[Battle] [PokeRed] Fix the 1/255 accuracy glitch in PokeRed disassembly

[NitroHedgehog]

Long time no see ;)

See, I an not so new to RBY compiling, but there is that one glitch that ticks me off. In Red, Blue and Yellow, there is a 1/255 chance that ANY move with 100% accuracy would fail, and I would like to know how to fix this in the PokeRed disassembly script. This is mainly just for a personal project, though. Could anyone help?

 

[BluRose]

Long time no see ;)

See, I an not so new to RBY compiling, but there is that one glitch that ticks me off. In Red, Blue and Yellow, there is a 1/255 chance that ANY move with 100% accuracy would fail, and I would like to know how to fix this in the PokeRed disassembly script. This is mainly just for a personal project, though. Could anyone help?

for this one you need to dig in the code in
engine/battle/core.asm, specifically .doAccuracyCheck

.doAccuracyCheck
; if the random number generated is greater than or equal to the scaled accuracy, the move misses
; note that this means that even the highest accuracy is still just a 255/256 chance, not 100%
	call BattleRandom
	cp b
	jr nc, .moveMissed
	ret

in order to fix it, you can do something like

.doAccuracyCheck
; if the random number generated is greater than or equal to the scaled accuracy, the move misses
; note that this means that even the highest accuracy is still just a 255/256 chance, not 100%
	call BattleRandom
	cp b
	jr z, .move_hit
	jr nc, .moveMissed
.move_hit
	ret

where the problem was that if accuracy was 0xFF and the number generated was also 0xFF then the check would say that it missed because the carry flag wasn't reset (b would not be larger than the register a that the cp instruction compares to)
so we then say that if the two numbers are equal, then make sure it hits instead
every move thus gets essentially a 1/255 boost but at the same time it ensures that moves with full accuracy hit