On Glitches on Cartridge and whether it should Apply to Smogon

DragonWhale

It's not a misplay, it's RNG manipulation
is a Smogon Social Media Contributoris a Forum Moderatoris a Community Contributoris a Battle Server Moderator
Moderator
#1
Currently in the games Pokemon Sun and Moon, when Z-Memento or Z-Parting Shot are used in online play the game will crash, disconnecting both players and ending the game in a tie.

Clearly this is a major glitch, and Pokemon has decided to ban the moves Memento and Parting Shot from all online battles with random matchings and VGC. It'll still crash if you play against a friend online. We still don't know if or when a patch to fix it will be applied.

The point of this thread is to ask: do we stay faithful to the cartridge mechanics on Pokemon Showdown to the point where these glitches are considered, and if so do we ban Memento and Parting Shot from all gen 7 battles? Or do we go for the complex ban with the combination of Darkinium Z + either of these moves? Or do we use the "HP percentage mod/sleep clause" argument and pretend like this glitch doesn't exist at all?

Just for reference, there's precedent of banning glitch-inducing moves with BW Doubles banning Sky Drop.
 

Karxrida

Travels at the speed of DARKNESS
is a Community Contributor
#2
There's also precedent for flat-out ignoring bugs, like how Showdown's Gen IV sim does not implement the easily triggered Acid Rain.

Does the glitch only occur online or does it also affect local wireless matches? If it does not occur with local play, then we could easily not do anything because there is an existing venue where we can play matches without the glitch occurring.
 
#4
I'm pretty sure there was an active decision made in the past to not implement game breaking bugs like acid rain. I can't find the definite thread with search function being wonky for some time periods :S

There are several threads/posts in PR mentioning it though. 1 2 3

Basically it was previously agreed that the simulator need not fully replicate bugged and unintentional mechanics. However, this is highly subjective, and doesn't always apply. Things like stat overflow in RBY as a technical limitation are implemented, while acid rain as a bug isn't implemented. Both of these don't break the game though. So it depends on the situation as to whether to ban / not implement. BW Doubles could have very well decided to have the simulator not implement the Sky Drop glitch and still allow it I think. You can both treat game breaking glitches as if they don't exist or you ban them.

While we don't always have to faithfully replicate game mechanics/glitches, it still requires an active decision to be made to either ban or not to implement, usually on the basis of net gain from the decision.
 
#5
I think it's worth mentioning that there's a fundamental difference between a glitch that doesn't impede the game like Acid Rain and the Skydrop glitch and a bug that straight-up crashes the game, especially when the crash is triggered by using one of two new moves. Just because some programmer(s) at Game Freak screwed up and missed a game-breaking glitch doesn't mean we should have to suffer for it, especially when they only banned the moves to stop games from crashing, not because the moves were being abused to win games
 

Quite Quiet

sleep life away
is a Tournament Directoris a member of the Site Staffis a Super Moderatoris a Community Contributoris a Tiering Contributoris a Smogon Media Contributoris a Battle Server Moderator
Articles Co-Leader
#6
I've spoken about this previously, but I don't understand why we should be selective when we decide what is implemented in a sim. Simulators, by their definition, attempt to replicate the original as closely as possible (in this case Pokémon games). Bugs and glitches have always been a part of those. Yes in this specific scenario the glitch is a bit more severe than before, but I don't think that should be a factor. If there's ever a call to not implement a glitch like this, then we've effectively moved to a metagame that we can't play on cart anymore. Sleep Clause already does that partly, but there's at least ways to adhere to that on cart (the one who slept more than 1 mon forfeit). This has no possible workaround on cart though, so not implementing it means we're throwing away playing gen 7 metagames on cart. Despite cart being less relevant that it was previously, I don't think that's something that should be done.
 

Steven Snype

Kunclord Supreme
is a Tutor Alumnusis a Forum Moderator Alumnusis a Community Contributor Alumnusis a Contributor Alumnus
#8
If it weren't for the bugs that happen in that generation's cartridge games, we would not have RBY the way it currently is and still have to force our Marowaks to suboptimal DVs in GSC to be able to SD. Sleep Clause iirc was how the games would handle Sleep Clause, not what is an actual metagame bug, which requires a different discussion from the rest of this, as this is focusing on in-game bugs. In a perfect world, we'd be playing on a simulator that replicates every detail, and Showdown does try to replicate the game's bugs, as evidenced in the past generations and how they function.

Obviously, what we play on is the simulator, and it's on the sim development team to replicate or address those bugs. We have the obligation to follow the technical limits of the means we agree to play on (most of the time, it's the simulator). If there's a discrepancy between the simulator and the cartridge, the player community can't actually do anything about it because the sim developers are the ones that can actually implement it. In practice, there's obv gonna be time between when we discover a bug in-game and with when it can/will actually be replicated on the simulator since there's all sorts of other dev priorities (broadly speaking) than implementing what is comparatively speaking, a feature when other shit is working.

This imo should be a sim developer call on when it's implemented. Until then, we play to the limits of what we got available with us. This was a TD decision posted publicly a few years back when the discrepancy between sim play and wifi play actually happened (don't have the link to it off the top of my head) I think it was back when Philip7086 wasn't a terribly old fart who nobody remembers.
 
#9
I've spoken about this previously, but I don't understand why we should be selective when we decide what is implemented in a sim. Simulators, by their definition, attempt to replicate the original as closely as possible (in this case Pokémon games). Bugs and glitches have always been a part of those. Yes in this specific scenario the glitch is a bit more severe than before, but I don't think that should be a factor. If there's ever a call to not implement a glitch like this, then we've effectively moved to a metagame that we can't play on cart anymore. Sleep Clause already does that partly, but there's at least ways to adhere to that on cart (the one who slept more than 1 mon forfeit). This has no possible workaround on cart though, so not implementing it means we're throwing away playing gen 7 metagames on cart. Despite cart being less relevant that it was previously, I don't think that's something that should be done.
I'll just like to clarify here. Simulators aim to replicate the game mechanics as much as possible. However, we must accept that the game we play on simulators aren't the exact same as the one we can play on cartridge. If we want to play on a pure cartridge restrictions, we wouldn't be playing with Wish event Pokemon etc. (that argument is a stretch though). So what we play on sims indeed is what we can't already play on cart. For that Sleep Clause example, I'm pretty sure that no it cannot be adhered on cart? On the sim you can spam your sleep move without fear of having to forfeit, but on cart you can't do the same. By adhere, yea they can adhere to Sleep Clause on both sim and cart but they are two different implementations of Sleep Clause. So we're still essentially playing two different games. We've never played the actual game since like Gen3?

If it weren't for the bugs that happen in that generation's cartridge games, we would not have RBY the way it currently is and still have to force our Marowaks to suboptimal DVs in GSC to be able to SD. Sleep Clause iirc was how the games would handle Sleep Clause, not what is an actual metagame bug, which requires a different discussion from the rest of this, as this is focusing on in-game bugs. In a perfect world, we'd be playing on a simulator that replicates every detail, and Showdown does try to replicate the game's bugs, as evidenced in the past generations and how they function.

Obviously, what we play on is the simulator, and it's on the sim development team to replicate or address those bugs. We have the obligation to follow the technical limits of the means we agree to play on (most of the time, it's the simulator). If there's a discrepancy between the simulator and the cartridge, the player community can't actually do anything about it because the sim developers are the ones that can actually implement it. In practice, there's obv gonna be time between when we discover a bug in-game and with when it can/will actually be replicated on the simulator since there's all sorts of other dev priorities (broadly speaking) than implementing what is comparatively speaking, a feature when other shit is working.

This imo should be a sim developer call on when it's implemented. Until then, we play to the limits of what we got available with us. This was a TD decision posted publicly a few years back when the discrepancy between sim play and wifi play actually happened (don't have the link to it off the top of my head) I think it was back when Philip7086 wasn't a terribly old fart who nobody remembers.
I'd think that things like stat overflow aren't so much of a bug but a technical limitation the old games had. Maybe Crystal_ would want to chime in on that. So in a sense, they're still basic core mechanics of those past generations. As for actual bugs that are possibly game breaking or exert a bad influence on the game, decisions are made to avoid the bug by either banning or not implementing. There have been people who argued for Acid Rain to be implemented. However, ultimately it was decided not to implement acid rain due to it having no benefit to the game other than staying true to the cart mechanics. The game we play on sim does not have any obligation to be faithful to all cart mechanics and we can deem things as irrelevant and unwanted if that's what improves the sim game. So yea there's a precedence where we actively chose not to implement sleep clause.

I don't think we need to discuss further as to whether or not to fully follow game mechanics since this was a precedent set back since DPP? Sorry for going off tangent.

Anyway it might be worthwhile to discuss whether we want to ban such glitches or just not implement the glitch for future guidelines. Alternatively it can just be handled on a case by case basis.
 

Nails

Nuke the Weebs
is a Forum Moderator Alumnusis a Tiering Contributor Alumnusis a Battle Server Moderator Alumnusis a Past WCoP Champion
#10
handle them on a case by case basis. not all cart bugs are created equal and implementing obvious bugs which are gamebreaking is silly. normally i lean pretty hard in favor of cart mechanics but people on battle spot use zarting shot as a means of forcing a draw if they're in a disadvantageous position because they crash the game. there's exceptions to every rule and something which can force end the game is one we should make.

also it's probably going to get patched anyways.
 

UltiMario

Out of Obscurity
is a Pokemon Researcher
#11
I mean I feel like we don't need to bother considering they're apparently planning on fixing this
In addition to this, I haven't heard any reports of crashes during Quick Link (local wireless) battles. So in this case, there's still a perfectly valid way to play with Z-Memento and Z-Parting shot since no limits 6v6 singles is a Quick Link option.
 

DragonWhale

It's not a misplay, it's RNG manipulation
is a Smogon Social Media Contributoris a Forum Moderatoris a Community Contributoris a Battle Server Moderator
Moderator
#12
In addition to this, I haven't heard any reports of crashes during Quick Link (local wireless) battles. So in this case, there's still a perfectly valid way to play with Z-Memento and Z-Parting shot since no limits 6v6 singles is a Quick Link option.
Actually in offline/quick link play, the incoming pokemon don't heal lol. It doesn't crash the game though. Do we just ignore it and have Pokemon Showdown emulate the post-patch cartridge play because we assume it'll be fixed? Or do we just have these z-moves not heal the incoming Pokemon the way it works on cartridge at the moment?

Another glitch that came up today is an announcement from PGL (only in Japanese atm) that says Rocky Helmet tiebreakers are not correctly implemented. Starting from gen 7, Rocky Helmet was supposed to favor the attacking Pokemon in the event of a tie, but the mechanic is still acting the way it did in gen 6 where the defending Pokemon wins, as shown by Marty's post here. Which Rocky Helmet tiebreaker do we implement in this case? The one where the attacker wins or the one where the defender wins?

As for how to deal with future glitch implementation, I agree with people saying it should be a case by case basis. It's not that common and it makes sense to treat glitches differently depending on the severity on the playability of the game.
 
Last edited:

Isa

CHINGUYA
is a Tournament Director Alumnusis a Smogon Social Media Contributor Alumnusis a Global Moderator Alumnusis a Community Contributor Alumnus
#13
Actually in offline/quick link play, the incoming pokemon don't heal lol. It doesn't crash the game though. Do we just ignore it and have Pokemon Showdown emulate the post-patch cartridge play because we assume it'll be fixed? Or do we just have these z-moves not heal the incoming Pokemon the way it works on cartridge at the moment?

Another glitch that came up today is an announcement from PGL (only in Japanese atm) that says Rocky Helmet tiebreakers are not correctly implemented. Starting from gen 7, Rocky Helmet was supposed to favor the attacking Pokemon in the event of a tie, but the mechanic is still acting the way it did in gen 6 where the defending Pokemon wins, as shown by Marty's post here. Which Rocky Helmet tiebreaker do we implement in this case? The one where the attacker wins or the one where the defender wins?
i am against implementing versions of moves that causes crashes but i have no issues with moves that dont work as intended. developer intentions are not important for our purposes - with regards to move functionality at least - and we do not adjust ourselves to that. see focus energy or ghost vs psychic in rby which was incorrectly coded, which we have not fixed - developer intentions were to raise the critical hit ratio and make ghost attacks super effective, but we stay true to the actual game and lower the crit rate/make psychic types immune to ghost moves.
as per the above, in my opinion it'd be the best if zarting shot, zemento and rocky helmet reflect cartridge play rather than intentions. if they're updated through a patch, we'd follow suit.
 
Last edited: