RBY Counter

[Hipmonlee]

There is a specific case where the RBY community has understood that PS has had a buggy counter implementation for a while. However, while investigating this Enigami has discovered that the correct implementation actually causes a desync.

https://www.smogon.com/forums/threads/past-gens-research-thread.3506992/page-8#post-8342135

The scenario is this:

Player 1 has a paralysed pokemon up against Player 2's pokemon with Counter.
On turn 1 Player 1 uses a counterable move while Player 2 uses some move other than Counter.
On the following turn, Player 2 selects Counter and Player 1 FPs.

On Showdown, then no matter what move Player 1 selected Counter will succeed.

However, on the cart, if Player 1 selected a non Counterable move then Player 1's Cart would consider Counter to have failed. But Player 2's Cart has no information about what move was selected (or at least failed to record what move was selected), and will consider Counter to have succeeded (because the last used move that it has recorded is counterable). This causes a desync.

Here's a replay of me attempting to recreate this on Showdown.
https://replay.pokemonshowdown.com/gen1ou-1047692234 - Jump to turn 20 (or watch the whole thing if you want to see me being an idiot).

This is abused in battles, I have definitely seen Nails KO an enemy Chansey by repeatedly Countering a Seismic Toss while it FPs (you often only need consecutive FPs to take out a Chansey).

Note: this is confirmed to occur for Paralysis and the turn ending. It presumably also happens in the case of Freeze, Sleep, Confusion, Flinch, or anything else that can cause a move to not occur. However, in the case of Sleep and Freeze, if the Pokemon is already Frozen or Sleeping, they do not select a move at all. This would only happen on the turn the Pokemon is put to Sleep or Frozen.

So what should be done with Counter?

I think we have 5 reasonable responses at this point:
1. Ban Counter

• This is the only true to cart option we have. I don't like this idea.

2. Patch Counter so that for both players it behaves in the way Player 1's Cart interprets the move.

• Counter would fail if the move selected was uncounterable.
• Note this will also create a scenario where if Player 1 were to use a non Counterable move the first turn, and a Counterable move the second turn as it FPs, then Counter could succeed if the last damage dealt (by either player) was non 0.
• This preserves the "spirit" of Counter, so that if you correctly predict your opponent to use a normal or fighting move your Counter has the potential to succeed, even if your opponent is Fully Paralyzed. Likewise you can't get an easy KO by just waiting for your opponent to FP the turn after using a Counterable attack.

3. Patch Counter so that for both players it behaves in the way Player 2's Cart interprets the move

• This is closest to the current implementation of Counter.
• This means that players have a little more certainty about when their Counter's will succeed. For instance if Player 1's Pokemon faints before attacking and then Player 1 sends out a sleeping Pokemon, Player 2 would know for certain whether Counter will succeed or not on the following turn despite not having seen the move Player 1 selected.

4. Patch Counter so that it will fail whenever a desync would occur.

• This is the closest to the Cart, while still allowing Counter.
• Doesnt have the arbitrariness of options 2 and 3.
• It could be quite confusing, where sometimes Countering against a paralyzed Pokemon would succeed and sometimes it would fail.

5. Patch Counter so that it always fails whenever Player 1's last selected move is unknown to Player 2's Cart.

• Similar to the last option, but more consistent.

Any other ideas I missed?

 

[teal6]

I think Option 4 is the clearest and best here without question, actually. I do not worry about the last bullet point because anyone willing to get significantly into RBY is going to have to wrap (haha) their head around bizarre and seemingly nonsense mechanics long before they en"counter" (haha) counter working weirdly. I don't think it'd be a huge stretch of imagination to go "oh, he clicked a move that isn't counterable and we have a mod that causes that to fail".

 

[Hipmonlee]

My view here is that I specifically do not like option 3 because introduces very powerful strategic options that do not exist in the cart.

Aside from the Chansey vs Chansey example, some other scenarios are things like:

Player 1 Explodes and kills Player 2's Pokemon.
Player 1 sends out a Paralysed Pokemon, and Player 2 sends out a Pokemon with Counter.
Player 2 can just fish for a 25% chance of a OHKO.

I have a team where one of my lines against Snorlax is to paralyse it, and then softboiled repeatedly with Chansey against it, and whenever they FP throw a Counter off for the 25% chance of finishing it off.

Option 2 is much less abusable because the Counteree can always protect themselves from a Counter. And 4 and 5 obviously prevent this as well.

 

[Zarel]

Note that unless we ban Counter or cause the game to end when this case comes up, this is going to be named "Counter Clause Mod".

Depending on how often this case comes up, maybe "cause the game to end" isn't that bad of a choice?

 

[Hipmonlee]

It's pretty common. Pokemon are paralysed about as often as they are unparalysed, and counter is probably on about a quarter of teams?

[Edit] - for the benefit of people unfamiliar with competitive RBY, I am absolutely confident that there would be no meaningful objection to a "cause the game to end" implementation resulting in the immediate banning of Counter for all competitive RBY.

 

[Isa]

How does this interact with player 1 switching as player 2 uses Counter? Also a desync?

 

[Lutra]

I think it's helpful to understand the desync scenarios that Crystal_ made videos on: https://www.youtube.com/user/CrystaL12RG/search?query=desync to help consider the bigger picture.

My first thought is that when considering the RBY tie situation, where each cart registers a loss for the user by default, I think the result would be corrected to a tie if each cart told the other that they lost and they knew how to interpret that. In this Counter scenario, where Player 1 confirms a non-counterable move, the information just doesn't seem to be updated for Player 2's cart, so I'd favour the Player 1 cart interpretation.

I then saw in Crystal's Counter desync video, it seems to imply the cursor can select a move without confirming it, and then make a switch and that move is stored as their last move. I assume this would allow virtual any of Player 1's mons, perhaps not named Hitmon, to change their move to something non-counterable before switching, which would probably cause a desync; but as far as Player 1's cart is concerned, they dodged a Counter. So it seems like if we take Player 1's side to the extreme, that's another away for Counter to fail. You could perhaps take the Player 1 cart interpretation only if a move is confirmed. I think whatever is decided, it's important for the players to have easy access to the information on how PS differs from cart link battles.

 

[Quite Quiet]

Given Enigami's other note regarding counter it doen't seem to me that there is any way to cleanly account for all ways Counter can desync. Either tie games that would end up in a desync if playable on cart, or ban Counter. Maybe we could come up with a complete spec of "evey way desyncs happen and how we decdied that the mechanic works", but then we're straying even further than how cart works. Arbitrarily deciding how a mechanic /should/ work according to taste is not how we should go about things. If a mechanic or feature behaves in unexpected or buggy ways we should just deal with it.

Obviously, this makes RBY less playable than before, but other issues in other gens doing the same aren't modded out of the game because of that. We should accept that this is how the game is and work with what we have.

 

[Stratos]

I dunno Quite Quiet, we're really pretty inconsistent with this stuff. In Platinum, Pursuit can cause a desync; in BW, Sky Drop can. Yet in DPP we mod the desync out while in BW we banned Sky Drop.

I think the RBY counter situation is closer to the Platinum Acid Rain one, since we've been playing with Counter (Pursuit) for years already. It should hopefully be possible to determine how Counter "should" work by finding an NPC who can use the move and just testing its behavior in PvE battles? I haven't watched Crystal_'s video yet so forgive me if that's already been done.

If that ends up nerfing it (because it will never work on switching) then, well, fine... but I don't think we need to delete the move.

 

[Amaranth]

I quite like Ortheore's take here but Stratos also makes a very fair point with the DPP Pursuit precedent. I personally feel like if we're accepting to mod the game to make things closer to what we would like them to be, then we might as well ban 256 misses and recovery fails, so my initial instinct is to avoid modding and just ban Counter - which leads me to my second point:

I have a serious issue with the OP ruling out banning counter because "he does not like it", which just translates to "I've been playing this tier with Counter for a long while and I do not want that to change". This tier already has several issues due to refusing to change. Research showing that some pokemon cannot get max DVs in all stats without the use of glitches has been completely ignored, and even though this survey confirmed that a majority of players would like to see something change in terms of tiering, absolute fuck all has been done to act on this in months. I understand that I don't have the full backstory, not being a council member and all, but I think it's not unreasonable to draw the conclusion that there is some sort of reluctance by the council to significantly change RBY, and frankly it bothers me, much like the reasoning of "I don't like it because it's change" also bothers me. Latias has been unbanned in DPP and sleep has been banned in BW but we're stuck here with every initiative getting choked out and eventually dropped due to a painfully slow council and a playerbase that doesn't seem to be bothered by this at all, despite voting in favor of change in the survey linked above. Don't rule out banning Counter because "you don't like it", come up with some good arguments against it - Stratos is helping you out here already.

These three issues are currently unanswered (Counter, impossible max DVs, survey) and this is a good moment to bring some attention back to them. DO SOMETHING. I don't care if we end up taking decisions that don't line up with my personal ideas, I just want to see the community sincerely and fully address these problems rather than brushing everything under the carpet because they don't care for change anyway. Start with Counter here, and move on to the others quickly, please.

 

[Lutra]

Given Enigami's other note regarding counter it doen't seem to me that there is any way to cleanly account for all ways Counter can desync.

It desyncs for Counter when the move isn’t communicated to the opponent. So it wouldn’t show up in the battle log, but a player chose it. The frustrating part is a player can change the choice of move before switching out, a bit like checking the mega evo box on, and it actually mega evolving while you switch. I think that’s all we know so far about Counter desyncs.

 

[Hipmonlee]

I have a serious issue with the OP ruling out banning counter because "he does not like it"

I havent ruled it out, I dont have the authority to rule anything out any more. I just dont like it.

For the record it was a long time before Counter ever worked on a sim, so I did play with Counter banned for years. I think it's a really fun addition to the game. I also think a Counter ban will push us back towards the reflect Chansey games that everyone seemed to hate so much. It will definitely make it much harder to use Poliwrath.

 

[SoulWind]

Posting this on behalf of Peasounay

Thanks to beautiful SoulWind for posting this on my behalf, since apparently council members can't post in PRs.

I would like to address two things from the post of Amaranth

The point of the survey was to gather data to continue the discussion of your "The future of RBY on smogon" thread in a more concrete way. It was not supposed to be used as a council proof to take decisions, it was supposed to give data as to what the RBY community wants to do about its tier to continue the "so what do we do" discussion, and no one bar marco posted anything. Like technically I suppose we, the council, could have came up with a Wrap suspect and stuff but it's an oligarchic way of doing things that I don't believe to be good : the people who answered the survey are the ones that make the community and the game exist and if change there should be they should be a part of it rather than just voters to whatever we throw at them. The council in my eyes should be organizing the process, we did start it with this survey and no one answered so I guess people did not want to change the tier that badly ? If they do it's not too late though, the results are still valid.

Secondly for this specific case, I get the whole "don't mod the game just because you're happy with the discovery" argument but unfortunately as long as we have freeze clause it will always be an intellectually dishonest argument (and i'm a pro freeze clause who would never want to see it go). I think you either accept modding or you don't at all but using that argument for a case like this is a bit hypocritical. I don't buy into the whole "but then you fix 255s and hyper beam" (or whatever) because modding one glitch doesn't mean you should necessarily mod another, and if you do, the argument saying "but we modded this" is bad because it's not about the process of modding, it's about what game you want to play. If a majority of players are fine with having counter modded, then I think we should, and if at the same they don't want to mod 255s, I don't think we should. It really is only a game on the internet that has no purposes other than having fun, so all the ethical debates to me are non-sense. If it were up to me I'd ask to the current playing community "what do you want to do with counter ?" and would give them the different options. Don't see the game as an untouchable entity, the simulators in themselves are modding...

PS : "causes the game to end" is the stupidest suggestion i've ever read

 

[Amaranth]

RE: Counter, Freeze Clause, ethical debates, modding
I'm not anti-mod as a whole, but I only see it as an extreme measure for cases that cannot be handled otherwise without significantly changing the game. The alternative to Freeze Clause mod would be banning Ice Beam and Blizzard - I don't need to explain why that's unviable - or playing with no restrictions on freeze at all, which I'm sure we all agree would significantly lower the competitiveness of the game, so I take it as a necessary evil. Sleep Clause is another mod and personally I'm not even on board with it (I believe I mentioned the idea of a complete sleep ban some time ago in the thread that eventually led to the survey), but even still I understand wanting to preserve the sleep mechanic in some form and modding the game to allow it. Those mods are necessary to preserve crucial parts of the game.
The argument is not intellectually dishonest because it's simply about where you draw the line for what things are acceptable to mod out. To me the line should have been drawn even before Sleep Clause, but nonetheless, I definitely feel like Counter is not worth modding our game for, because it's a relatively unimportant move - it's a cool tech that shows up from time to time but it's nowhere near as commonplace as Ice-type attacks and even Sleep moves. It's not "intellectually dishonest", it's just arbitrary - when you have to draw a line somewhere whatever solution you choose is going to be arbitrary, so this is inevitable.
"It really is only a game on the internet that has no purposes other than having fun" - does this not give more credit to the idea of modding 256 misses then? Surely we can all agree the game would be more competitive and more fun without them? Just modding things for fun is a dangerous slippery slope in a tier where there is a fair amount of unfun mechanics.
Ethical debates aside, I think we should try to agree on which way we want to mod Counter if we're modding Counter, and then have a vote between mod or no mod. Or maybe it would make more sense to vote on mod or no mod first, and then if we decide we want to mod we can argue about which way to do it? This is exactly what the council is here for, pick a way to move forward and make a post about it please and thank you

RE: Survey
If the idea was to continue discussion this wasn't communicated very clearly at all. The months of stalling made me figure you guys were working on something and the survey was taken in preparation for actual action, apparently you just wanted to see us fill more threads? Fine then, I'll fill some threads in this subforum after SPL, we'll work something out. It's good to know that this whole thing was stalled due to misunderstandings rather than actual deliberate decisions to stall things I suppose.

 

[MetalGro$$]

Insane how 20+ years later we're still discovering new mechanics!

Anyway, I think this is a bigger issue than people think it is. Counter might have been banned for years due to not being implemented properly, it is now a HUGE part of the metagame. How many games are there without at least 1 Body Slam fired off at a Chansey? Scouting / deriving the move Counter on a Chansey's moveset has big, game changing effects. I was very impressed at SoulWind's superb patience when playing Snorlax vs Chansey sequences during his RBY Cup winning run. He'd bring a ton of Earthquake Snorlax and never click Body Slam until he was 100% sure it wasn't carrying Counter. Outright banning this move will certainly have a ripple effect on the rest of the tier. Snorlax/Tauros might use Earthquake a lot less and Gengar might start rising in usage is something that could change. I don't have to remind you how a slight change in the Body Slam mechanic completely changed the tier, so my advice first of all is to do this process slowly, carefully and correctly.

Now for my personal opinion: I know the whole ethical stance about how the simulators should stay as closely to the cartridge as possible, but I ask you: why should we punish ourselves for a gamebreaking mistake made by the programmers? I personally feel it would go against the spirit of the game to ban a move they wrongly implemented. Nowadays, they can fix their own bugs, like they did with Z-Parting Shot and if they could do that in RBY they most certainly would do it, so why would we not do it for them? And before someone throws the 1/256, Focus Energy and recovery glitches in my face: these quirks don't cause a desync and keep the game playable. I'd definitely be in favor of any sort of Counter Mod and I feel we should definitely at least hold it to a vote within the playerbase.

Big thank you for whoever gave me the ability to post in Policy Review btw

 

[Zarel]

It's pretty common. Pokemon are paralysed about as often as they are unparalysed, and counter is probably on about a quarter of teams?

[Edit] - for the benefit of people unfamiliar with competitive RBY, I am absolutely confident that there would be no meaningful objection to a "cause the game to end" implementation resulting in the immediate banning of Counter for all competitive RBY.

If the rule is "the player who uses Counter loses if it causes a desync", I see no need to ban Counter. I would imagine this would be how people would play it on cart, after all. It would let the player decide between "don't use Counter" or "only use it in safe situations" or "risk losing".

edit: someone else said "cause the game to end" is a stupid suggestion; so I think I need to point out that I'm a sim coder and all I'm doing is pointing out possibilities. Which possibility you end up choosing is not something I care to have a hand in.

 

[Hipmonlee]

I don't like idea of people suddenly instalosing because they didnt realise that an FP on an earlier turn primed Counter for one player and not the other, and even though the current pokemon they were facing when they selected Counter was not Paralysed, because their opponent switched instead of attacking like they expected Counter looked back to that earlier turn and caused a desync.

Most people already dont really understand the mechanics of Counter particularly well.

I think we'd be better off just banning it.

 

[Hipmonlee]

So, out of this discussion there seem to be only two options that have any support whatsoever.

1. Ban Counter
or
4. Patch Counter so that it will fail whenever a desync would occur.

That's pretty convenient because it means we have basically agreed what a patch would look like if we decide to patch this. And the only question now is whether or not Counter is such an important part of the game that we should go to the extreme measure of patching how it functions in order to keep playing with it.

To me that seems like a pretty good question to put to a community vote. That is obviously a decision for the Council to make though.

 

[Isa]

if the council chooses option 4 i petition that psywave be legal again and patched to fail when it'd cause a desync (based on my current understanding that psywave is in fact currently illegal to use due to causing desyncs). i see no reason to modify one move to prevent desyncs but not another.

 

[Hipmonlee]

Psywave is definitely not such an important part of the game that we should go to the extreme measure of patching how it functions in order to keep playing with it.

 

[Lusch]

Bumping this thread because I seriously believe something should be done about Counter. Basically, I don't have something really new to add, but MetalGro$$'s post sums up my position perfectly. We should absolutely not ban Counter. This would be huge and makes Snorlax's life even easier. Not to mention the scouting porcess and metagame changing effects this would likely draw in. I am also pretty sure that those changes would make the metagame worse. Not gonna go into much detail here, because it would only repeat Metal's post...

Option 4 is the one we should go with. I thought I'd just try to spell out what this means in detail.
First, there is the situation with the paralyzed Pokemon. Whenever this Pokeon gets fully paralyzed on a turn it used something uncounterable, while the opponent uses Counter, Counter simply has to fail, because a desync happens otherwise. Not that we should put overly much weight on that, but personally, I think this is also a great change to the move's mechs, since it allows for counterplay in that situation. It's just stupid that you have to hope you get trough para, even when you click the right move to avoid Counter.

Second, there is the situation where one player switches out when the other one uses Counter. Currently, if the switching Pokemon used a counterable move in the turn before switching, Counter will work on the incoming Pokemon. However, in game you can just put the curser on a non-counterable move and cause a desync. (watch Crystal's video linked by Lutra above). So it is not entirely clear how to mod Counter in that situation, since there is no fight butten and curser to select moves with. My suggestion would be: Whenever the switching Pokemon in that situation has a move on its moveset that is not counterable (which admittedly nearly every Pokemon has) Counter should fail, because you gotta give a knowledgable RBY player the credit that he would (if he could) select a non-counterable move before switching in that situation. It would practically mean: no more countering something on a switch-in. Personally, I always thought that was cheap anyway. Surely there are situations where the Counter user wants to provoke exactly that situation (for example after an Explosion), but sometimes it's just blind luck to use the move randomly when the other guy switched and catching something else.

I think it is totally acceptable to mod Counter to let it fail, when the circumstances are met. Surely Counter gets a bit weaker because of it, but it still makes Snorlaxes check for it and does not lose its primary purpose which is hitting Snorlaxes (or other Pokemon) on the same turn they use a counterable move, the way Counter clearly is meant to work.

If people can agree on this kind of mod (make Counter fail, whenever is causes a desync), then this is to me clearly the surperior option. It keeps a crucial part of the metagame intact. In my opinion it's something that should be implemented. It's important to make a differnece here between arbitrarily modding the game (which should only be done in extreme circumstances. Freeze and Sleep clause both are necessary, modding 256s is not. Call it arbitrary, if you want, I don't agree.). Modding desyncs (especially when it's such an imprtant part of the meta) is not comparable to "making the game how we like it". On that note, I would not be opposed to mod Psywave in the same way, make it fail whenever it causes a desync, sure.

 

[Heroic Troller]

Counter causes a bug? Counter should be banned.
Psywave, Fly and Dig never got patched, as much as i like Counter's role in the meta i'd like to keep the tiering at least somewhat coherent with itself, it's not like we can arbitrary decide that THIS move is too good for our ends to get the same treatment every other move had prior. If you want to go for the most enjoyable metagame and twist your own rules, fine by me, but then i want to have a good talk of 1/255 misses and other stuff as well following. Because you all know if it was any other move it was getting a quick ban and be done with it instead of this roundabout way to keep it where it is, coherence above all i say.

 

[Hipmonlee]

Second, there is the situation where one player switches out when the other one uses Counter. Currently, if the switching Pokemon used a counterable move in the turn before switching, Counter will work on the incoming Pokemon. However, in game you can just put the curser on a non-counterable move and cause a desync. (watch Crystal's video linked by Lutra above). So it is not entirely clear how to mod Counter in that situation, since there is no fight butten and curser to select moves with. My suggestion would be: Whenever the switching Pokemon in that situation has a move on its moveset that is not counterable (which admittedly nearly every Pokemon has) Counter should fail, because you gotta give a knowledgable RBY player the credit that he would (if he could) select a non-counterable move before switching in that situation. It would practically mean: no more countering something on a switch-in. Personally, I always thought that was cheap anyway. Surely there are situations where the Counter user wants to provoke exactly that situation (for example after an Explosion), but sometimes it's just blind luck to use the move randomly when the other guy switched and catching something else.

This seems like an overwrought solution to something that is not a problem. There is a trivial solution here, which is on cart battles you ban players from changing the selected move in the attack menu before switching. This is implemented by default on Showdown.

I get that we would be changing the mechanism of counter anyway, but the goal should be to keep things as similar to the cart as possible. Where there's an easy win we should stick with that.

 

[Lutra]

This seems like an overwrought solution to something that is not a problem. There is a trivial solution here, which is on cart battles you ban players from changing the selected move in the attack menu before switching. This is implemented by default on Showdown.

Why should the counterplay to counter be prevented though and not counter itself? It's not a valid counterplay if it causes a desync, but at the same time that usage of counter isn't either. I don't think counter should be made stronger just because the counterplay has no other utility or appears to break some previous assumption and seem stealthy.

 

[Hipmonlee]

There is no counter desync on that turn without that specific counterplay, and that specific counterplay is trivial to ban. Patching in that situation takes us further away from how the game would be played on the cart.

I'm gonna take a step back and talk about why I think option 4 is acceptable. Which might help clarify my position.

So lets assume we have two players playing on cartridge. The two players want counter to be legal and both would rather play the game to its conclusion and lose than try to intentionally cause a desync.

With option 4 the only reason that I can think of that those two players would chose to use counter differently than on Showdown is if they were intentionally trying to remove any risk of an accidental desync. In this situation using counter on a desync turn is always a negative for the user of counter, a desync is obviously going to be a stronger negative than wasting a turn with a failed counter, but it is as close as we can make it. I would call this a satisfactory outcome.

To justify your scenario we have to assume one of the players is intentionally trying to cause a desync, therefore the way the game is played between cart and showdown is suddenly very different. One player could be incentivised to use a move in the cart game (countering to cause a desync) that they would be incentivised to not use on Showdown (counter would just waste a turn). If we went down this line then I would be in favour of counter being banned.