[20:58] ParaSlam: the link battle prng follows a predictable pattern
[20:58] ParaSlam: it might e exploitable
[20:59] ParaSlam: the next pseudo-random number called is n × 9 + 1
[21:35] ParaSlam: the first 10 PRNG calls are based on the divider register, so they are hard to predict
[21:36] Piexplode: aight
[21:36] Piexplode: this is interesting o.o
[21:36] ParaSlam: but as soon as you get a 1/256 miss, you can guess what will be the next pseudo-random numbers if I'm right
[21:36] Piexplode: wow!_!
[21:36] ParaSlam: that would be sooooooooo broken
[21:36] Piexplode: however, a legitimate game element.
[21:37] ParaSlam: just take 0xFF, multiply it by 9, add 1
[21:37] ParaSlam: ta dah, you have 0xF8
[21:37] ParaSlam: then B9
[21:37] ParaSlam: etc
[21:38] Piexplode: that's amazing!_!
[21:42] ParaSlam: so if BSlam misses, Hypnosis will hit, then Hyper Beam won't critical, then 0x82 0x93 0x2C 0x8D are rerolled, then it will get a 96.47% damage roll (0xF6)
[21:43] ParaSlam: this means that with a simple pocket calculator aside your game boy, you can predict what will happen
[21:43] ParaSlam: that is, if your faster or predict what the opponent will do if his/her pokémon is faster
[21:44] ParaSlam: that wouldn't work for the first turns of battle, as the 10 first calls use with the normal prng
[21:44] ParaSlam: use the normal prng*
[21:45] ParaSlam: I'm going to check this just to be sure
[21:47] ParaSlam: also, if the PRNG calls at link battle initialization are actually used in-battle as Novus suggests it, that means the host has more chances to win speed ties at the beginning of the game
[21:47] ParaSlam: and slightly increased Critical Hit rate as well
[21:48] Piexplode: oml
[21:48] Piexplode: this is
[21:48] Piexplode: insanely crazy sutff
[21:48] Piexplode: oml
[21:48] Piexplode: oml
[21:48] Piexplode: wow
[21:48] ParaSlam: and increased chance of full paralysis as well