msg195220.html#msg195220

I would like to thank the following people:

voltage - for discovering the access Pokémon beyond the sixth slot glitch.
luckytyphlosion - for letting me know that you can get Bad EGGs and EGGs with it.
VaeporSage - for his amazing Pomeg glitch research in general
werster - for his great experimentation with the glitch. He demonstrated that move to Pokémon corruption is possible, and later that EVs & conditon to Pokémon is possible.
Sanqui - for his research into how the glitch works.
TheZZAZZGlitch - for his research into how the glitch works.
Other people, including people on werster's stream.

As for myself, I verified simply selecting the post six-Pokémon will corrupt them into Bad EGGs/EGGs, and I documented position FF as the first corrupted Pokémon where scrolling up will corrupt them in a backwards order starting from box 2 position 23, and that ~D7 affects the first Pokémon in box 1. I didn't do that much. The bulk of the research is from Sanqui/TheZZAZZGlitch and werster.