im_a_blisy’s avatarim_a_blisy’s Twitter Archive—№ 13,729

                              1. A few people are asking for a more detailed explanation, so here's my best go at it. A 🧵 Gen 3 Pokemon games have 6 "common" methods for generating a Pokemon. Method 1, Method 2, and Method 4. This is because a Pokemon is generated in 4 steps. @im_a_blisy/1787663961879384434
                            1. …in reply to @im_a_blisy
                              [FIRST HALF OF PID] [SECOND HALF OF PID] [FIRST HALF OF IVS] [SECOND HALF OF IVS] Shortened to[PID1][PID2][IV1][IV2]. This is Method 1. However, because these are generated in steps, the game can interupt itself with the screen drawing function while building a Pokemon, skipping
                          1. …in reply to @im_a_blisy
                            An RNG advance. So Method 2 looks like this: [PID1][PID2][SKIP][IV1][IV2] And Method 4 is [PID1][PID2][IV1][SKIP][IV2]. Method 3 is [PID1][SKIP][PID2][IV1][IV2], but it hadn't been observed in-game before. But the idea of it had of course been known.
                        1. …in reply to @im_a_blisy
                          All of this was just for static encounters. Stuff like Kecleon, or Kyogre. Random Encounters actually work a bit differently. They use the Method 1/2/3/4 system as a base, but as a community we call them Method W1/W2/W3/W4 for Wild (formerly known as H Methods).
                      1. …in reply to @im_a_blisy
                        The Wild Methods are different because they have extra information to be checked when a battle starts, and that eliminates some PID and IV combos from existing. Stuff like encounter slots do this.
                    1. …in reply to @im_a_blisy
                      Recently, Mucks and Shao had discovered the existence of Method W5, which seems weird because, well, there's nowhere else for a [SKIP] to go in the generation process, right? Well, Method W5 is actually just a Wild Pokemon using Method 1 generation while ignoring encounter slots!
                  1. …in reply to @im_a_blisy
                    What's happening here is that for wild Pokemon, the game picks a nature, then if the PID on that RNG Advance is not the desired nature, the game rerolls the RNG. It'll do this until it hits the desired nature with the desired species. Eventually, around 100~ it quits.
                1. …in reply to @im_a_blisy
                  It just stops doing rerolls and generates what is effectively a static encounter, so you can get Pokemon with Nature and IV spreads that would normally be impossible for it to get in the wild. But what does this have to do with Method 3!
              1. …in reply to @im_a_blisy
                Shao realized if [SKIP] happened between [PID1] and [PID2], after 100~ or so rerolls happened, it would cause a nature mismatch and force method W5. This led him to speculate about Method W3 because that's WHERE the skip needs to be for it to happen.
            1. …in reply to @im_a_blisy
              So he made a custom program that would find a scenario where, on rerolls 100-103, if a [SKIP] happened it'd actually be a nature MATCH, and the Pokemon would continue to generate instead of having method W5.
          1. …in reply to @im_a_blisy
            Shao estimates the odds of W3 happening instead of W5 are around 2.5%. I had to RNG 35 Lairon before I saw W3, and they were all W5. There's a lot of variables we still don't understand fully.
        1. …in reply to @im_a_blisy
          The [SKIP] that's happening is because of the games Vertical Blanking function, which might happen at different rates in different areas of the game. I tried for a while in BF1 of victory road and got nowhere, but it happened relatively fast in F1. So, who knows what's possible.
      1. …in reply to @im_a_blisy
        Also, people are asking how meaningful this is or w/e. Method 3 has been a kind of rumor / theory for years, but practicality wise, it's close to nil. Best case is this will help you find what RNG Advance you landed on if you got no results, same with W5. Or use it to flex.
    1. …in reply to @im_a_blisy
      I meant it when I said it was for RNG Nerds, it's basically just cool to see it happen in game, not an emulator, for real. It's literally not implemented in PokeFinder or RNG Reporter lol. You can't search for it. I was surprised PKHeX accepts it as legal.
  1. …in reply to @im_a_blisy
    Looks like someone submitted a case in 2019 of having it one. github.com/kwsch/PKHeX/pull/2328 It seems to be ignoring much of what I understand can cause Method 3, but, the [SKIP] can happen at any time without cause. I suppose what we just did was aiming for best case conditions.
    1. …in reply to @im_a_blisy
      What is the most confusing to me is this persons generation probably shouldn't be making a spinda. I'll investigate further!
      1. …in reply to @im_a_blisy
        it looks like if they used a syncronizer on seed 762a with 11599 advances it's doing 93 rerolls, which is much more plausible. Gonna try that!
        1. …in reply to @im_a_blisy
          Update, it's hittable. Seed 762a advance 11391 in PokeFinder, you'll hit several Method 5 spindas but eventually should get this one.