RNG Manipulation in FireRed/LeafGreen: Wild Pokémon Supported in RNG Reporter 9.93

Status
Not open for further replies.
Code: 
n = # of beats
T = tempo in BPM
c = # of cycles from 0x0000 to 0xFFFF

256Hz*65536clks = 16,777,216Hz

65536clks/16,777,216Hz = 0.00390625s

0.00390625s = 1 / 256s

60s / 1 / 256s = 15360s^2

(n-1)*15360s^2 / T = c

modx(15360*(n-1),T) =  15360*(n-1)/T - floor(15360*(n-1)/T)

[B] seed[/B](n,T) = { modx(15360*(n-1),T) }*65536
Things to add to this:

n = t*T/60
where t is the elapsed time

in this post I found the time length between the first button press and the turn over to the intro starting up again was roughly 224 beats at 256 BPM.


224/256 = .875, .875*60 = 52.5 seconds from the button press to start the "press start" screen, and stop the intro, until it starts up again.

52.5s*256 = 13440 times the game overflows seeds from 0x0000 to 0xFFFF, during the time between the button press to start it and when the intro starts up again.

you have 13440 chances to hit your target seed, out of 880803840 possible starting places

1 / 65536 odds isn't good. Factor in inconsistency issues, where you press the first time, and with a metronome, inexactness of the formula hitting beats, and you have an almost impossible chance to hit the same seed twice or more.

so we have the problem of where you start the first button press; well let's take a look at this mathematically.

we know that before the intro starts up, nothing important is going on. We'll call that time length t1. If we reset on the first beat of a measure(in musical terms) then the number of beats that occurs is:

t1*T/60s = n1

this is constant; the length of time between reset and the intro doesn't change. What comes next is variable, because of the user pressing a button at various points to stop the intro.

t2*T/60s = n2

so the time length between where the intro starts, ie after t1, and where the first button press is, is:

t2-1 = (n2-n1)*60s/T

we need to figure out a formula that utilizes the above equation, in order to figure out how the placement of the first button press affects the seed generation.

EDIT1: t1 is approximately equivalent to 16 beats at 256 BPM, which is 3.75s in length

Now we have t1 and the length of time t0, which is 52.5s.

we can break up t0 into two segments, the first skippable, the second fixed in length.

that brings me to another question we must discuss:

does pressing A to skip the intro into the "press start" screen affect the generation of the starting seed?


I'll edit more into this post if I discover anything.
 
Good news. I finally received my Gameboy Player for my cube and I did a bunch of soft-resets, about 30, on Mewtwo last night. Unfortunately there's only 1 turbo setting on my controller, but it doesn't really matter to me. Anyway, I hit TONS of duplicate seeds, and many duplicate IV spreads, without much of a hitch. After 10 or so soft resets, I started getting many duplicates in a row to the point of where when i would go to the party menu, I would see the amount of Mewtwo's HP, say 244/244, and predict every time what the spread was and confirming with only one line of stats on metalkid's iv calculator. I'm not at my laptop right now, which has all my findings, but I'll post them later tonight if necessary...

I'll explain exactly how I do this, to those who are thinking of getting one of these controllers, and to compare my method to mattj. Again, I am using the Gamestop Chameleon Gamecube controller, which I'm pretty sure is different from mattj's. Here it goes:

First, I assign my turbo button to "A" and "B" buttons by holding the turbo shoulder button and pressing A, then holding turbo again and pressing B.

Second, stand in front of Mewtwo, hold Start, B, and Y (or X, meaning select), then press and hold A until the battle with Mewtwo starts. Catch Mewtwo, hold B to skip all the nickname and stuff (this is why I set B to turbo).

Third, check its stats and put it into the first row of stats on metalkid's IV calculator. Keep rare candying it up (holding A so I don't have to watch the animation with Mewtwo and the rare candy), for 2 more levels, which is usually fine. Check seed in RNG reporter, plug it into FRLGSeedFinder, document results.

For step 1, specifically, I hold B+START+Y then hold A and quickly release the others. There were two different general ranges of seeds, almost all lying within either the fe0x range or abbx range. I found some kickass and very early Adamant Mewtwo spreads to use, so I am very excited. The point is though that I want to modify step 1 to test the current clock theory. This too has been bothering me for a while, and I believe it is the same question as Zari and mattj's: When does the clock start? I know it goes against flovv's discovery of the clock starting at 0 right when A is pressed after the intro, but the results just don't to line up. Here is what I propose not only to myself but to mattj and anyone else trying this as well.

1. SR as usual, but don't hold A. Simultaneously start an emloop timer equivalent to the number of beats associated with some established tempos (225, 240, etc.) and enough beats to get past the change of screens between the copyright info and the Gamefreak star.

2. When the clock reaches 0, THEN hold A until you encounter Mewtwo.

3. Document a bunch of spreads (I know how tired of it you are, mattj) and compare them to the ones you have been achieving successfully. I know this puts us at a familiar point of stabbing at the dark and endlessly writing stuff down, but if we can continue to hit consistent, albeit different, spreads by incorporating this delay, then I'd say we're at a comfortable position again. If however we hit the same seeds we usually do (c04'ish in mattj's case, and feox/abbx'ish in my case), then... well I hope that leads us back to what we have recently established, namely that the clock starts when A is pressed on the intro sequence.
 
Great to hear you finally got all the gear! :D

That's weird that you only have 1 turbo setting. Must just be the newer controllers or something (mine's pretty old). However, if you hadn't noticed (I just realized it myself) you can change the consistent starting seeds you get by keeping the same BPM and simply changing the point at which you begin holding
"Turbo A".

When I set my turbo A on the second setting and start holding it the moment I SR I get one set of starting seeds.
When I set my turbo A on the second setting and start holding it at the final note of the Nido/gar scene I get completely different, yet still consistent starting seeds.

No idea WHY that happens, but I'm 100% sure of it, so if you aren't quite happy with the target spreads you find in the consistent starting seeds you happen to be hitting, try to find different points at which to start holding "Turbo A" in order to widen your pool of consistent starting seeds that you know how to hit. :)

Helped me a lot! Oh, and just me personally, I used stuff like the music in the startup to chose a point to start holding A.

Hopefully I'll have time to SR some more tonight.

3. Document a bunch of spreads (I know how tired of it you are, mattj)
Click to expand...
haha just saw this! yes. yes it is very very tiring T_T
Make sure to post your results, and make note of what point you began holding A, etc, etc. I'm really interested in what seeds you happen to be hitting. I still don't know 100% for sure if YOU can hit the same seeds I can on my cart (even though I assume you can if you have the same BPMs and start Holding Turbo A at the same point)...
 
Ok just to help you guys out when doing this research, the length of the copyright info is 16 beats at 256 BPM, and the gamefreak intro (up until the intro with the nido-gar scene) is ~37 beats at 256 BPM for 53 beats total or .20703125

so to determine the amount of beats needed to last through the copyrights + the gamefreak thing is ceil(.20703125*T) = # of beats needed to wait
 
mattj said:
missed my target by 1 frame....
*cries softly...*
Click to expand...
welcome back to emerald with one thousand times more difficulty. As promised, here are all my spreads in order of encounter:
hardy, 2,5,28,18,25,21, f4e1, 561 **
serious, 30,7,25,30,19,18, 6257, 562
lax, 29,20,10,27,12,3, abaf, 562
naive, 3,4,18,5,12,19, f4e2, 561 **
adamant, 13,25,12,10,30,6, abb6, 562 *** !!
sassy, 2,10,114,23,17,0, f4e9, 562 **** !!
adamant, 13,25,12,10,30,6, abb6, 562 *** !!
naughty, 18,10,3,7,13,23, aba6, 562
careful, 2,15,5,22,30,25, abad, 562 ******** !!
docile, 21,18,25,5,25,24, f4e1, 562 ** !!
naughty, 18,10,3,7,13,23, aba6, 562
careful, 8,29,19,0,19,3, abb0, 563
docile, 21,18,25,5,25,24, f4e1, 562 ** !!
sassy, 2,10,14,23,17,0, f4e9, 562 **** !!
docile, 21,18,25,5,25,24, f4e1, 562 ** !!
docile, 30,24,5,18,26,27, abae, 563
timid, 18,5,6,23,12,3, aba5, 563 *****
calm, 6,10,30,7,28,30, abb6, 563
quiet, 23,31,6,3,18,19, f4e0, 562 ****** !!
hardy, 21,23,16,5,6,18, aba5, 562 *****
quiet, 23,31,6,3,18,19, f4e0, 562 ****** !!
sassy, 2,10,14,23,17,0, f4e9, 562 **** !!
sassy, 2,10,14,23,17,0, f4e9, 562 **** !!
quiet, 23,31,6,3,18,19, f4e0, 562 ****** !!
hasty, 10,12,31,12,5,11, abb7, 562 ******* !!
quiet, 23,31,6,3,18,19, f4e0, 562 ****** !!
adamant, 13,25,12,10,30,6, abb6, 562 *** !!
hasty, 10,12,31,12,5,11, abb7, 562 ******* !!
adamant, 13,25,12,10,30,6, abb6, 562 *** !!
careful, 2,15,5,22,30,25, abad, 562 ******** !!
quiet, 23,31,6,3,18,19, f4e0, 562 ****** !!
sassy, 2,10,14,23,17,0, f4e9, 562 **** !!
and breakdown:
duplicate spreads:
abb7, 562 (x2)
f4e1, 562 (x3)
abad, 562 (x2)
f4e0, 562 (x5)
f4e9, 562 (x5)
abb6, 562 (x4)
aba6, 562 (x2)

duplicate/close seeds:
f4e9 (x5)
f4e1 (x4)
f4e0 (x5)
---------
abb7 (x2)
abb6 (x4)
abb0 (x1)
abaf (x1)
abae (x1)
abad (x2)
aba6 (x3)
aba5 (x2)

flop:
6257 (x1) lol
 
Finally got a chance to glance over those results Nix. Those look GREAT! :D Glad they're coming out consistent within reason!!

Have you found any decent target spreads?
 
So I was chugging some more numbers while playing through FR (I'm almost to mewtwo!!) and I noticed something interesting: When breeding in FR/LG only the gender and ability are preset for the egg, unlike in DPPt/HGSS where nature is too. This is a mixed blessing, because you can SR for a better nature, but most likely you won't hit the same nature twice in a row. You not only have to hope for good IVs, but a good nature as well!

As a side note, while I was searching tempos for the spread I found in this post, I found this nugget of gold:

Code: 
Tempo, beats, # of seeds offset from desired seed
115, 119.0000094, [B]0.6160384[/B]

I was SRing for zapdos(at the 158 BPM tempo), and I was getting consistent results, but I'm currently under a DS ban for a day because I've been playing it too much :[

I'll start up again tomorrow, hopefully with some good results. The one thing I noticed from my limited SRs was that I was getting an offset of about 4-5 thousand seeds, consistently. I think that's further confirmation that we need to find out where the clock starts. Once I get back up and running, I'm going to try a few different tests, then post my results here when I'm finished.
 
I'll post all the info soon, if not when I actually catch the bugger, but I think I may have had a very good final sign.

On reset 21, with emloop set up this certain way, I hit my seed and missed my frame by just 1. I realized I might be close, so I left Emloop the same and on reset 33, I hit the exact same spread. Looks like I just need to adjust by 1 and the next time I hit my seed (about 1/10 SRs) I "ought" to hit it...

*plzjesusplzjesusplzjesusplzjesusplzjesusplzjesusplzjesusplzjesusplzjesusplzjesusplzjesus*
 
Misdreavus said:
Is this the spread you're after for Mewtwo?

Hasty, 25/31/30/26/29/31
Click to expand...

That was my previous target spread, but it's unfortunately 26 mins or so into the game. For whatever reason, I was unable to hone my Soft Resets down to within at least 10 frames when Emloop was set to that time.

I switched to a much lower frame target (Naive- 26/25/27/27/29/31 S:55a(5/33 "P.E.") F:35939 (9:58:98!!)) and decided to do a series of SRs just to see if a lower frame target lead to more consistent SRs, and it has so far.

I've got it to within 1 frame (on my seed) it seems now, but I'm having a hard time finding free time to actually do the SRs. It seems to be just a matter of time and luck now.
 
mattj said:
That was my previous target spread, but it's unfortunately 26 mins or so into the game. For whatever reason, I was unable to hone my Soft Resets down to within at least 10 frames when Emloop was set to that time.

I switched to a much lower frame target (Naive- 26/25/27/27/29/31 S:55a(5/33 "P.E.") F:35939 (9:58:98!!)) and decided to do a series of SRs just to see if a lower frame target lead to more consistent SRs, and it has so far.

I've got it to within 1 frame (on my seed) it seems now, but I'm having a hard time finding free time to actually do the SRs. It seems to be just a matter of time and luck now.
Click to expand...
Ah, okay. Do we know for sure that these are the best Hasty/Naive spreads available within a reasonable time frame?

I know that in Emerald, quint-flawless and hex-flawless spreads exist, but occur at absurdly high frames (moths and even YEARS after SR), but since Emerald restricts you to one seed, while FR/LG does not, I would imagine (from what I know, at least) that there are even better Hasty/Naive spreads available at a reasonable frame.

I'd love to get a good mixed Mewtwo in 3rd gen for the VGC! I'd love one with the 31/31/30/31/31/31 Hasty spread that I have for one of my Palkia, but I imagine that's just wishful thinking. :p
 
Misdreavus said:
Ah, okay. Do we know for sure that these are the best Hasty/Naive spreads available within a reasonable time frame?

I know that in Emerald, quint-flawless and hex-flawless spreads exist, but occur at absurdly high frames (moths and even YEARS after SR), but since Emerald restricts you to one seed, while FR/LG does not, I would imagine (from what I know, at least) that there are even better Hasty/Naive spreads available at a reasonable frame.

I'd love to get a good mixed Mewtwo in 3rd gen for the VGC! I'd love one with the 31/31/30/31/31/31 Hasty spread that I have for one of my Palkia, but I imagine that's just wishful thinking. :p
Click to expand...

Those are most likely the best spreads he's found within the seeds he consistently hits with his GC turbo controller; there most likely are better spreads out there, but that one has a manageable SR time for a decent spread.

doing a little more work for my zapdos (finally decided to go with 30 / 23 / 31 / 30 / 31 / 31 bold instead of timid, since I haven't seen any bold ones on smogon :/), I'm still coming up with the hitch over where the clock starts. I just finished getting to the point where I can start SRing for mewtwo, so zapdos may be put on hold for a bit.

also, this is just idle speculation, but would breeding in FR line up with some of the wild methods (obv. with some IVs removed/replaced by passing stats)? The reason I ask is because unlike DPPt, the natures aren't set upon egg generation in FR/LG, only the gender / ability. I find this odd, because in emerald (not sure on RS) the natures are set. Would this have any connection to the erratic method generation for wild pokemon?
 
Oh, no that's not nearly the best spread, just the quickest/good spread (VGC locations NEEEEXT WEEEEK!! D=) I've got something like a...
Hasty, 29/30/39/31/29/31 spread, but it's at a friggin 3 hour 43 minute Soft Reset.

No Thank you. :P

Of course, if I had more time *sigh* I'm really really REALLY sure there are even better low frame spreads out there, I'd just have to take the time to adjust my SRing methods and viola! I'd have new consistent starting seeds.

After the VGCs I'll come back to this and do more reasearch/find new seeds/etc and do another runthrough for a better Mewtwo (and a friggin SD Snorlax too! :P)
 
So, if I understand correctly, the initial seed is determined entirely from the length of time spent on the "PRESS START" title screen, and then the frame is generated in the same manner that it is in Emerald?

Does this basically make it twice as hard to abuse as Emerald? *shudders* I know your turbo controller helps you, but I imagine that to get the best spreads available, we'll need to linger at the title screen for longer than your turbo setting will allow you to do so automatically.

Do you believe that there is an equation that directly relates the initial seed to the time spent at the title screen, such that we will eventually be able to take a seed and use it to determine how long we must linger there in order to hit that seed?
 
Misdreavus said:
So, if I understand correctly, the initial seed is determined entirely from the length of time spent on the "PRESS START" title screen, and then the frame is generated in the same manner that it is in Emerald?

Does this basically make it twice as hard to abuse as Emerald? *shudders* I know your turbo controller helps you, but I imagine that to get the best spreads available, we'll need to linger at the title screen for longer than your turbo setting will allow you to do so automatically.

Do you believe that there is an equation that directly relates the initial seed to the time spent at the title screen, such that we will eventually be able to take a seed and use it to determine how long we must linger there in order to hit that seed?
Click to expand...


n = t*T/60

n = s / frac(15360, T)*65536 + 1

s / frac(15360, T)*65536 + 1 = t*T/60

(s*15 ) / (frac(15360, T)*16384) + 60 = t*T

t = (s*15) / (16384T*frac(15360, T)) + 60/T

Technically that would be right, if only the n formula was right. That's not what the real answer is, but that would be how you would go about solving for it. Currently I'm trying to remove the bug in the n formula, where the first portion is >1. I think if I get that problem fixed, then getting t should be no problem.

EDIT: Misdreavus, I believe the answer to your question is yes, provided we do it with a time variable, such as tempo

EDIT2: now that I think about it, the n formula we have currently does nothing more than tell us if it hits the seed on the second beat.

EDIT3: I JUST FOUND THE EQUATION YOU WERE LOOKING FOR, MISDREAVUS:

Code: 
n = number of beats, t = elapsed time at the menu screen, T = tempo in BPM, s = the desired seed

okay so we know that:

s = { frac(15360n, T) }*65536

and:

n = t*T / 60

so what did I do? I remembered [URL="http://www.smogon.com/forums/showpost.php?p=2557458&postcount=192"]Colin's post[/URL], about the seed formula only giving values of n / T

n = t*T / 60

n / T = t / 60

[COLOR=Red]Warning: what comes next is mind-blowing[COLOR=Black]

s = (15360*t / 60 - floor(15360*t / 60))*65536

[B]s = (256*t - floor(256*t))*65536[/B][/COLOR][/COLOR]

there's your formula, Misdreavus.

EDIT4: Misdreavus, it's not the amount of time you wait at the continue screen, it's when you press the A button; because the game loops through all of the seeds 256 times per second, there are 256 different times that satisfy the equation I just found per second.

Because a great many(256) of these instances happen within 1 second, the degree of accuracy that a computer print would out for time results needs to be within 1 x 10^-9 to be accurate to JUST 1 decimal place, 1x10^-10 for accuracy of two decimal places. The reason it has to be so accurate is because if t<1s we are multiplying the time by 16,777,216! (this is due to the floor function equaling 0 at such low times).
 
Yeah pretty much what he said. It's hard to really comprehend, but apparently it rolls from 0000 through FFFF in like 3 milliseconds. So basically, it's waaaaaaaaaaaay quicker than my Turbo Controller.

I wish it was something like 4th Gen where you could wait X seconds to target Y delay, but unfortunately that doesn't seem practical here (at least so far until somebody verifies that equation). It's a matter of hitting it as it rolls by again at light speed.

Oh, and I wouldn't say "twice as hard as Emerald". Just being honest, it looks to be about "10 times as hard" if not more, just based on the fact that the majority of my consistent starting seeds only appear about 1/10 or so times (some more some less). Oh, and I"m using a Turbo Controller to help me out. I have no idea what the consistency would be for Metronome guys. :/

Oh and just another wrench to throw in there, I don't know why, but I have confirmed that the time at which you begin tapping A makes a difference in what seeds you get too. If I begin holding Turbo A at the moment I SR I get 1 set of starting seeds. If I SR, then let the Nido/gar intro play through a bit, and THEN start holding Turbo A, I get another set of starting seeds.
 
mattj said:
Oh and just another wrench to throw in there, I don't know why, but I have confirmed that the time at which you begin tapping A makes a difference in what seeds you get too. If I begin holding Turbo A at the moment I SR I get 1 set of starting seeds. If I SR, then let the Nido/gar intro play through a bit, and THEN start holding Turbo A, I get another set of starting seeds.
Click to expand...

I have an idea to find out where the clock starts. We need to pick a seed, plug that into the seed to time equation I just found, find a t value that works for our purposes, then put that value into this equation:

n / T = t / 60

if we can find a ratio that matches t / 60, then we have both our beats and our tempo already solved for!

Then, since we would know the beats and tempo required to hit the seed, run a few tests, and find what seeds you would hit.

the next step would be to plug that seed into the seed to time formula, and solve for the smallest time possible, that makes both sides of the equation equivalent.

We also need the smallest time possible for our target seed.

Once both of those values are found, subtract one from the other, add that value to the seed to time formula(note: it goes inside the parentheses, not outside). What that value is is the time offset from the desired seed.

Now that we have the offset factored into the seed to time equation, solve for t once again, then find a ratio of n / T that equals that t / 60

then of course you would check to see if you hit your seed, but if my hunch is correct, you should hit your seed after that final calculation.
 
>_? Caaaaaaaan't beeeeeeeeeee consiiiiiiiiiiiiiistent?!!?()^&(&*

Code: 
1) (Rash): 18 / 11 / 3 / 30 / 5 / 16, 572, 35932
2) (Hasty): 29 / 25 / 22 / 11 / 1 / 21, 935d, 35947
3) (Lax): 22 / 4 / 26 / 8 / 6 / 29, 569, 35935
4) (Gentle): 15 / 30 / 20 / 13 / 30 / 28, 562, 35939
5) (Naïve): 15 / 7 / 31 / 19 / 29 / 0, 556, 35939
6) (Rash): 30 / 9 / 31 / 25 / 5 / 14, 565, 35933
7) (Rash): 30 / 9 / 31 / 25 / 5 / 14, 565, 35933
8) (Lax): 7 / 16 / 3 / 22 / 26 / 18, 556, 35935
9) (Impish): 6 / 3 / 19 / 15 / 27 / 14, dc90, 35934
10) (Careful): 14 / 0 / 30 / 5 / 7 / 27, 55c, 35940  close!!
11) (Serious): 0 / 30 / 0 / 2 / 10 / 18, dc8b, 35937
12) (Rash): 25 / 23 / 16 / 27 / 13 / 24, 7de7, 35930
13) (Calm): 6 / 10 / 14 / 14 / 15 / 31, 571, 35934
14) (Quiet): 3 / 16 / 21 / 12 / 28 / 14, 17f1, 35948
15) (Serious): 0 / 30 / 1 / 14 / 26 / 11, 572, 35943
16) (Impish): 7 / 6 / 25 / 27 / 3 / 13, 935d, 35945
17) (Rash): 30 / 9 / 31 / 25 / 5 / 14, 565, 35933
18) (Mild): 11 / 7 / 23 / 10 / 27 / 3, [B]55a[/B], 35933 whatendefurk?!
19) (Impish): 8 / 26 / 3 / 27 / 13 / 6, 565, 35942
20) (Sassy): 13 / 27 / 3 / 25 / 22 / 29, 935d, 35946
21) (Hasty): 24 / 3 / 0 / 20 / 27 / 8, 565, 35937
I'm using this convoluted mess of apps all together and it's exceedingly difficult to hit consistent frames, these are the best I've gotten so far. I'm still shooting for
55a F35939 Naive 26/25/27/27/29/31
Click to expand...
but I'm scattered +/- 5. It's fairly frustrating.

Does anybody else here have this kind of inconsistency with Emloop? Because I'm not using a portable GBA, I have to use Emloop on my Wii which is probably the problem, as it's fairly difficult to set up my metronome (60 BPM) on my computer in the other room, turn up the volume, synch Emloop's pre-timer with the metronome, Reset/Hold Turbo A EXACTLY on the start of the Real Emloop Timer, get to Mewtwo, press A once to "activate him", then wait 10 minutes and Hit A again to "actually begin the battle".

There's got to be an easier/more consistent way. I'm right "in range" but still scattering all over the place. >_?

[edit]
Oh, and those are just the "latest" SRs, I've done...150+ now just specifically aiming for the two target spreads I've had. @_@
 
The problem with emloop is that its timer is based on a clock rate of exactly 60 fps. As seen by my data waaay back on page 4 or so, the frequencies can be as low as 59 Hz in some cases. This seems to be why higher frames are difficult to obtain even in Emerald, which constanty has me adding or subtracting frames from the pretimer.

Side note: I just panicked at what I said, but then I realized I was being stupid. What I meant by a varying rates is for the RNG itself, not for the clock (seeding). I can breathe now.

I'm sorry that it's been a while since I've actually done any work on this. I've been in between SoulSilver and I'm on a RNG breeding frenzy in Diamond/Platinum... not working towards the VGC's because I am way too late for that train. This is still really important to me as I really want that Sd Mewtwo, and there's some pretty awesome spreads on those seeds I hit. I'm going to bust out my real GBA in the next few days and start putting all this metronome stuff i've been drooling over a shot.
 
so is my rangeof frames within reason for a ten minute soft reset? Or should i be looking for something that i might be doing wrong? This is the first time i've ever used em loop.
 
Yeah, +/-5 frames is pretty damn good... 1/12 = .0833... seconds off is not bad. Since you are generally missing your frame by 1, I find it easier to just NOT adjust by 1 frame, as that 1 off frame is probably human error, and you just might hit it next time without adjusting. Keep at it, I know you'll get it.
 
I hate to double post, but I've been stewing over this idea for the past couple weeks. The idea is using Sweet Scent to check the seed to save time for really long frames. This really only works for Mewtwo. This could work either for tempo tappers or turbo abusers. Here goes.

What you need:

  • 2 open party slots
  • 1 sweet scenter
  • 2 Master Balls (clone in emerald!) (absolutely necessary)
  • Either around 10 Rare Candies (clone!) or Stat-up Vitamins
  • IV checker, either metalkid's (for those with with candies) or marrilland's (for those with Stat-up vitamins)
  • Emloop

Method:


  1. Calibration: Soft Reset as usual, do your usual tempo/turbo stuff, but instead of encountering Mewtwo, use sweet scent, and check the seed of the Pokemon you encounter. Time yourself. Repeat this process several times and use your longest time as tMaxSS (time of maximum sweet scent process). Note that the frame number does not matter, since you are timing a very liberal number of the time it takes you comfortably do this process).
  2. Shoot for a frame,
    Code: 
    F >> tMaxSS*60
    giving you a bit of a cushion. Since most good frames tend to be high, this should work fine.
  3. Find your desired seed and frame, and enter said frame into emloop's timer. Convert to time, yada yada yada.
  4. Soft Reset as usual and start emloop at the time of seed generation, aka pressing start/A on the Venu/zard screen.
  5. Use sweet scent, capture whatever Pokémon you encounter, and check its seed. If the seed is NOT your desired seed, you are free to go back to step 4. If it indeed does match your desired seed, wait until emloop reaches 0 and beeps its final beep, and encounter, hopefully, your desired Mewtwo.
The justification for this method is that it can save you a lot of time and headaches. Let's say that your tMaxSS is around 3 minutes, and your desired frame is is around 30 minutes. If you miss your seed, you can start over, saving you 27 minutes that otherwise would have led to you crying for hitting a very close or exact frame but missing your seed. This is pretty much the closest thing we have to the coin flip trick in DPPt, so we should use it to our advantage. Tell me what y'all think.
 
I'd quote it but that's pretty huge!

That is a great idea and I REEEEEEEALLY hope we can figure out a way to use SSing to find our seed. I had that same idea before, (especially since I used SSing on my Jap HG Cart before we knew about verifying your seed with phonecalls and roamer placement), but I noticed a problem with it. Maybe we can figure out a way to work around it though...

I wish I could draw a picture to illustrate the problem, but this'll have to do. It's a list of the actions you'd be taking. THIS IS WHERE THE FRAMES ARE ADVANCING.
Reset>>GF Scene>>Nido/gar Scene>>Begin Venu/zard Scene>>Seed Selection>>End Venu/zard Scene>>
CONTINUE SCREEN>>REVIEW>>STANDING IN FRONT OF MEWTWO>>
Open Menu>>Sweet Scent>>Battle/Catch Poke>>BACK TO OVERWORLD>>
Open Menu>>Check IVs>>CLOSE MENU>>WAIT IN FRONT OF MEWTWO>>
Battle Mewtwo>>Catch Mewtwo
Click to expand...
The reason I gave up on this idea is that it seemed to me that it would be insanely hard to accurately account for...
1) the time from the Venuzard Scene to Opening your menu to SS
2) the time from Catching your SS poke to Opening your menu to check IVs
3) the time from Closing your menu to battling Mewtwo

I figured, even with missing my seed most of the time and wasting many many MANY Soft Resets, I thought it would still be easier to hit my frame, because there are really (I assume) only 2 points at which I need to focus on...
1) When I start the timer
2) When I Battle Mewtwo

Any ideas on how to get around this problem? I'd REEEEEEEEALLY like to use SS to check my seed. I know of that awesome 29+, 31Spd Hasty Mewtwo, but it's at almost 4 hours x_x
 
Damn, I knew there would be something about frames not advancing at certain points. You know what... how about this? I think this makes the process even easier. There would only be one unknown and you would only have to "worry" about it if you hit your seed! Let me outline this as best as I can.

1. Soft Reset
2. Sweet Scent
3. Battle/Catch
4. Set a timer for a comfortable amount of frames, say FM = 600 (Menu Frame), start the timer and press B to exit the nickname question simultaneously and open your menu when the timer ends. This time, you don't have to keep track of anything!
5. Check IVs, seed, then frame. NOW the frame matters! we'll call this frame F1.
6. Add F1 + FM = dF (informal math speak for change in F).
7. Subtract dF from your target frame, F. This we will call F' (F Prime).
8. Set up an emloop countdown of F' frames
9. Simultaneously exit the MAIN MENU, pressing B and starting the emloop timer at the same time. I'll lay this all out.
Code: 
FM = 600 (10 seconds)
F1 = your encounter frame, around 600 for me so I'll use that.
dF = F1 + FM = 1200
We define F' here, as the final value.
F = F' + F1 + FM = F' + dF ---> F' = F - dF
This accounts for all the time that the frames spend while running. 
F1 is:
[quote][U][B] CONTINUE SCREEN[/B][/U]>>[B][U]REVIEW[/U][/B]>>[U][B]STANDING IN FRONT OF MEWTWO[/B][/U]>>[/quote] FM is: 
[quote][U][B]BACK TO OVERWORLD[/B][/U]>>[/quote]and F' is:
[quote][U][B]CLOSE MENU[/B][/U]>>[U][B]WAIT IN FRONT OF MEWTWO[/B][/U]>>[/quote]
In case you didn't notice, this is akin to finding your starting frame (what I call "Egg Frame") in DPPt Breeding, and also to the starting (Monster) frames for different legendaries in DPPt. Not to toot my own horn, but this looks good to me. We'll have to put this into practice and see what emerges.
 
Status
Not open for further replies.
Back
Top