The accuracy presented in the RBY Teambuilder is inherently flawed.
As many know, the 1/256 glitch (or "Gen 1 miss") causes moves with 100% accuracy to have a 1/256 chance to not work. As such, moves with 100% accuracy become effectively 99.6%. However, this isn't just for moves with this value; the glitch occurs because the game does a "less than 255" comparison, not because the attack happens to have this 255 accuracy value. Due to this fact, every single accuracy value on Pokemon Showdown's RBY sim is wholly inaccurate. You're essentially writing down the Stadium implementation, in which all of the numbers are correct. I believe this disparity to be a pressing issue for risk assessment when choosing what to do next, and a problem with cartridge accuracy.
So my proposition is to simply fix this in some form. You could change all the accuracy numbers, or put the "real" accuracy in brackets. Some folks over at Pokemon Battle Historia did a nice conversion list for the x/256 uncertainty. From the left, the numbers are; given accuracy, "real" value/256, accuracy considering 1/256 uncertainty. As such, a lot of the work is already done.
Implementing this is a bit of a mixed bag, though; you could also argue that listing accuracy in this way could make it harder to read. In my opinion, that's better than giving what can be seen as a misleading value. The 1/256 uncertainty is much worse for moves with lower accuracy, and is a much bigger problem than the 0.04% chance a "100%" accurate move can miss. I also believe this makes the message you get when you attack with a 100% accurate move and miss a lot more infuriating for a newer player. It's telling you "hey this move is all consistent n all that y'know?" and then slapping you for thinking it's correct. Not to mention PS already calculates accuracy checks with this whole thing in mind for every move. I have to think about potential .3-.5 drops in accuracy for my moves despite these conflicting accuracy values that aren't even being used right in front of me. Sure, it's a small difference, but people make a massive deal out of the x/256 uncertainty already, and the fact of the matter is that this is inaccurate to the cartridge.
Option 1: Outright change all the accuracy values
The most obvious choice would be to do this. I do not believe there is a precedent saying to blindly use the game's specified move values instead of accurate ones, so I don't think this is a stretch. In fact, Pain Split in Gen 3 has 100% listed accuracy in-game, but it bypasses checks. On Showdown, as you can see that it's listed to have --% accuracy, which is accurate to how the game actually processes it. In my opinion, not blindly following the game is a good thing, as there's a lot of cases where it will lie about accuracy (eg. Nature Power has 95% Accuracy despite calling other moves, which in turn follow their own calcs).
One issue I have with this option is that I believe this does have a knock-on effect on Stadium's teambuilder, which could be a problem. I remember Marty saying something about them using the same movedex or something, which has resulted in issues with Focus Energy's description. It just makes an already mixed bag of a teambuilder (many effects aren't noted due to pulling from RBY) worse.
Option 2: Brackets
If it's too big of a problem to fix the numbers outright, the actual accuracy could just be put in brackets like this mock-up, right? This way, you can get the "friendly" numbers without decimals, while allowing players to do an accurate risk assessment using the "less friendly" numbers. I think this would be the least invasive solution, but it's still there, and could be a bit confusing to those out of the loop. Maybe something like /256: x or...idk.
EDIT: As an aside, this Articuno analysis cites DEdge as 99.6% acc, which only adds to the confusion.
As many know, the 1/256 glitch (or "Gen 1 miss") causes moves with 100% accuracy to have a 1/256 chance to not work. As such, moves with 100% accuracy become effectively 99.6%. However, this isn't just for moves with this value; the glitch occurs because the game does a "less than 255" comparison, not because the attack happens to have this 255 accuracy value. Due to this fact, every single accuracy value on Pokemon Showdown's RBY sim is wholly inaccurate. You're essentially writing down the Stadium implementation, in which all of the numbers are correct. I believe this disparity to be a pressing issue for risk assessment when choosing what to do next, and a problem with cartridge accuracy.
So my proposition is to simply fix this in some form. You could change all the accuracy numbers, or put the "real" accuracy in brackets. Some folks over at Pokemon Battle Historia did a nice conversion list for the x/256 uncertainty. From the left, the numbers are; given accuracy, "real" value/256, accuracy considering 1/256 uncertainty. As such, a lot of the work is already done.
Option 1: Outright change all the accuracy values
The most obvious choice would be to do this. I do not believe there is a precedent saying to blindly use the game's specified move values instead of accurate ones, so I don't think this is a stretch. In fact, Pain Split in Gen 3 has 100% listed accuracy in-game, but it bypasses checks. On Showdown, as you can see that it's listed to have --% accuracy, which is accurate to how the game actually processes it. In my opinion, not blindly following the game is a good thing, as there's a lot of cases where it will lie about accuracy (eg. Nature Power has 95% Accuracy despite calling other moves, which in turn follow their own calcs).
Option 2: Brackets
If it's too big of a problem to fix the numbers outright, the actual accuracy could just be put in brackets like this mock-up, right? This way, you can get the "friendly" numbers without decimals, while allowing players to do an accurate risk assessment using the "less friendly" numbers. I think this would be the least invasive solution, but it's still there, and could be a bit confusing to those out of the loop. Maybe something like /256: x or...idk.
EDIT: As an aside, this Articuno analysis cites DEdge as 99.6% acc, which only adds to the confusion.
Last edited:
