It seems to me that there's at least 8 different models to handling sleep that could be considered. I'll be assessing these models based on the criteria I explained previously.
Standard Model
The standard model is the current implementation of Sleep Clause on Showdown for BW DOU.
Cartridge Accuracy
Of course, to no one's surprise, the standard model fails the criterion of cartridge accuracy. It does no good to reply that Sleep Clause existed in some form in past official titles; in Generation 5, Sleep Clause is in fact nowhere to be found.
Metagame Impact
Also to no one's surprise, it passes the criterion of metagame impact. This just is how the metagame is currently. I also think the metagame with respect to sleep is perfectly healthy right now, so maintaining the status quo is just fine.
Being Easy to Understand
Perhaps more controversially, I do think it fails the criterion of being easy to understand. Because all other modern generations of Doubles OU do not have sleep clause, playing in a tier without it can often feel foreign to newer players and even veterans, to the point where it is a common mistake to straight-up forget about Sleep Clause. Take two tournament examples from 2023: Doubles Premier League Week 1 - tennisace vs Lunarr (
https://replay.pokemonshowdown.com/smogtours-gen5doublesou-680147) and [I would have included
BW Cup round 1 - fangame vs teals, but the replay is lost. tl;dr is that fangame Spored in a situation where Sleep Clause would have always applied no matter what).
In tennisace's game, after Heatran took a sleep turn, it could have potentially woken up. The reaction from tennisace makes it clear that they had forgotten about Sleep Clause however, trying to put Jellicent to sleep instead of Giga Draining for damage. Examples of this can be multiplied, including where players attempt to break Sleep Clause
multiple times in the same game. Now, in one sense, this is the player's fault; you might compare it to the player who doesn't know that Prankster moves weren't blocked by Quick Guard in BW. But I'd like to give these players the benefit of the doubt and say it's at least partially due to how non-obvious Sleep Clause is.
I also have heard players who think that Sleep Clause means you can only Sleep one Pokemon for the entire game, and don't realize it just means you can't have two of their Pokemon currently put to sleep. As much as I love to joke about people not understanding Sleep Clause, at the end of the day, it's because it's not that simple.
Being Easy to Implement
The fourth criterion of being easy to implement is obviously passed, because it is already implemented on Showdown.
Self-Sleep Standard Model
Our first competitor to the standard model is what I'll call the self-Sleep standard model. This is the old implementation that the Pokemon home console games like Stadium and XD used, where self-inflicted sleep like Rest counts towards Sleep Clause. Although it's similar, there is some nuance in how the criteria are applied.
Cartridge Accuracy
The criterion of cartridge accuracy of course is still not met, since Generation 5 didn't have Sleep Clause. Again, it does no good to reply that this is how it was implemented when Sleep Clause was actually a thing – in fact, there is no Sleep Clause available in Generation 5 cartridge play.
Metagame Impact
How about the criteria of metagame impact? Here the impact is admittedly rather low. It buffs Rest to a degree, but BW doesn't really see many Rest users. Cresselia and Suicune have used Rest in the past, but they're definitely on the more niche end. It would also open up the option to side-Spore your own Pokemon with Amoonguss, letting you choose what gets Sleep sacked. I don't anticipate this being a play you'd regularly see. Overall, you get some niche additional counterplay.
Being Easy to Understand
What about being easy to understand? Well, it is a deviation from the standard Sleep Clause implementation, so you might confuse some players used to the standard model who try to Spore an opponent when their ally has used Rest. But really, this model suffers from the same problem of players forgetting it exists that the standard model has. If anything, it makes the problem worse.
Being Easy to Implement
Fortunately for this model, it is already implemented on Showdown, since Showdown supports simulating Pokemon Stadium. So it passes the criterion of being easy to implement with flying colors.
Removing Sleep Clause
Cartridge Accuracy
Next, let's consider that alternative the old VGC players brought up: removing Sleep Clause in its entirety. You could put as many Pokemon to sleep on the opponent's side as you wanted. Obviously, this passes the criterion of cartridge accuracy, since it is possible to do this on cartridge.
Metagame Impact
The real issue with this implementation is the criterion of metagame impact. Unrestricted sleep increases the relative strength of Amoonguss significantly in BW. This is because for the most part, there are basically no switchins to Spore, unlike in future titles with Safety Goggles, Grass-types, or Misty and Electric Terrain. Most of the existing counterplay, like Taunt or Substitute, requires the Pokemon already be on the field. Take this example [would have used the example of
DPL 9 finals, me vs mishiimono, but the replay is lost. I used it as an example of what Amoonguss could do in TR without Sleep Clause active]. Now, my point really isn't so much that this is better or worse – it's just that the dynamic of this game in particular changes significantly without Sleep Clause. Now, granted, you can look at plenty of BW games and find matches where the opponent pins down Amoonguss and never lets it get out of hand. But I think lots of players would want to give additional respect to Amoonguss in the teambuilder and that changes teambuilding significantly. For example, I think that Jellicent gets significantly worse with Sleep Clause enabled – it never runs Protect, and would have to resort to dropping a critical status move for Taunt if it wanted to keep up with the increased mushroom usage.
But perhaps I'm overreacting – after all, VGC didn't have Sleep Clause, and their metagame was very similar. So how did VGC players react? Well, in 2013 at least, the format most similar to BW DOU, Amoonguss consistently stayed towards the top of the usage stats – across the top of
European Nationals,
US Nationals, and
that year's World Championships, Amoonguss was the 5th most used Pokemon. An Amoonguss or Breloom was featured on the winning teams of every champion. Top 8 Worlds finisher Enosh Sachar was using an unorthodox Timid max Speed Amoonguss to outspeed Pokemon like Latios in Tailwind! Teams would come well-prepared to handle Amoonguss, be it Taunt Thundurus, Substitute, their own Lum Berry Amoonguss, and sometimes even Lum and Chesto Berry on the same team. More unique counterplay included Ray Rizzo's Fire Gem Eruption Torkoal, which keep in mind was pre-Drought, and Matthias Suchodolski's Safeguard Bronzong. It's not as if, however, Sleep was such an overwhelming issue that you saw, say, Lum Berry Safeguard, which was a metagame staple on Bronzong in VGC 2016 as one way of dealing with the oppressive Dark Void Smeargle.
I do think there is an important difference between the two metagames, however, and that's how impactful just 1 sleep is. In VGC, it is proportionally worse for one Pokemon to become inaccessible on your team than in Doubles OU; you're comparing 1/4 to 1/6. Because of sleep's heightened impact, that means teambuilding to maximize prevention of
any sleep is super valuable, unlike in 6v6. Take a look at
Arash Ommati's winning Worlds team from that year. He had not only 2 defensive measures to Amoonguss in Substitute Heatran and his own Amoonguss with Lum Berry, but also several OHKO measures, including Tornadus and several double target options – in fact, every team member had a way to threaten Amoonguss in some meaningful way. This isn't even touched on in his report, likely because it's just part of teambuilding in that era.
Personally, I am quite happy with the metagame we have in BW DOU right now, so I am not eager to change it if I don't think the benefits are there. In my judgment, I would suspect removing Sleep Clause doesn't improve the player's experience in BW DOU, either in play or in teambuilding. I think it's not something that can be completely ruled out on theory alone, but I do think the impact of this change is likely negative.
Being Easy to Understand / Being Easy to Implement
Our other two criteria are as straightforward as the first. The only way it doesn't pass the criterion of being easy to understand is if you're considering folks used to BW who might be caught unaware by the rule change, but consistency with other Doubles formats is a big plus. And of course, it's trivially easy to implement on Showdown, since all you have to do is remove that clause from the tier.
Sleep Clause as a Win Condition
Our third model to consider is to enforce Sleep Clause by treating it as a win condition. If you put two of your opponent's Pokemon to sleep, you lose the game. On this model, you could enforce such a win for both controllable and uncontrollable Sleep Clause violations, or just for controllable violations. What do I mean by controllable and uncontrollable here? Well, suppose your opponent clicked two contact moves into your Effect Spore Amoonguss, and by chance the Ability triggered and put both opponents to sleep. In such a case, you might consider this an uncontrollable violation of Sleep Clause and not award a win. I don't think that's a good thing myself, but I could imagine someone arguing for it.
Cartridge Accuracy
First up, the criterion of cartridge accuracy. While there's of course no mode in the Generation 5 games, or any Pokemon game for that matter, that awards a win if the opponent successfully puts two of your Pokemon to sleep, nevertheless this is cartridge-accurate in the sense of it being enforceable. It's just like how both players show up to a match agreeing not to turn the game into an endless battle, or how both players could show up in a game of Generation 8 and simply agree not to Dynamax. Traditionally, this is how Sleep Clause is "enforced" on cartridge across generations. If you end up putting two Pokemon to sleep, you lose the game, because both players agreed to follow the Sleep Clause rule, even if the game doesn't explicitly enforce it.
Metagame Impact
How does this sort of Sleep Clause stand up to the criterion of metagame impact? I think it has very marginal impact. Basically, it makes plays where you try to resleep an opposing Pokemon to cover a wakeup extremely risky. For example, suppose a Cresselia has already taken 3 turns of sleep, and you want to use Spore to cover it waking up and attacking, putting it back to sleep. This play is not safe - if Cresselia switches out, you just cost yourself the game, because you just put two Pokemon to sleep. I highly doubt this would work as a strategy you could abuse intentionally to
force the opponent to put multiple of your Pokemon to sleep. One hypothetical way to do it would be to Trick a Choice item onto Amoonguss with Gothitelle, trapping it in, and then switching in another Pokemon with Follow Me or Rage Powder to try and get both of your Pokemon sleeping.
Being Easy to Understand
But realistically, it's very likely, in my opinion, that the vast majority of Sleep Clause violations are going to be from player error, which leads to the third criterion of being easy to understand. In my judgment, this model fails this criterion miserably. This is unlike how Sleep Clause is implemented everywhere else on Showdown. This means newer players who don't understand the nuances are punished not just with a wasted turn, but an entire game loss. That's a pretty steep price for the player who decided to try out the tier for fun in a room tour.
Being Easy to Implement
I also think this model faces problems with the criterion of being easy to implement. What happens if a player breaks Sleep Clause? Does the turn just immediately end, right then and there? Imagine if you play the turn out to its completion. You could end up with scenarios where you break Sleep Clause mid-turn, but say, KO the other sleeping Pokemon on that same turn so you don't end up breaking it, which seems wrong. Or consider what happens if both players try to break Sleep Clause on the same turn - what should happen then - a tie? Keep in mind that Sleep Clause violations don't have to be mistakes - they could just be calculated risks. Both players might just happen to go for a risky play of attempting to resleep their opponents. So then you're ending games during the middle of the turn, which isn't exactly cartridge accurate either and may cause some trouble for PS, though I don't think it would be the end of the world. If you try to implement the distinction between controllable and uncontrollable methods of sleep, that's also going to cause additional implementation headaches. Minimally, it's not as straightforward as other models to implement.
Can't Select Sleep Moves model
The fifth model we'll consider is very similar to the previous model, but differs in a crucial way: Sleep Clause would prevent a player from clicking a Sleep move on the opponent if they have already put at least one Pokemon to sleep. Another way to put it is that sleep is prevented on move selection, but not move execution. This would be implemented via user interface and server-side restrictions. You simply wouldn't be able to click Spore, then target the opponent, if one of their mons was already put to sleep by you – Showdown just wouldn't let you click the target or type out the equivalent command. The crucial difference between this model and the previous one is that Sleep Clause would never result in a game loss.
Now, you might initially think we could just disable the move, and just gray out Spore like how Choice items work for example. But that won't work because there's some board states where you
have to click Spore twice. For example, suppose you have a Choice Scarf Breloom, and lock into Spore. But oh no! The opponent has switched into Gothitelle. If you make Spore unclickable, what happens here? Does Breloom have to use Struggle? You don't really want to allow the Breloom to click Spore on the opponent either though, because what if the sleeping Pokemon just switches out? Luckily for us, this is
Doubles OU, and there's still another target available – your ally! You can always target your ally, even a fainted ally, and never break Sleep Clause. We'll talk later about whether or not this is a good thing, but at this point I'm just trying to outline the model.
As you'll recall with the "breaking Sleep Clause is a loss" model, uncontrollable elements like Effect Spore or Relic Song are also factors to consider. The current model we're discussing doesn't assign losses for breaking Sleep Clause, so it does allow for potentially multiple Pokemon to be put to sleep via random means.
Also worth noting is that the restriction for selecting a Sleep move goes away if the opponent only has one Pokemon remaining, as it's impossible to break Sleep Clause in that scenario.
Cartridge Accuracy
How does the "can't select Sleep moves" model fare against our criteria? First, it passes the criterion of cartridge accuracy. It's very similar to the "breaking Sleep Clause is a loss" model in that regard; the only difference is that Showdown prevents the player from clicking a move, in a similar way to how it prevents the player from clicking Dynamax in formats with Dynamax banned. The game is never modded to make sleep moves fail, force a Pokemon to Struggle when they otherwise wouldn't, and so on. It's an agreement made by both players going into a match, similar to the agreement they make not to bring Zekrom or Gravity that's also automatically enforced.
Metagame Impact
What about the criterion of metagame impact? I think it's important to note that the vast majority of players already play Sleep Clause in this way. When you click a sleep move, you're already not going to click it again, because that would just be a wasted turn in most scenarios. However, what about those scenarios where it might not be a wasted turn? Take for example a board state where the opponent already has taken 3 turns of sleep. You might be tempted to Spore this turn, because it's pretty likely the opponent will want to take advantage of the guaranteed wakeup and attack. However, on the "can't select Sleep moves" model, this play would be restricted, because the opponent could switch, which would break Sleep Clause. A similarly restricted play would be using Spore with Amoonguss and Breloom on the same turn into a Pokemon you're confident has Lum Berry. This is because theoretically the Pokemon could have Eject Button and be hit by their ally's attack, forcing them to switch out mid-turn and thus breaking Sleep Clause. Other examples where you
probably wouldn't break Sleep Clause, but nevertheless hypothetically could exist, such as feeling very confident the Chansey with Natural Cure will switch out, but what if it doesn't, etc.
Remember again our hypothetical of a Choice-locked sleep user locked in vs Shadow Tag. As we said there, this means the Shadow Tag user has a devastating punish, since in order to avoid potentially breaking Sleep Clause, the sleep user is forced to target their own ally, putting it to sleep. Now, this isn't so different to the "breaking Sleep Clause is a loss" model, since your opponent might see the opportunity and switch out their sleeping mon to try to cheese a win, but it is pretty significantly different from the standard model, where the Choiced Pokemon in such a situation is just punished with their attacks failing a few times.
Now, all of these scenarios are pretty strange. But the proponent of the model will likely point out that, while strange, this just is what you would have to do on vanilla cartridge mechanics if you didn't want to risk a loss. Plus, it's not the responsibility of the player to
know not to click the sleep move. Showdown is enforcing this move selection restriction automatically.
Also of note is that in my research for this video, I surveyed the top 8 of all 4 BW Cup Smogon has held, which is the primary standalone annual BW DOU tournament, as well as all of the BW games of Doubles Premier League for the last 3 years [note: remember this was as of May 2023]. This was a fairly large sample size among intelligent players, and I could not find a single game of a player intentionally Sporing their opponent when they already had already put one Pokemon asleep. I wanted to pull tournament replays to showcase examples of this concept, but I literally could not find one. That's not to say they don't exist, but I do think it eases worries of concerns about the impact of intentional Sleep clause breaking plays on the metagame.
This implementation does potentially face problems with another game mechanic though: moves that call other moves. Sleep Talk, Assist, and Copycat are most notable. The problem with these moves is that they don't target the opponent – they target the
user. When these moves are used and call a single-target attack, they actually target either opponent randomly. So, if your Choice Band Sleep Talk Kyurem-Black happens to call Dragon Claw, you don't know if Dragon Claw will target the opponent's left slot or their right slot. This is a problem because we can't restrict these moves that call other moves at move selection in the same way that we could for the single-target sleep moves. Here's a basic example: suppose a Choice locked Amoonguss only has the moves Sleep Talk and Spore, and becomes Choice locked into Sleep Talk while trapped by Shadow Tag. In such a case, it has to click Sleep Talk – there's literally no other options! That means it could Spore both Pokemon if the rolls happen to align just right. Is this likely to happen? No, but it and related situations means you can't restrict moves that call other moves with this clause, even if there's the potential for the player to put more than one opponent to sleep on the opponent's side. So, what does that actually mean? Well, you could build a gimmick team built around Assist, where the only callable move is Spore, and the rest of the team's moves are filled with attacks like Thief, Dig, or Focus Punch to deal damage. I think such a team would obviously suck and not likely to actually win games. Why? Because you're using untargetable Spore paired with terrible attacking moves - eventually, the opponent will just KO Liepard and Purrloin and wipe out the rest of the cheese without breaking a sweat. What about Copycat? It's not really a concern because the only ranked Pokemon that even gets Copycat in Gen 5 is Chansey. What about Sleep Talk + Spore, or perhaps Sleep Talk Rest Psycho Shift Cresselia or Deoxys-Defense? Well, if you want this to be reliable, you have to run only those moves – any more ruins the odds of the combo working, and you also have to take the time to set up those janky board states, requiring at least one other Pokemon to either put your ally to sleep or self-damage for Rest, and all for the reward for potentially putting multiple Pokemon to sleep via 50/50 RNG calls. Overall, although this model of Sleep Clause means these sorts of strategies are
possible, and can't be restricted programmatically, nevertheless their impact is so low that any player who plays BW would recognize they'd have no actual effect on the metagame.
What about RNG sleep-inducing effects like Effect Spore or Relic Song? Technically, this model of Sleep Clause gives a marginal buff for those two, since those effects could bypass it. However, nearly everyone recognizes the superiority of Regenerator over Effect Spore on Amoonguss and likewise for Technician on Breloom. Effect Spore randomly picks between poison, paralysis, and sleep, and will inflict one of those statuses 30% of the time to an attacker than makes contact with Amoonguss. So to break Sleep Clause with Effect Spore, the opponent has to use a contact move into Amoonguss, and it has to roll the chance for a sleep. Keep in mind that if it rolls poison or paralysis, it could block a Spore aimed at that Pokemon, causing Effect Spore to backfire! As for Relic Song, I don't even know Meloetta's stats because it's so bad that I've never bothered to look at it, even for this video. So it's not a concern either.
Overall, in terms of metagame impact, while these edge case scenarios are funny to think about, realistically not a whole lot changes. There's a few plays restricted from a player that are available on the standard model, but the tier itself doesn't really change a whole lot. I think even with the hypothetical ways to sleep multiple opponents, this model actually represents a net nerf to Sleep, and overall it passes the criterion of metagame impact.
Being Easy to Understand
What about the criterion of being easy to understand? This one is much more of a mixed bag. On the one hand, you literally can't screw it up – you can't waste a turn or lose the game clicking a sleep move when you shouldn't, because Showdown will prevent you from clicking it. On the other hand, if you don't understand how sleep clause works, you might be confused as to
why you can't target the opponent. I would compare it to someone who tries to bring a Pokemon with Trick and Assault Vest – it might not immediately click why they can't Trick the item away to prevent status moves, but the game prevents you from doing it. There's also something to be said about all the hypothetical scenarios we talked about and the "gotchas" they bring. It may not be immediately obvious to a player how it would work mid-battle if say, they're the ones with a sleeping Pokemon and can't see that the opponent can't target them anymore with sleep moves. Overall, I don't think it passes the criterion of being easy to understand, but I do think this is mitigated significantly by the game handholding you through it.
Being Easy to Implement
What about the criterion of being easy to implement? Well, right now, there is nothing comparable to this implementation in the codebase. However, the footage you've been shown to represent this model is from code I wrote myself! [note: obviously there is no video, it would have been used as a visual aid in the video]. It took me a couple hours, but it all works great! The logic for Sleep Clause preventing you from selecting a sleep move if the opponent is already asleep is based on how Curse changes target based on whether or not the user is Ghost-type at the time of move selection. Preventing players from double clicking Spore on the same turn was based on how we prevent, say, trying to click 2 Z-moves on the same turn. Now, this hasn't gone through standard code review or anything yet, but I think the criterion of being easy to implement is reasonably met, because, well, I've done it!
Ban Sleep Moves
Shifting gears to our sixth model, we encounter something strikingly simple: simply ban all Sleep moves. So in a similar way to how you can't bring Dark Void to a game of BW DOU, you also wouldn't able to bring Spore, Hypnosis, and so on. As you'll recall, this is the implementation singles decided to adopt to deal with how powerful sleep was even with the standard model active. Of note, Relic Song and Effect Spore are not banned in BW OU, though this could an option if deemed necessary.
Cartridge Accuracy
Just like any other thing banned in a Smogon metagame, this model passes the criterion of cartridge accuracy. Both players agree that when playing BW DOU, they won't bring Zekrom, or Fissure, and this is just another agreement.
Metagame Impact
Of course, the big debate in this model is the criterion of metagame impact. What this model effectively does is kill off Amoonguss. Although Amoonguss gets other status moves like Stun Spore or Toxic, realistically that isn't going to save its viability. Take a look at Togekiss, which could hypothetically do the same thing with Follow Me and Thunder Wave, but nobody does that and Togekiss isn't even a ranked Pokemon. Now, the defender of this model might reply that the
reason nobody does that is precisely
because Spore Amoonguss fills that role better. But if you took Spore away, Amoonguss is still viable, just nerfed. I think that's just mistaken. Amoonguss's offensive pressure in BW especially is driven by Spore, as opposed to just redirection. This is most apparent in the concept of sleep sacking, where you're intentionally giving up a Pokemon to Spore in an attempt to make the game a 5v5. The only reason that works is because Amoonguss's value drops so much after using its initial Spore. That's not replicable with Toxic or the inaccurate Stun Spore. There's already so many viable Thunder Wave users that nobody is going to use a version of it that's only 75% accurate. The team composition that requires specifically Amoonguss with paralysis is obviously niche. As for Breloom, while I do think it represents a nerf, it doesn't really change much else. Scarf sets are probably impossible, and all Breloom sets would probably just use Bullet Seed, Mach Punch, Superpower, and Protect, instead of deciding between Superpower and Protect. That just means Breloom's not as threatening vs Pokemon like Cresselia or Metagross, or in trading Sash for a sleep. Breloom is less likely to appear on a team vs competitors like Conkeldurr, but it's not like losing Spore makes Breloom unviable. So it all comes down to Amoonguss's impact on the metagame. Rain teams which rely on Amoonguss as a defensive pivot are now forced into either Volcarona or Togekiss as a redirection Pokemon, likely leading to more offensive rain compositions, which now have one less answer with a key Water resist plucked out of the metagame. Balance teams which use Amoonguss as support to get up Trick Room or Substitutes are similarly left with less than ideal options. Trick Room teams that would be slowed down by Spore trades from Amoonguss have freer reign. The list goes on. Whether you find these things good or not is up to the player's opinion, but BW's metagame does clearly shift, and I personally am quite happy with its current state. Minimally, even if you don't like the current BW DOU metagame, it's unlikely that's because of Amoonguss, and probably more because you don't like how ubiquitous Thunder Wave is as Speed control. So I don't think banning Sleep moves passes the criterion of metagame impact.
Being Easy to Understand / Being Easy to Implement
It does however pass the criteria of being easy to understand and easy to implement with flying colors. Since it's enforced at the teambuilder level, you can't screw up at all and lose a turn or the game. And it's really straightforward to implement: just ban the moves.
Ban individual Sleep move abusers
The next model of Sleep Clause is similar to the last one, but instead of banning all Sleep moves, you instead ban individual abusers of Sleep moves. To be clear, this isn't directly restricting sleep in any way; it's just removing Sleep Clause and banning broken Pokemon that appear as a result of a lack of Sleep Clause.
Once again, this model passes the criterion of cartridge accuracy with flying colors, but struggles with the same issues as the "ban Sleep Moves" model in terms of metagame impact. Except this is much worse, because you're likely going to end up banning way more than just Amoonguss. Scarf Breloom's Spore is broken - just ban it. Politoed spamming Hypnosis as a filler move is broken - just ban it. People using Eviolite Foongus? Just ban it. If I was already concerned about the potential negative impact that banning just Spore from Amoonguss would have, you'd better believe that ripping out Politoed from the metagame would just be ridiculous. It's also not simpler to understand or simpler to implement. Instead of targeting the actual issue with the ban, Sleep moves, you're targeting individual Pokemon that can abuse the mechanic. It's not hard to implement on Showdown in the sense of banning the individual Pokemon, but the whole process of suspecting each individual Sleep user would be a time-consuming, repetitive process likely not fit for an oldgen like BW DOU. I find this model unlikely to have many supporters; it's not even really a model at all.
Mod Sleep in another way
The final model to discuss is a more novel idea; if we're OK with modding sleep through the standard model, why not simply mod in other ways instead? We could modify sleep to behave like the generations following black and white, where Sleep turns are not reset on switchout. If that's not enough, we could introduce additional nerfs, such as making it so the number of sleep turns is known and displayed visibility to both opponents. This information is known by the game already, since the number of turns a Pokemon is asleep is determined when that Pokemon is put to sleep. And if that's not enough, perhaps the maximum number of sleep turns could be reduced to 2.
Cartridge Accuracy
This is our first model in a while that does not pass the criteria of cartridge accuracy; in fact, it sort of flaunts it altogether by suggesting we just modify the mechanic in different ways. While novel, I think most players would agree that it goes considerably too far. Modding how the game's mechanics works is supposed to be a last resort, when no other options are available. At least Sleep Clause has some precedent, even if not quite the same implementation and no direct support in the mainline series, but modifying how the Sleep condition itself works is too radical a move to be consistent with tiering policy.
Metagame Impact
What about metagame impact? It probably depends on how much you nerf sleep, but even with all the nerfs I laid out in my opening comments, where you don't reset on switches, reveal the Sleep turns to both players and even reduce the maximum number of sleep turns, I think Amoonguss and Breloom would actually get better, because the fundamental issue of being unable to switch defensively vs Spore still remains.
Being Easy to Understand / Being Easy to Implement
It also fails the criterion of being easy to understand because, ya know, you're changing how the game works, and that's likely going to be confusing for players that anticipate Showdown to simulate how the game works. It may or may not pass the criteria of being easy to implement, depending on what nerfs you assign to sleep. Really, the criterion of cartridge accuracy weighs this model down so much that it would take significant benefits in other areas to counterbalance, and they just aren't here.
