Past Gen RNG Research

[LightningFusion]

This allows you to send up to two Individual Values to your baby Pokémon, leaving the rest determined the same way as in the past with some random and some selected randomly from the parents. However, if you do it with both parents, you will not be able to control the baby's Nature through the everstone method.

The pokemon IVs are still generated the old way, and with the power items, those IVs are passed no matter what with the rest generated the old way. So that can cause problems, but there is no change in the breeding.

Syberia tested having 2 parents with a different item each and see if both IVs would pass along with the average 3 from the 2 parents, but resulted with just 3 IVs passing each time. This proves that holding the power items only force those IVs to pass and 1 is still passed from the parent, and the last 3 generated by the breeding RNG.

 

[ΩDonut]

I mentioned a while back that we could use the pattern of responses you get from talking to your Pokemon to verify your seed, like with coin flips.

This might be too complicated to implement in RNG Reporter at this time. For one, we don't even know how the RNG number corresponds to each response. Furthermore, I think the pattern changes depending on the happiness of your Pokemon AND the location you talk to them in, because happy Pokemon give happy responses and sad Pokemon... don't.

But they can still be used.

We just have to catalog the responses for useful seeds, using recently-traded or recently-caught Pokemon in a cave or something. I've used the responses from a recently-traded Espeon to verify my seed for a shiny Timid Lugia (which I still haven't caught because advancing the RNG by walking is tricky, but at least I know I hit the seed without catching it!)

I've documented the frame advances for some of the responses. These seem to stay the same regardless of the Pokemon's happiness.

Advance	Response
14	...
19	! + ...
22	dancing box
90	happy dance
29	!
99	look around
88	music note
30	look around + happy

 

[SnapperTT]

Stuff

Data

Hang on, these pokemon responeses are ambigious. I've noticed 3 differnt kinds of text that go with the same animation. Things like "music box" can mean many things as well.

I propose we compile a list of responses, and assign a id number to these for easy, unambigious referance. Then we can test these responses for frame advancement, etc.

EDIT:

I've composed a list of responses and put them here:
http://snapperthetwig.com/poke/partnerPokemonResponses.xls

This is by no means an exhaustive list (I'm yet to see a '"!" + "..."'), but it contains 12 responses for a pokemon with 255 happiness, and 11 responses for a freshly caught pokemon.

I've assigned ID numbers to responses, as this is a much more unambigious way of referring to responses. "What, exactly, is '...'?".

Please notice that sometimes a very similar (or the same) animation is used for several different responses.

From here, it would be easy to test against every response.

Cheers,
SnapperTT

 

[3.14159]

I tried to reset a nice adamant Ho-oh on my heartgold cartridge. The frame for the spread is 35 so I had Latias and Raikou roaming and had to walk 8*128steps with 4 pkm in my party. I constantly hit frame 36 ... I fainted Raikou to get to the 35 but then I landed on frame 34. So Raikou seems to give a +2 at the frames but that contradicts with other reports where you get +3 for all roamers. Perhaps you get +2 if one of the beasts or both are out? Or the frame of Ho-oh is simply acting weird how Syberia state a long time ago:

EDIT: It's definitely doing something with no syncher, but with a syncher my results may be inconsistent
EDIT2: I see what's happening. With a seed of F30004C2, no syncher, and 3 roamers out, I get a starting frame of 4. With a syncher, playing from the same save state, I get a starting frame of 5. With a syncher, walking 256 steps, I get a frame of 15. Without, I get 14.

EDIT3: Maybe my calculations were off, as now I am getting a frame of 15 after walking 256 steps with 5 pokes in my party both with and without a syncher. However, in previous testing (with different seeds), I was getting an offset of 4 with a syncher, and either 4 or 5 without. This is weird.

I will look further into it for Suicune when I hit my stupid delay for Ho-oh one day ...

 

[SnapperTT]

I've updated the spreadsheet with a couple more messages:
http://snapperthetwig.com/poke/partn...nResponses.xls

I've tried map some of the responses to frame advances, but I'm failing rapidly with a Sync (been getting nothing but garbage for the last two hours). I'll give some of the more skilled people a shot. :D

 

[mattj]

I hope it's not out of place to ask this here, but Snapper, (just so I understand what you're linking to) is that just a compilation of all possible responses?

 

[LightningFusion]

This is by no means an exhaustive list (I'm yet to see a '"!" + "..."'), but it contains 12 responses for a pokemon with 255 happiness, and 11 responses for a freshly caught pokemon.

Please notice that sometimes a very similar (or the same) animation is used for several different responses.

 

[mattj]

^^^ Thanks. As soon as my HG comes in the mail I'll start testing them for advancement too.

[edit]
I know most people here are using the ROMs and whatnot, and may not actually have the poke-walker, but has anybody looked into wether or not the pokewalker diary (in the HG/SS game, not the walker itself) could be used similarly to how the journal in DPPt is used to advance frames? Here's the link to the info over on serebii (bottom of page). Maybe I'm wrong, but it looks pretty much like the journal from DPPt, looks like the pages in it can be flipped back and forth, and includes information on caught pokemon apparently.

 

[SnapperTT]

^^^ Thanks. As soon as my HG comes in the mail I'll start testing them for advancement too.

[edit]
I know most people here are using the ROMs and whatnot, and may not actually have the poke-walker, but has anybody looked into wether or not the pokewalker diary (in the HG/SS game, not the walker itself) could be used similarly to how the journal in DPPt is used to advance frames? Here's the link to the info over on serebii (bottom of page). Maybe I'm wrong, but it looks pretty much like the journal from DPPt, looks like the pages in it can be flipped back and forth, and includes information on caught pokemon apparently.

The Pokewalker is used like Mystery Gift is -- its selected from the main game menu, and is inaccessable ingame.

Props for getting the legit game (it really is nice even if I don't understand 90% of the text :P), but I don't see why you can't test it on a rom. Heck, its probably easier as you can use save states and use AR codes without having to do all that fiddling that you have to do with an actual cartrige. If anyone wants, I can upload a save (.sav) file of me infront of Suicune for testing with.

 

[ΩDonut]

I've updated the spreadsheet with a couple more messages:
http://snapperthetwig.com/poke/partn...nResponses.xls

I've tried map some of the responses to frame advances, but I'm failing rapidly with a Sync (been getting nothing but garbage for the last two hours). I'll give some of the more skilled people a shot. :D

Try testing using the Dratini from Dragon's Den. It's Method 1, so you'll know exactly how many frames advanced.

Earlier in the main help thread I posted saying that with a certain ID\SID combination, the odds of getting a shiny Pokemon on the Pokewalker shoot up to 50%, as opposed to 0% for most combinations. This isn't the case. It's actually 1-in-32, which is still pretty decent.

It boils down to all of the Pokewalker PIDs being of the form D5FFFF??. For this example, I used D5FFFF66 and ID\SID combo 57126 and 62911.

1101111100100|110 ID
1101010111111|111 First 16 bits of PID
11111111?????|??? Last 16 bits of PID
1111010110111|111 SID

In each column (except for the last three), the number of 1s have to add up to 0, 2, or 4. So the first 5 ?s have to be 01100 in this case, making the odds 1/2^5, or 1-in-32.

 

[SnapperTT]

I've been doing investigations into frame advances due to partner pokemon responses. Its on the Second Sheet of the following Spreadsheet:
http://snapperthetwig.com/poke/partnerPokemonResponses.xls

Feel free to run your own tests, but try to get more than one data point for each response. (This is to ensure consistancy and make sure that errors don't propagate).

EDIT: I tested these responses against an Abra at the Game Corner, which I am assuming is METHOD 1. I have included the Seed and Spread information, though.
DOULBE EDIT: I'm begining to suspect that the types of responses you get vary with location.
TRIPPLE EDIT: I've discovered a new Response, and a possible frame advancement for it. I've had enough for one night, and I should get studying for exams. Have fun guys.

 

[mattj]

I've been doing investigations into frame advances due to partner pokemon responses. Its on the Second Sheet of the following Spreadsheet:
http://snapperthetwig.com/poke/partnerPokemonResponses.xls

Really great research. It certainly looks fairly consistent. As soon as mine comes in this weekend, I'll hurry up and get to a point where I can do the same.

Just to clarify, what do you mean by "time" in your spreadsheet? Is that how many seconds you were past the nearest minute, or something else?

 

[SnapperTT]

Just to clarify, what do you mean by "time" in your spreadsheet? Is that how many seconds you were past the nearest minute, or something else?

Time is seconds past the previous minute. With the seed, seconds, delay and spread info, you should have enough information to recreate the experiment.

 

[Deschain]

this isn't exaclty "PRNG" research, but i'm not sure where else to post this.

i've noticed that in every single battle i've had on Platinum, if i get frozen (10% chance) or my opponent gets frozen, the pokemon remains frozen for the maxium amount of turns.

i believe Platinum is glitched to make this happen.

has anyone else noticed this?

 

[Syberia]

I still cannot, on any seed I have tested with, get my egg IVs to appear on a reasonable frame in RNGReporter. Kazo, when you mention spreads "that work in HG/SS," do you mean to imply that they are guaranteed to work, even if I can't get anything else to show up except on ungodly high frames?

 

[Romsstar]

Yeah they are guaranteed to work.

D2080210 hit that seed
Frame 7 |
B/31/31/B/B/31 | -> This means that your Egg IVs will show up on Frame 7.
Btw: it doesn't matter if you parent is male or female if you give it 2nd the IVs will be inherited.

Use Breeding DPPt to check. It really works... I did it already 4 times...
Only thing that sucks is getting Shiny Eggs. But even that is not THAT hard.

 

[SnapperTT]

Hey guys, I've discovered an interesting anomaly. Some people have been complaining about spreads "not working", and some spreads "working". I've managed to hit working and not working spreads. I only realised this after hitting this:

A: 22/31/19/7/27/31
B: 30/14/31/8/1/31

SEED: 18000206; FRAME: 1

RNG REPORTER PREDICTION:
3/26/19/B/B/2
ACTUAL:
3/26/31/8/27/2
= 3/26/B/B/A/2 ?

No items were equipped to the parents

I suspect that Breeding in HGSS follows a different method than in DPPt, (maybe due to the power items buisness?). Anyways, the HGSS Breeding RNG Method Thingo Magjigy is probably the same as the one in DPPt but with a slight tweak. As such, it may be possible to correct this with another tweak.

 

[mingot]

@ SnapperTT -- Yes, breeding seems to have changed. Likely the base algorithm is the same, but there are likely extra RNG calls before the ones that are actually used. I hope. From your test, it seems pretty clear that the base IV layout follows the same pattern.

@ Syberia -- It's not that you are getting something on a high frame, it's that the algorithm (be it extra rng calls before or something deeper) are different.

 

[Romsstar]

@mingot: So do I get that right that in DPPt I called the RNG with the happiness application and the coin flips and this advanced the egg frame? (for shiny not IV)
Cause I tried many things and nothing seems to advance the egg frame at all. I always land on the first egg of the seed no matter what and can't advance it. only way is taking the egg and take all eggs afterwards till I came to my shiny egg. means frame 50: take 49 eggs before. which can be sort of frustrating btw:/

 

[Syberia]

@ Syberia -- It's not that you are getting something on a high frame, it's that the algorithm (be it extra rng calls before or something deeper) are different.

I understand that, but it just frustrates me to no end that others are getting the DPPt IVs to work, and I have not once gotten anything that matches up, on any seed. I'm even starting to think that, however unlikely, there may be a difference in mechanics between Heart Gold (which LF used, and got to work) and Soul Silver, which I am using and cannot get to work. I may test the exact same breed with the exact same parents (hell, even the exact same save file) on Heart Gold later today just to see if it works there.

 

[mingot]

I understand that, but it just frustrates me to no end that others are getting the DPPt IVs to work, and I have not once gotten anything that matches up, on any seed.

Who, other than LF, are you refering to as "others"?

@ Rommstar -- As someone who has abused a shiny egg, let me ask you, do you normally find that the IV frames are the same as reported by RNG reporter? Most people have had no luck with it, but you hatched a flawless monster. We are trying to figure out what is going on here.

 

[Syberia]

I'm not sure, but I've seen other flawless HG/SS breeds going around. Didn't Kazo do a couple? I'm also wondering if those seeds/frames he posted for Hidden Powers are verified to work, or if he just did Time Finder for them and hasn't tested them?

 

[mattj]

I'm not sure, but I've seen other flawless HG/SS breeds going around. Didn't Kazo do a couple? I'm also wondering if those seeds/frames he posted for Hidden Powers are verified to work, or if he just did Time Finder for them and hasn't tested them?

Yeah they are guaranteed to work.

D2080210 hit that seed
Frame 7 |
B/31/31/B/B/31 | -> This means that your Egg IVs will show up on Frame 7.
Btw: it doesn't matter if you parent is male or female if you give it 2nd the IVs will be inherited.

Use Breeding DPPt to check. It really works... I did it already 4 times...
Only thing that sucks is getting Shiny Eggs. But even that is not THAT hard.

 

[Romsstar]

@mingot: Yeah the frames are the same for Kazos Spreads. (Only with Kazos so far) But if you use a spread where you have an A parent as well(As I use only Kazos B only spreads for flawless) I noticed that sometimes it sorta seem to match with the Split spreads of Emerald. Sometimes split, sometimes alternate I can't tell what it influences.
So I can tell you: If you have a spread where you have both, A and B inherits, your IVs will probably be screwed. Also I noticed, when I used IV pass items (like power bracelet and stuff), it also screwed up the IVs. Couldn't find then the Frame at all. But I figure that if you use the B spreads only it will be fine even for Hidden Powers although I didn't test yet but will do so.
Actually we won't need any other spreads do we?
But in theory I think: you gotta look for a spread where you have only B parents to be sure that your IVs will actually be the way you want them. But well as I said; didn't test it yet.

@Syberia: Actually I can't tell about the Hidden Powers yet but I plan to test some oft them today or tomorrow so I probably can tell you more by then.

@mattj: yeah the flawless Spread totally works. So if you have an flawless Poke as your B parent you will get flawless pokes. But what Syberia actually meant I think are the Hidden Power Grass, Electric and Fire and stuff spreads. As I abused only for flawless so far. Kazos Spreads are intented to work with a flawless B parent. Although I reckon you probably could also use the flawless spread if you have a parent with a Hidden Power that might match the three 31s.

 

[Syberia]

I was testing a different spread with all B inheritance (sorry I don't remember the seed, and since it didn't work, I didn't think to write it down), and it didn't work, though. So I'm not sure that "all from one parent" has much of anything to do with it.