Since this was recently brought up in the Clauses in Generation V thread, and has significant potential to bring that thread off topic, I thought it was best to move to a separate thread. Important background reading is this thread. Making a good Sleep Clause is not as easy as it may seem. I think that most players can agree that an ideal Sleep Clause would have the following attributes (though feel free to suggest changes): 1. Prevent sleep from becoming an overbearing strategy. If we implement a Sleep Clause the first thought is to prevent sleep based strategies being overpowered. We do not need to force sleep to be unusable, or necessarily to prevent multiple Pokemon from being put to sleep. 2. Be consistent with ingame mechanics. If there is no "hard" Sleep Clause (one which simply causes sleep moves to fail after one Pokemon is asleep), then we should not interfere with mechanics and cause the move to fail when ingame it would not. Likewise, we cannot force sleep moves to miss or extra effects to not activate. 3. Be fairly simple both for the user and in programing terms, and have a clear answer to every situation. An overly complicated Sleep Clause will be harder to implement, more problematic for users to work with, and in all likelihood more arbitrary and prone to interpretation. 4. To not interfere with the game more than necessary. The aim of a game of Pokemon is to defeat your opponent's six (or fewer in some metagames) Pokemon. I think it would be counterproductive, distracting, confusing, and harmful to introduce a second forcible win condition. If you can force your opponent to break the Sleep Clause, even only in fairly unusual situations, and win without KOing your opponent's Pokemon I think there is a flaw in the ruleset. Additionally, the Sleep Clause should not remove any more non-broken strategies than required to have a consistent, simple, all encompassing rule. 5. To avoid adding an extra element of chance. I think that having a clause which is "lose if your Sleep Powder hits, don't otherwise", or "lose if your opponent runs Serene Grace Blissey, don't otherwise" adds a very artificial and arbitrary layer of strategy to the metagame. Ideally, the clause would deal with situations like this fairly without causing coin toss wins based not on the actual mechanics, but on a clause we make. And to give you a few examples to consider (one Pokemon has already been put to sleep on Player B's side in each case): Player A's Pokemon uses Magic Coat and reflects Sleep back. What if Player A knew that the foe had no other option but to use a sleep move? What if Player A did not know the foe had a sleep move? What if Player A were choice locked/encored/out of PP on all other moves and trapped? What if the foe had a Lum Berry/other way to circumvent sleep (e.g. Shed Skin activated)? What if Player A knew the foe had a Lum Berry (it could have been tricked on, or found by frisk) and needed to have it deactivated for another strategy (burn it with another Pokemon)? Magic Coat's, and by extension Magic Mirror's, case is interesting in that it requires both players to make a specific move in order to cause a double sleep (note: I do not say break the Sleep Clause, since that is up to us). It's not a simple chance like many possible violations. Player A's Pokemon uses Metronome/Assist/Sleep Talk/other move which may call a sleep move indirectly (non forced). These can be seen two ways. Either you could say that since it has a chance of putting the opponent to sleep, it should be disallowed for consistency despite none of the strategies that could be based upon them being remotely broken. Or you could say that the move selected does not actually cause the foe to go to sleep, that it calls a separate attack, so that with the correct wording there is no need to call these violations of the Sleep Clause. Player A's Pokemon uses Yawn (non forced) and Player B does not switch out (non forced). Again, there are two possible viewpoints for the scenario. First, you could argue that since Player B had the option to switch out it is not Player A's "fault", and they should not be punished. Player B could have avoided it. However, Player A did use the move which put a second of Player B's Pokemon to sleep, so it is also reasonable to argue that since they could have avoided it the rule should find in Player B's favor. Player A's Pokemon's Effect Spore activates and puts the foe to sleep. This is an interesting case, a small chance to put something to sleep each turn you have the Pokemon in (assuming they hit you with a contact move). I do not think anyone could argue that Effect Spore sleep is broken, however Player A could have avoided the situation by switching out. The Effect Spore Pokemon would still be back in play at the end, as the last Pokemon, but at this point they have no choice but to keep it in, so having the sleep clause activate would be, technically, a forcible win which I believe is best avoided. There are several other tricky situations, especially in doubles or triples, but I believe there is a version of the clause which will cover all the niche situations in one, prevent sleep from being an overbearing strategy, be simple enough, and give a clear answer while not allowing any forcible wins.A player may not select an attack which has a chance to directly cause more than one opposing Pokemon to have been put to sleep directly by a move used by their team unless no other move is selectable.Directly is defined as having a % chance to cause sleep in the move data. Calling another move is not considered to be directly causing sleep, nor is anything related to abilities or reflection of sleep moves. On a simulator a player attempting to use a move which violates the Sleep Clause would receive a warning, and they would lose if they chose to go ahead and make the move anyway. On wifi, they would lose the game for selecting a move which may directly cause a second direct sleep. This eliminates all forcible wins, is fairly simple to understand, prevents the kind of likely broken strategies we aim to avoid, gives a clear answer in every situation (please try and prove me wrong on this one, lets make it watertight), and even eliminates the extra luck/unknown information dependent wins as far as I can see.