If shiny odds are 1/8192 and DPPt runs at 30 FPS, does that mean there is a shiny frame roughly every 273 seconds?

Question

DPPt or any game with full odds. I remember hearing about RNG manipulation and finding the specific frames that shinies appear on.

I'm thinking about a soft reset for example at 1/8192 odds. If there is a 1/8192 chance every frame that the Pokemon you encounter is shiny, and there are 30 FPS, that means there is one shiny frame approximately every 273 seconds, or FOUR MINUTES AND THIRTY THREE SECONDS! That seems like an insane amount of time for there to be a single shiny frame in, but then again, these odds are a lot longer than I think.

Is my thinking on this right, or do I have the logistics of this completely wrong? Thanks!

Comments

You might get a better answer if you ask here.
https://www.smogon.com/forums/threads/most-generation-prng-help-information.52180/

---

I believe you are right, if not close. I know RNG manipulators only need to wait a few seconds to minutes to hit the right frame most of the time.

---

I figured you or fizz probably just know whether this is accurate or not

Answer 1

Your understanding is good for Gen 3 but not Gen 4.

In Gen 3, the frame (as in: the RNG frame, defined here) advances each time the game draws a new frame (as in: the picture that refreshes at 60 Hz) to the screen.

You can use an RNG tool to see that this does indeed lead to shiny frames being generated every few minutes on average (and often in big clumps due to the simplicity of the PRNG used), but it's important to note that your odds are still bound by the 1/8192 probability, not the rate at which the frame is turned over.

In Gen 4, frame advances aren't tied to the refresh rate and instead occur at the particular times the game needs to create a random event. That event could be something like generating the stats of your wild encounter, or something as simple as which direction an NPC randomly turns next. More examples relevant to RNG manip are listed here.

Comments

Huh, my answer might be wrong then, since most of my experience with RNG manipulation is with Emerald. I just assumed it was the same since RNG Reporter has methods for Gen 4 that, from my cursory glances, seemed to be very similar to the Gen 3 ones, but I guess it's a lot more complex than that.

---

Yeah Gen 3 and 4 both use LCGs seeded using system time (minus the Emerald seed oversight), so they have lots in common. It's just frame advancement and how the numbers are realised as stats etc that changes it up.

---

Thanks Fizz, I appreciate it. When you say that in Gen 3 the odds are still bound by 1/8192 and not the rate that the frame increases, are you saying that every frame is obviously the same as any other frame or am I interpreting that incorrectly?

The Gen 4 part makes sense, at least to the degree that I want to understand

---

What you said is correct insofar as shiny probability. That line in the answer is just trying to emphasise that although "shiny available every few minutes" sounds exciting, it's the overarching probability (and how quick you can trigger encounters) that actually matters for getting a shiny faster.

Answer 2

You're on the right track, but remember that each shiny frame is completely independent of the frames around it. So while it's true that there is, on average, one shiny frame every 273 seconds, outliers can be very common. From my own experience using the RNG Reporter, there are some cases where two shiny frames will be less than a second apart, as well as 30+ minute spans with no shiny frames at all.

So it might be inaccurate, or at least misleading, to say "there's a shiny frame roughly every 273 seconds," because the game doesn't look at groups of 8192 frames and choose one of them to be shiny; each frame is randomly decided with an independent 8192 chance.

Edit: Most of my experience with RNG manipulation comes from Gen 3, and as Fizz pointed out, may not apply to Gen 4 as much as I originally thought.

Comments

Right, I understand and agree with this. Good to know that that is how Gen 3 works and not Gen 4. Thanks Krlw