-
Check out the relaunch of our general collection, with classic designs and new ones by our very own Pissog!
-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
Cartridge Sleep Clause mechanics
- Thread starter RBG
- Start date
I agree. For such a big change such as this we should test it out first instead of theorymoning different things about it. It would allow for more users to get involved, used to it, and it would reduce the chances of us missing something important. Perhaps we could try:
- Make a test server with a ladder implemented to test it out.
- Run some tournaments with players giving feedback after each round or after the tournament of how it goes. (Hmm tournament feedback... that might be a nifty sub-forum idea.)
And I agree with Earthworm that achieving victory other than being able to KO all of your opponent's mons does not seem right.
I'm going to play the Devil's Advocate here for a moment and ask... How broken is it to allow just one pokémon to be put asleep by the opponent and not two?
Most wifi tournaments have used essentially these rules,with maybe some hairs split over the finer details.
The only time this is reasonably exploitable is in Ubers with Wobbuffet. Except the most common sleeper in Ubers is Darkrai who destroys Wobb.
You wont even notice it, aside from the situations that have been mentioned as problems with the current clause (spamming sleep to catch a poke as it wakes, or against sleeped natural curers.
Have a nice day.
The only time this is reasonably exploitable is in Ubers with Wobbuffet. Except the most common sleeper in Ubers is Darkrai who destroys Wobb.
You wont even notice it, aside from the situations that have been mentioned as problems with the current clause (spamming sleep to catch a poke as it wakes, or against sleeped natural curers.
Have a nice day.
Why not include an exception?
"Sleep-inducing moves used while the user is unable to switch moves and trapped by the opponent do not trigger Sleep Clause."
Essentially, if the user is Choiced or Encored and trapped by either Mean Look, Block, or Shadow Tag (but not Ingrain!), Sleep moves that they perform will be exempt from Sleep Clause. This doesn't affect play much because a: Wobbuffet carries Safeguard, allowing it to protect its switchin from Dark Void coming from Darkrai, and b: The obligation is on the player using Mean Look to ensure that they do not trap someone into a Sleep move.
"Sleep-inducing moves used while the user is unable to switch moves and trapped by the opponent do not trigger Sleep Clause."
Essentially, if the user is Choiced or Encored and trapped by either Mean Look, Block, or Shadow Tag (but not Ingrain!), Sleep moves that they perform will be exempt from Sleep Clause. This doesn't affect play much because a: Wobbuffet carries Safeguard, allowing it to protect its switchin from Dark Void coming from Darkrai, and b: The obligation is on the player using Mean Look to ensure that they do not trap someone into a Sleep move.
Well if you killed something with Wobbuffet, then I could send out scarf Darkrai and you wouldnt be able to use Wobbuffet to take the sleep hit in that situation..
It would encourage people to send in Darkrai to sleep things in that situation. Other pokes wouldnt be able to do it unless the trapper was dying from hail or something.. Something extremely rare anyway..
But other than that I dont really have an issue with it. It's not overly complex, and it is basically never going to happen except in Ubers anyway.. And the effect on Ubers shouldnt have much impact on the rules we make.
Have a nice day.
It would encourage people to send in Darkrai to sleep things in that situation. Other pokes wouldnt be able to do it unless the trapper was dying from hail or something.. Something extremely rare anyway..
But other than that I dont really have an issue with it. It's not overly complex, and it is basically never going to happen except in Ubers anyway.. And the effect on Ubers shouldnt have much impact on the rules we make.
Have a nice day.
Actually it wouldn't. Choiced Darkrai locked into Dark Void gets stalled out by Wobbuffet with Safeguard, so it's never in the user's best interest to use choiced Darkrai to sleep Wobbuffet. Non-choiced Darkrai is irrelevant in this argument because it can choose any move, so it breaks sleep clause either way.
Let's say Darkrai is choiced. (that's the only Darkrai that matters here)
Bad Dreams deals 12.5% per turn, Leftovers heals 6.25%.
Sleep lasts 1-4 turns, Safeguard lasts 5 turns. Wobbuffet loses 6.25%-25% HP while asleep, while he is guaranteed to gain at least 31% back when he Safeguards. As such, unless Wobbuffet is at 25% or less, it can ALWAYS stall out Darkrai's Dark Void. Choiced Darkrai versus Wobbuffet is a really shitty matchup for the Darkrai user even if it doesn't cause auto-loss, since the Wobb user has nothing but options. He can either force Struggle or set up Safeguard and let something scary set up.
Even if Wobbuffet is the second thing slept on the guy's team, it's nothing but good for him.
Bad Dreams deals 12.5% per turn, Leftovers heals 6.25%.
Sleep lasts 1-4 turns, Safeguard lasts 5 turns. Wobbuffet loses 6.25%-25% HP while asleep, while he is guaranteed to gain at least 31% back when he Safeguards. As such, unless Wobbuffet is at 25% or less, it can ALWAYS stall out Darkrai's Dark Void. Choiced Darkrai versus Wobbuffet is a really shitty matchup for the Darkrai user even if it doesn't cause auto-loss, since the Wobb user has nothing but options. He can either force Struggle or set up Safeguard and let something scary set up.
Even if Wobbuffet is the second thing slept on the guy's team, it's nothing but good for him.
Just a thought, how would this affect Little Cup with Wynaut? A Wynaut holding a Lum/Chesto Berry and some Speed EVs can easily abuse this auto-forfeit clause.
For example, Player 1 has a Paras out, and no Pokemon are currently asleep. Player 2 has a Wynaut in his party.
Player 2 switches in Wynaut
Paras uses Spore
Wynaut falls alseep
Wynaut's Lum Berry cures it of status
Wynaut uses Encore
Paras receives an Encore
Paras uses Spore
Wynaut falls asleep
Player 2 sends out a Pokemon
Paras uses Spore
Pokemon falls asleep
2 Pokemon have been put asleep by Player 1, so Player 1 auto-forfeits
Player 2 wins!!
The same could happen with any Pokemon in place of Paras that has a sleeping move e.g. Bronzor, Shroomish. People can think of strategies to abuse this new clause that would never be implemented in real link battles. Only a few Speed EVs are needed on Wynaut to outrun Paras, Bronzor and Shroomish, all of whom almost never run Speed EVs. Every time one of these Pokemon uses a Sleep move, they risk an auto-forfeit, and severely limiting their usefulness, whilst this would not be a problem in normal fights.
There might be some other scenarios like this in other tiers, but this was the one that struck my mind.
For example, Player 1 has a Paras out, and no Pokemon are currently asleep. Player 2 has a Wynaut in his party.
Player 2 switches in Wynaut
Paras uses Spore
Wynaut falls alseep
Wynaut's Lum Berry cures it of status
Wynaut uses Encore
Paras receives an Encore
Paras uses Spore
Wynaut falls asleep
Player 2 sends out a Pokemon
Paras uses Spore
Pokemon falls asleep
2 Pokemon have been put asleep by Player 1, so Player 1 auto-forfeits
Player 2 wins!!
The same could happen with any Pokemon in place of Paras that has a sleeping move e.g. Bronzor, Shroomish. People can think of strategies to abuse this new clause that would never be implemented in real link battles. Only a few Speed EVs are needed on Wynaut to outrun Paras, Bronzor and Shroomish, all of whom almost never run Speed EVs. Every time one of these Pokemon uses a Sleep move, they risk an auto-forfeit, and severely limiting their usefulness, whilst this would not be a problem in normal fights.
There might be some other scenarios like this in other tiers, but this was the one that struck my mind.
I would just like to point out that no Wynaut in Little Cup would run a Lum Berry over Oran as it needs Oran to turn many of those 2HKOes into 3HKOes. I also wouldn't fret too much about Paras as it isn't used often, and Bronzor, as it is a genderless pokemon, cannot access the move Hypnosis. Sleep inducing pokemon in Little Cup are far and few, and even them fitting them in a team is pretty difficult. In all honesty I don't think bringing up Little Cup into this debate makes sense, sleep moves aren't used as often, and Wynaut is only used some of the time.
[QUOTE="Seven Deadly Sins]"Sleep-inducing moves used while the user is unable to switch moves and trapped by the opponent do not trigger Sleep Clause."[/QUOTE]
If this is implemented (and it should because it punishes people trying to exploit instead of rewarding them and does not affect normal play), it takes care of the situation you mentioned.
Also what Gen said.
(btw hypnosis is illegal on bronzor so before bringing up lc stuff do a little research kthx)
If this is implemented (and it should because it punishes people trying to exploit instead of rewarding them and does not affect normal play), it takes care of the situation you mentioned.
Also what Gen said.
(btw hypnosis is illegal on bronzor so before bringing up lc stuff do a little research kthx)
I don't see what the big issue is for making an opponent instant-lose because they play in a manner that allows them to be manipulated to the point of an instant-loss. The more "manipulatable" scenario occurs in ubers, a place where broken pokemon and strategies are allowed.
And honestly, these kind of "loopholes" are so situational that a player has to dedicate many team slots and positions in the OU metagame to make it even feasible to manipulate.
And honestly, these kind of "loopholes" are so situational that a player has to dedicate many team slots and positions in the OU metagame to make it even feasible to manipulate.
I agree completely. Pretty much what I've been saying since the last topic.
Whilst I do agree that loopholes regarding this clause are situational, and are usually unorthodox, they promote the use of strategies that would otherwise be completely unfeasible in standard play. Such strategies will never be seen in link battles for obvious reasons, and the purpose of the change in the sleep clause is to make shoddy battles more similar to link battles.
I think a definition along the lines of this might be more appropriate:
If a player has two of his Pokemon put asleep by his opponent, and the opponent was 100% to blame for putting the opponent to sleep through the use of moves, then the opponent automatically loses the game.
Compare it with the "current" definition:
This definition has a little more focus on the opponent. This would be saying that if the player has some role in putting two of his own Pokemon to sleep, an auto-loss would not occur. This prevents people from abusing loopholes and otherwise unorthodox strategies, whilst keeping the battles more similar to link battles. The use of moves is mentioned so that a switch cannot contribute to the sleep. Otherwise, one could argue that a switch at any point in the match would influence the use of Sleep moves, and no one player would be at 100% fault. Plus it makes the clause much easier to implement on shoddy; only moves need to be taken into account instead of switches and abilities.
Considering the moves that have already been debated (these assume that the player uses the move; not the opponent):
Magic Coat - Doesn't break
The opponent uses a Sleep move that was reflected back at him. As the player used a Sleep move, he played a role in putting two of his own Pokemon to Sleep. Whilst the opponent may or may not have known that he could have potentially put the opponent to sleep, the sleep involved both the use of Magic Coat and the use of a Sleep move.
Metronome - Breaks it
The player uses this move, and the opponent does nothing to influence this action. Therefore, the player is 100% responsible for this move and loses.
Assist - Breaks it
Same as metronome
Sleep Talk - Breaks it
Same as metronome
Encored sleep - Doesn't break it
The opponent used Encore, so it's not 100% the player's fault, even if he could swap out. Whilst many of you will disagree with this, I believe this should be the case, because by using Encore, the opponent is intentionally letting the player put one of his Pokemon to sleep. Even if the player switches out, he used a move that let himself have his Pokemon put to sleep. Also, in a real trade link match, an opponent may not switch their Pokemon out against a player stuck using Encored sleep, for the simple reason that he risks having many of their Pokemon then risk being put to sleep. They would probably keep their Pokemon in, knowing they would wake up before the Encore ends.
Choiced Sleep - Breaks it
The opponent did not use a move to influence this sleep; therefore it is not his fault.
Last pp sleep move -Breaks it
Same as above; plus the player would probably end up losing anyways.
Effect Spore - Doesn't break it
The opponent has to use a move in order for Effect Spore to put the opponent to Sleep. It's not the player's fault if the opponent continuously uses attacks like Rapid Spin for the sole purpose of trying to put himself to sleep against a player's last Pokemon (Breloom or Parasect).
Whilst I largely agree with Jumpman's post, it would be very hard to implement it into shoddy. My definition is somewhere along the line's of his post, as by having a Pokemon using a move that could put himself to sleep e.g. reflected Sleep Powder, the opponent cannot have an auto-loss for having his opponent intentionally putting himself to sleep.
I think a definition along the lines of this might be more appropriate:
If a player has two of his Pokemon put asleep by his opponent, and the opponent was 100% to blame for putting the opponent to sleep through the use of moves, then the opponent automatically loses the game.
Compare it with the "current" definition:
This definition has a little more focus on the opponent. This would be saying that if the player has some role in putting two of his own Pokemon to sleep, an auto-loss would not occur. This prevents people from abusing loopholes and otherwise unorthodox strategies, whilst keeping the battles more similar to link battles. The use of moves is mentioned so that a switch cannot contribute to the sleep. Otherwise, one could argue that a switch at any point in the match would influence the use of Sleep moves, and no one player would be at 100% fault. Plus it makes the clause much easier to implement on shoddy; only moves need to be taken into account instead of switches and abilities.
Considering the moves that have already been debated (these assume that the player uses the move; not the opponent):
Magic Coat - Doesn't break
The opponent uses a Sleep move that was reflected back at him. As the player used a Sleep move, he played a role in putting two of his own Pokemon to Sleep. Whilst the opponent may or may not have known that he could have potentially put the opponent to sleep, the sleep involved both the use of Magic Coat and the use of a Sleep move.
Metronome - Breaks it
The player uses this move, and the opponent does nothing to influence this action. Therefore, the player is 100% responsible for this move and loses.
Assist - Breaks it
Same as metronome
Sleep Talk - Breaks it
Same as metronome
Encored sleep - Doesn't break it
The opponent used Encore, so it's not 100% the player's fault, even if he could swap out. Whilst many of you will disagree with this, I believe this should be the case, because by using Encore, the opponent is intentionally letting the player put one of his Pokemon to sleep. Even if the player switches out, he used a move that let himself have his Pokemon put to sleep. Also, in a real trade link match, an opponent may not switch their Pokemon out against a player stuck using Encored sleep, for the simple reason that he risks having many of their Pokemon then risk being put to sleep. They would probably keep their Pokemon in, knowing they would wake up before the Encore ends.
Choiced Sleep - Breaks it
The opponent did not use a move to influence this sleep; therefore it is not his fault.
Last pp sleep move -Breaks it
Same as above; plus the player would probably end up losing anyways.
Effect Spore - Doesn't break it
The opponent has to use a move in order for Effect Spore to put the opponent to Sleep. It's not the player's fault if the opponent continuously uses attacks like Rapid Spin for the sole purpose of trying to put himself to sleep against a player's last Pokemon (Breloom or Parasect).
Whilst I largely agree with Jumpman's post, it would be very hard to implement it into shoddy. My definition is somewhere along the line's of his post, as by having a Pokemon using a move that could put himself to sleep e.g. reflected Sleep Powder, the opponent cannot have an auto-loss for having his opponent intentionally putting himself to sleep.
The only reason they aren't used in link battles is because the abuse of the current sleep clause on wifi hasn't really hit the majority of players, since they're so used to the "safe zone" that Shoddy Battle 1 affords them. They may know the "Blissey has Serene Grace" dilemma, but the idea of using Encore and a Trapping pokemon to force an instant loss doesn't usually occur. Even when it's successfully pulled off, the opponent who it happened to usually would pack a fit and just DC rather than take the (single) loss. Or so would your "everyday" wifier, anyway.
I say let the sleep clause through as it is and let it play itself out. While at first some people may try to abuse it, they'll find themselves often just limiting their overall options rather than try to play normally. Eventually people will realize that it doesn't work and the people who try to abuse it will drop to maybe a 0.01% of players, which only about a small percentage of them will even be able to successfully pull it off consistently.
Worse come to worse, the use of Choiced sleep moves drop drastically. Oh dear...
I don't see why it will never be seen in link battles, so it's not obvious. Please explain why no one will ever use this on a cartridge?
There is never a situation where the opponent is 100% to blame. "He could have switched to his sleeping Pokemon!" or, if trapped "He could have not switched out his original sleeping Pokemon to begin with!". Trying to assign blame is silly, and impossible to do.
I figured I would post the wording of how I think Sleep Clause should be implemented:
Unless you had no other possible move on the turn you used the move that would otherwise cause you to lose, you lose if you put an enemy pokemon asleep with a move that specifically induces sleep, and an enemy pokemon you previously put asleep with a move that specifically induces sleep is still sleeping.
'No other possible move' doesn't include forfeiting obviously (my Computer Science professor says forfeiting is always a valid move and counts as a move so I figured I should mention it, it probably isn't worth mentioning in the definition though).
The idea behind this is that there is never a situation where a player is forced to lose by his opponent; s/he must always be able to choose whether s/he loses or not. For me, this takes preference to the chance that there may be a legal situation where two Pokemon are put to sleep by the opponent. There is an issue with this where the player would attempt to set up a situation where he could legally sleep two Pokemon, but that's what I mean by taking preference - I think the capability of 'forcing' losses through anything other than KOing 6 Pokemon shouldn't be a result of any rule we use, and that allowing two Pokemon to fall asleep under the clause is more fair than being forced to lose.
Unless you had no other possible move on the turn you used the move that would otherwise cause you to lose, you lose if you put an enemy pokemon asleep with a move that specifically induces sleep, and an enemy pokemon you previously put asleep with a move that specifically induces sleep is still sleeping.
'No other possible move' doesn't include forfeiting obviously (my Computer Science professor says forfeiting is always a valid move and counts as a move so I figured I should mention it, it probably isn't worth mentioning in the definition though).
The idea behind this is that there is never a situation where a player is forced to lose by his opponent; s/he must always be able to choose whether s/he loses or not. For me, this takes preference to the chance that there may be a legal situation where two Pokemon are put to sleep by the opponent. There is an issue with this where the player would attempt to set up a situation where he could legally sleep two Pokemon, but that's what I mean by taking preference - I think the capability of 'forcing' losses through anything other than KOing 6 Pokemon shouldn't be a result of any rule we use, and that allowing two Pokemon to fall asleep under the clause is more fair than being forced to lose.
When referring to blame, I only mean blame "through the use of moves." That is to say, it only refers to when accounting for the moves made by the Pokemon; I probably should have made it clearer in my post that switching (and forfeiting) are not a part of this.
Most people who play by link cartridges use the standard in-game rules. That is to say, they don't care if they put two opposing Pokemon to sleep, if they use Double Team etc. If we disqualify people for people who abuse our clause on shoddy, people will try to avoid breaking this clause, and may change their Pokemon teams and strategies accordingly. As well, new strategies may become viable as a result of this clause i.e. loopholes in the clause, in which people abuse the clause to force the opponent into disqualification. Most people will adjust their teams so that the clause cannot be used to exploit the team and force the player into disqualification, but there will always be some teams that could be exploited, probably noob teams. Strategies involving exploiting the clause wouldn't work in link battles, because there is no clause to exploit in the battle.
If you are talking about a link battle where the players do agree clause is in effect, you could reasonably assume that players who want to play with proper rules (i.e. the clauses) are serious players. They aim to have the highest chance as possible of winning against one random team (in a similar way to how teams are designed for tournament use). Trying to use strategies that abuse the clause won't work against the majority of teams, as very few teams would have a weakness to the clause that could be exploited (seeing how teams would be designed so the clause can't be exploited against them). Hence players would use a more reliable team strategy that doesn't involve exploiting the clause, and strategies involving exploiting the clause would not be used in link battles.
The only case in which I could think in the clause would be exploited in a link battle with the clause is when a players knows the opposing team beforehand, and if the player knows the team can be exploited. You could then argue that the match wasn't competitive, as he knows the opposing team beforehand.
I agree with your logic, and that this clause should be as fair as possible i.e. you shouldn't lose if your opponent forced you to put two Pokemon to sleep. However, the word "specifically" concerns me, and it could be argued that moves like Sleep Talk and Metronome are allowed, as inducing Sleep through such moves is not intentional and is not the specific use of the move, but rather the move could have many potential uses; the move is not specific in nature. And it wouldn't be fair if a player has two of his Pokemon put to sleep by Sleep Talk, and there was nothing he could to do avoid this, so he would be playing the rest of the game with two Pokemon out of commission.
..I don't know where you're basing your assumption on, but you're terribly misinformed. People "enforce" sleep clause in link battles all the time, forcing the person who breaks it to forfeit, usually. You make it sound like people that don't use shoddy are a bunch of random lawless players that spam Mewtwo and Rayquaza in standard battles, lol. Anyways, the fact that WiFiers also use sleep clause makes the universal applicability of this new clause a priority, albeit a low one.
This probably depends where your from. America and Japan would have many less lawless players than anywhere rest in the world. Most Pokemon players in Australia, where I'm from, are actually random lawless players (if they decide to play with rules, they usually just ban the Pokemon that Battle Tower bans), and I'd figure this would be the case in most countries, but not in America and Japan, where most professional Pokemon battlers are from.
Either way, I already discussed what I would believe would happen if the clause would be upheld in link battles.
Smogon has no control over the rules of any battling community other than its own, so I'm not really sure why you brought that up. This discussion is for the sole purpose of developing the parameters of competitive play within Smogon, which, last time I checked, includes the WiFi forum as well. Also, as far as I know, people battling through Smogon's WiFi forum follow standard clauses, unless otherwise agreed upon. That being said, I really feel like we've strayed away from productive discussion, so I will leave it at that.
Sorry I was tired and misinterpreted something, I only noticed just now. Ignore the previous proposed definition I gave, I meant to change Colin's, not the current one. Here is what I meant:
Unless you had no other possible move on the turn you used the move that would otherwise cause you to lose, at the end of every turn, if more than one pokemon is asleep on your opponent's side, and if at least two of those pokemon were put asleep by your side's moves, excluding Magic Coat, but including Yawn, then you automatically lose. If both sides qualify, the game ends as a draw.
Sorry about the confusion!
edit: responding to below post
I meant to include switching as a 'possible move', so this defintions already encompasses that.
Unless you had no other possible move on the turn you used the move that would otherwise cause you to lose, at the end of every turn, if more than one pokemon is asleep on your opponent's side, and if at least two of those pokemon were put asleep by your side's moves, excluding Magic Coat, but including Yawn, then you automatically lose. If both sides qualify, the game ends as a draw.
Sorry about the confusion!
edit: responding to below post
I meant to include switching as a 'possible move', so this defintions already encompasses that.
This definition isn't bad. My main concern about it is Encore. It doesn't prevent people from losing due to Encore, or more specifically, Mean Look + Encore, trapping a Pokemon and forcing it to repeat it's sleep move. The same applies to Wobbuffet. I'm alright if a person loses if one of his sleep moves is Encored, but not when he's trapped at the same time.
I'm also unsure if there should be a draw if both sides qualify for breaking the clause instead of just continuing the match. Would it be fair if a draw was declared when one player was winning 6-2?
I've given this some thought, and I think the following should be the basic sleep clause, for singles cartridge play.
Metronome: Does not break. Using Metronome is assumed to be intending nothing.
Magic coat: Breaks if and only if you know for certain your opponent is move-locked. A Pokemon reaching a speed or damage output only obtainable by a choice item counts as knowing.
Assist: Breaks if and only if there is more than some percentage chance of it selecting a sleep move. The burden of proof lies on the Assist user.
Sleep Talk selecting a sleep move or Psycho Shift: Breaks; it is assumed that by using that moveset, you intended to put things to sleep.
Own choiced sleep move, generally: Breaks. You could have switched
Own choiced sleep move, trapped or last Pokemon. Does not break.
Encored sleep move, generally: Breaks. You could have switched.
Encored sleep move, trapped or last Pokemon: Does not break. You had no option.
Last PP sleep move, generally: Breaks. You could have switched.
Last PP sleep move, trapped or last Pokemon: Does not break. You had no option.
Yawn: Breaks. It's a sleep move.
I think the distinguishing things here are:
Metronome being legal. Frankly, anyone gutsy enough to use Metronome deserves what they get, good or bad).
Assist being sometimes legal. Note anyone using Assist when they have another Pokemon with a sleep move is going to risk having to reveal their team, to prove they didn't load up on sleep moves to try and slip by the clause.
Magic Coat being sometimes legal, sometimes not.
Then we can have some clarifications, based on this as the guiding principle. Note not all of these are finalised. You are welcome to argue that my own principle should lead to different conclusions for them.
Metronome: Does not break. Using Metronome is assumed to be intending nothing.
Magic coat: Breaks if and only if you know for certain your opponent is move-locked. A Pokemon reaching a speed or damage output only obtainable by a choice item counts as knowing.
Assist: Breaks if and only if there is more than some percentage chance of it selecting a sleep move. The burden of proof lies on the Assist user.
Sleep Talk selecting a sleep move or Psycho Shift: Breaks; it is assumed that by using that moveset, you intended to put things to sleep.
Own choiced sleep move, generally: Breaks. You could have switched
Own choiced sleep move, trapped or last Pokemon. Does not break.
Encored sleep move, generally: Breaks. You could have switched.
Encored sleep move, trapped or last Pokemon: Does not break. You had no option.
Last PP sleep move, generally: Breaks. You could have switched.
Last PP sleep move, trapped or last Pokemon: Does not break. You had no option.
Yawn: Breaks. It's a sleep move.
I think the distinguishing things here are:
Metronome being legal. Frankly, anyone gutsy enough to use Metronome deserves what they get, good or bad).
Assist being sometimes legal. Note anyone using Assist when they have another Pokemon with a sleep move is going to risk having to reveal their team, to prove they didn't load up on sleep moves to try and slip by the clause.
Magic Coat being sometimes legal, sometimes not.