One of Magic the Gathering's rules is

104.3f If a player would both win and lose the game simultaneously, he or she loses the game.

However, it seems to me like it "ought to be" the case that if there are only 2 players and [A loses] ​ [B loses] ​ [B wins] would occur simultaneously, then B should win, since win-and-lose-simultaneously should be better than just-lose.

For that reason, it seems like this would make more sense:

Proposed Alternative If, simultaneously, all players would lose the game and some players would win the game, then the players who would not win the game lose the game, and the game is a draw between the remaining players. ​ Otherwise, if a player would both win and lose the game simultaneously, he or she loses the game.

However, I can't think of any situations in which that would matter.

Are there any possible situations in which my proposed alternative would yield a different result from the actual rule?

  • It seems to me that you described in your question a situation where it behaves differently: one player loses, and another player wins and loses, all simultaneously.
    – murgatroid99
    Commented Oct 21, 2015 at 21:33
  • I believe it is not possible to win and lose simultaneously. There is a long discussion about this here, which ended in a forum moderator locking the thread because so many people thought they came up with a situation where the rule applied but were wrong for one reason or another.
    – Rainbolt
    Commented Oct 21, 2015 at 21:54
  • The thread lock message says "Within legitimate games, you can't force this situation." That doesn't mean that it's actually impossible, depending on your definition of "legitimate"
    – murgatroid99
    Commented Oct 21, 2015 at 21:57
  • 2
    That thread was closed 5 years ago. They've printed a lot of cards since then (most notably Laboratory Maniac, which allows you to win the game during the resolution of another spell or ability)
    – murgatroid99
    Commented Oct 21, 2015 at 22:01
  • 4
    The Comprehensive Rules strive to be comprehensive; this includes hedging a bit against potential corner-cases that will only occur because of some future design or as the result of an interaction of quirky circumstances (like specialized game mode or tournament rules). The appeal of 104.3f as written is that it's written very simply, to be a quick fall-back for truly weird stuff that doesn't require lots of brainpower to read and understand. It's not strange for the game to emphasize "losing" over "winning" because winning in MTG is usually just the state of "everyone else has already lost."
    – Alex P
    Commented Oct 21, 2015 at 23:17

1 Answer 1


Your hypothetical rule change would never matter, because there is currently no way for a player to win and lose the game simultaneously and therefore it is impossible for 104.3f to apply.

Generally, the steps of an effect in Magic are followed in the order they are written, often APNAP order. An effect stating "you lose the game and you win the game" would cause you to lose the game strictly first, before the "you win the game" effect could apply.

There are two exceptions to this generalization: damage events (510.2) and state-based actions (704.3). Replacement effects can replace only parts of these events, so theoretically by filtering through multiple replacement effects we could change one part of the event to "you lose the game" and another part to "you win the game".

Unfortunately, the only way to win the game with replacement effects is through Laboratory Maniac or Jace, Wielder of Mysteries. Both of these require us to draw a card via a replacement effect, and a very specific rule gets in our way:

121.7. Some replacement effects and prevention effects result in one or more card draws. In such a case, if there are any parts of the original event that haven’t been replaced, those parts occur first, then the card draws happen one at a time.

Consider the following situation. A player who has an empty library and graveyard controls:

Their opponent attacks with a 2/2 creature with trample, and they block with their 1/1. The damage event starts out as

(1 damage dealt to the 1/1 AND SIMULTANEOUSLY 1 damage is dealt to the 2/2 AND SIMULTANEOUSLY 1 damage is dealt to the player AND SIMULTANEOUSLY the player gains 1 life)

The former two are irrelevant for our purposes. Nefarious Lich's replacement effects change the latter two events to

(the player exiles 1 card from their graveyard, if they can't they lose the game AND SIMULTANEOUSLY the player draws 1 card)

If you didn't know about 121.7, you might assume that Laboratory Maniac replaces the draw here resulting in a simultaneous win and loss. But what actually happens is that 121.7 notices a replacement effect resulting in a card draw and results in

(the player loses the game.) then,
(the player draws 1 card.)

So the player simply loses.

If you're particularly astute, you might seize upon 121.7's wording "the original event" and think that it would not apply if we completely replaced every part of the starting effect. However, I have personally consulted with a judge I know. He says that the consensus in the judging community is that "original" in this context refers not to the effect that was originally generated before replacement effects started to apply, but to the effect after the replacement effect that results in card draws. In other words, it does not mean "if every part of the effect that was generated has been overwritten by replacement effects", but instead means "if any part of the effect that has now been partially replaced with card draws consists of something other than card draws".

You must log in to answer this question.