Why event pokemons have different IVs? Shiny Celebi? (GEN III)

[aire]

I thought that the event pokemons can not be shiny because their have the same PIDs but it seems that every time that you take a pokemon via event a new set of IVs are generated automatically. Also the GameCube in game events. So it is possible get a legit shiny Celebi?

I talk about 10ANIV or AGETO Celebis for example, because Jirachi CAN be Shiny.

[theSLAYER]

17 minutes ago, aire said:

I thought that the event pokemons can not be shiny because their have the same PIDs but it seems that every time that you take a pokemon via event a new set of IVs are generated automatically. Also the GameCube in game events. So it is possible get a legit shiny Celebi?

I talk about 10ANIV or AGETO Celebis for example, because Jirachi CAN be Shiny.

Those events have their anti-shiny feature working. So it's possible to get random details, but absolutely no shinies there.

Additional context: Do remember that Channel and WISHMAKR Jirachi used a different method of generation, and different method of distribution. IIRC, they tried to implement anti-shiny but borked the code somewhere.. GBA carts distribution isn't that case.

[aire]

Thank you, i see. But how it is know that the distribution cartridges have shiny-lock?, the celebi Mintsurin/10ANNIV and PokéPark cartdriges are no revealed yet.

Perhaps they are similar to the Aura Mew or the 10ANNIV without Celebi?

[theSLAYER]

6 minutes ago, aire said:

Thank you, i see. But how it is know that the distribution cartridges have shiny-lock?, the celebi Mintsurin/10ANNIV and PokéPark cartdriges are no revealed yet.

Perhaps they are similar to the Aura Mew or the 10ANNIV without Celebi?

To my understanding, there's a few things working hand in hand.
Firstly, based on analysis of the existing cart's code, and the underlying RNG of the stats of the third gen mons distributed by cart, it appears that the cart distributions were effectively using the same code as previous iterations. Meaning, they just tweaked the species and OT for each unique event, while the code for distributions remained mostly unchanged. [In other words, we don't need some of those unreleased carts to be able to tell their shiny potential, if they share RNG with previous known events. It is likely that they were distributed on a cart that was based off a known event.]

Next, event distribution ran on a "restricted seed" basis, which means there could only be up to 255 unique copies of each mon for that particular event (IIRC). This doesn't fulfill every PID combination, and could be surmised that it is technically pretty rare for a PSV and TSV combo to be even shiny. (and remember, this is on top of anti-shiny code)

On top of those 2 points (and this is more of a throwaway point), it has been what, more than 15 years? If a legitimate shiny for those exists, we probably would have heard about it.

Remember, we are talking about algorithmic precision, combined with the likelihood that most distro carts were a derivative of a previous working distro cart, combined with supremely extreme low odds, combined with over a decade's worth of time passing.

[aire]

3 hours ago, theSLAYER said:

To my understanding, there's a few things working hand in hand.
Firstly, based on analysis of the existing cart's code, and the underlying RNG of the stats of the third gen mons distributed by cart, it appears that the cart distributions were effectively using the same code as previous iterations. Meaning, they just tweaked the species and OT for each unique event, while the code for distributions remained mostly unchanged. [In other words, we don't need some of those unreleased carts to be able to tell their shiny potential, if they share RNG with previous known events. It is likely that they were distributed on a cart that was based off a known event.]

Next, event distribution ran on a "restricted seed" basis, which means there could only be up to 255 unique copies of each mon for that particular event (IIRC). This doesn't fulfill every PID combination, and could be surmised that it is technically pretty rare for a PSV and TSV combo to be even shiny. (and remember, this is on top of anti-shiny code)

On top of those 2 points (and this is more of a throwaway point), it has been what, more than 15 years? If a legitimate shiny for those exists, we probably would have heard about it.

Remember, we are talking about algorithmic precision, combined with the likelihood that most distro carts were a derivative of a previous working distro cart, combined with supremely extreme low odds, combined with over a decade's worth of time passing.

Thanks for the very detailed explanation, I think I understand a little better the generation mechanism, if the window of available PIDs is so small, it is very unlikely that there is a shiny one, but it could be, of course. You say only 255, but if I create one that looks legit, I understand that just by looking at the PID you could check that it doesn't match the supposed limited seed.
Anyway the anti-shiny seems to me a very rudimentary procedure but that way it could be effective.

I would like to know a bit more about the back-end of event distribution and triggering, it has always seemed to me a mysterious and attractive topic.

[theSLAYER]

5 hours ago, aire said:

Thanks for the very detailed explanation, I think I understand a little better the generation mechanism

no worries :3

Quote

if the window of available PIDs is so small, it is very unlikely that there is a shiny one, but it could be, of course

Well, that's ignoring the anti-shiny mechanism. Yes, the way probability works, is that nothing is definite.
There's always a chance of something to happen, no matter how abysmal

5 hours ago, aire said:

I understand that just by looking at the PID you could check that it doesn't match the supposed limited seed.

lf the RNG mechanism is known, then yeah it could be checked upon whether it was called from the restricted seed

5 hours ago, aire said:

Anyway the anti-shiny seems to me a very rudimentary procedure but that way it could be effective.

I can't recall the details, but basically if the PID results in a shiny, flip the furthermost bit, and it'll result in a non-shiny.
(either that, or reroll the PID, but I don't think it's this)

 

5 hours ago, aire said:

I would like to know a bit more about the back-end of event distribution and triggering, it has always seemed to me a mysterious and attractive topic.

I kinda get it, it is pretty fascinating.