I wasn't aware of this but from my testing it appears resetting through the touch screen is a hard reset and resetting with A B START SELECT is a soft reset. Is this common knowledge? It was killing me that I kept getting duplicate TIDs, and I knew the seed only gets reset on cart and emulator versions with a full hard reset. So I switched over to the button reset and voila, no dupes after 100 input-buffered attempts.
So what could this mean for people SRing for shinies with the touch screen? It depends. If the Pokemon is a stationary where the DV spread is generated with minimal margin for entropy, you may unknowingly be hitting a few duplicate spreads, especially if you're buffering A through the intro screens. But even mashing can cause duplicate code branching. This applies to Gen 2. If you're on Gen 1 there is too much margin for entropy to affect the average player who isn't using multiple input buffering techniques.
The upside is this could mean Gen 2 time-based single entry frame RNG is a little more convenient to pull off, assuming you match your hardware to a generated list of TIDs or DV spreads.
Update: Just got a dupe after 121 resets. Could be coincidence. At the very least the button reset somehow causes tons of RNG branching. Would welcome any additional testing!
Update 2: Another dupe at 201 resets (different value).
Update 3: 300 resets, 2 dupes. Big improvement over 20 resets 17 dupes from touch screen resetting. Will test this with DV spreads next time I have the chance. For now I would encourage anyone still SRing on the VC to use the button method, no matter how ridiculously uncomfortable it is....
Update 4: 500 more buffered resets, 2 dupes. So a total of 4 different duplicates in 800 resets, which is consistent with normal SR odds. This concludes my testing as far as TIDs go.
So what could this mean for people SRing for shinies with the touch screen? It depends. If the Pokemon is a stationary where the DV spread is generated with minimal margin for entropy, you may unknowingly be hitting a few duplicate spreads, especially if you're buffering A through the intro screens. But even mashing can cause duplicate code branching. This applies to Gen 2. If you're on Gen 1 there is too much margin for entropy to affect the average player who isn't using multiple input buffering techniques.
The upside is this could mean Gen 2 time-based single entry frame RNG is a little more convenient to pull off, assuming you match your hardware to a generated list of TIDs or DV spreads.
Update: Just got a dupe after 121 resets. Could be coincidence. At the very least the button reset somehow causes tons of RNG branching. Would welcome any additional testing!
Update 2: Another dupe at 201 resets (different value).
Update 3: 300 resets, 2 dupes. Big improvement over 20 resets 17 dupes from touch screen resetting. Will test this with DV spreads next time I have the chance. For now I would encourage anyone still SRing on the VC to use the button method, no matter how ridiculously uncomfortable it is....
Update 4: 500 more buffered resets, 2 dupes. So a total of 4 different duplicates in 800 resets, which is consistent with normal SR odds. This concludes my testing as far as TIDs go.
Last edited: