Why is there no ruling to offset Mandatory Loops?

As pointed out in a recent question: Marauding Raptor and Polyraptor - what happens when a loop has no exit? Which states this ruling

721.4. If a loop contains only mandatory actions, the game is a draw. (See rules 104.4b and 104.4f.)

Now I understand within the confines of the game that certain cards have been created and balanced with mandatory actions that once cast can not be deviated from. This is part of card design- which is a whole different topic, but sometimes these have the unintended consequence of looping forever.

Non-optional unintended loops are presumably outside of the preferred intended play by the designers, and it leads me to wonder if any effort has been done to investigate outs for when these situations arise (eg. after a mandatory-loop, loops x100 then remove any further triggered/replacement effects.)

Allowing a hard break to certain loops might allow some niche decks to arise, perhaps this might lead to a negative impact on certain formats, which might be why this decision was done?

So ideally I am looking for any sources which point to why they have decided to leave this rule as such, and for so long. Has any past attempts at reanalyzing this rule come up?

Disclaimers: Mainly looking for reasons within sanctions tournament rules, or other official events, but creditable sources outside that may be acceptable answers.

  • 5
    Don't have any official sources, but I'd suspect with game like Magic that has so many possible interactions (both currently and with future cards) it's probably safer to just have a safety net like this. Say those loops exited after 100 iterations, where in that loop do you exit? And how do you determine game state (an already ambiguous term) after that? For a two card combo like the one cited in your question, that might not be too hard. But what about 3, 4, 5 or 10+ card combos? Also, why is going to a draw always a bad thing? If you are losing and can force a draw, is that really bad?
    – Becuzz
    Commented Mar 17, 2020 at 16:26
  • 1
    I suspect the real answer is rather uninteresting - this is a niche problem which would require a very inelegant and technical solution (see also murgatroid99's answer why it's not so easy as "stop after X iterations"). Just defining these games to be a draw is the easiest and cleanest solution that always works. Also, Magic design has started to favour optional triggers over mandatory ones in recent years, so the problem is bound to become even more niche.
    – xLeitix
    Commented Mar 18, 2020 at 15:27
  • For optional triggers the solution is simple - if choices you make lead to an endless loop, you can decide how often you want it to happen, but you do need to make a decision that breaks the loop eventually.
    – xLeitix
    Commented Mar 18, 2020 at 15:29

2 Answers 2


There is no reasonable way to eliminate all mandatory loops, so this rule would still be needed for loops that can't be eliminated. For simplicity, it's better to just have the one rule.

The biggest problem with trying to stop mandatory loops is listing what exactly stops happening, and putting the game in a valid state afterward. The simplest loops are the kind you describe, where one ability triggers, and the resolution of that ability causes another ability to trigger, and so on until the first ability triggers again and you end up in a loop. But there are other kinds of mandatory loops that are a lot harder to cut off.

One of the strangest mandatory loops in the game involves the card Lich's Mirror. If a player controls a Lich's Mirror that they do not own, and gets 10 poison counters, the Mirror's ability replaces the state-based action that would make the player lose, but the Mirror itself stays on the battlefield, so when state-based actions are re-evaluated, exactly the same thing happens. Stopping this loop would require either suspending the rule that causes state-based actions to be rechecked after they do something, or making the Lich's Mirror's replacement effect no longer function. In other words, that player either loses immediately, or becomes nearly unkillable, and both changes have other unpleasant implications. Ending the game in a draw is a reasonable compromise.

There are also several cards with "state-triggered abilities", which trigger on a particular game state, instead of an event. Some combination of those cards allow those abilities to trigger and resolve without changing the game state, so they immediately trigger again. There's no good way to change the game state so that those abilities stop triggering.

In addition to that, there are more technical issues with defining such a rule. In particular, it's not clear how you would define an "iteration" of the loop, or how you would determine what part of an iteration to end it on. For player-defined looping shortcuts, the player makes those decisions, but a mandatory loop can involve both players equally, leading to difficult-to-resolve disagreements about how exactly such a loop should end.

  • 1
    Essentially you would need to create a directed graph of each interaction to show what a single loop is.
    – Max Young
    Commented Mar 18, 2020 at 3:35
  • "Stopping this loop would require either suspending the rule that causes state-based actions to be rechecked after they do something" There's a card that does that in silver-border land: Rules Lawyer.
    – nick012000
    Commented Mar 18, 2020 at 6:30
  • Rules Lawyer suspends checking state-based actions entirely. That would probably be too extreme a solution to trying to control one card looping. I was talking about the rule (the second sentence of 704.3) that says that after state-based actions are checked, if any of them did something, they get checked again (and again, and again, until none of them take any actions).
    – murgatroid99
    Commented Mar 18, 2020 at 6:33
  • But in this case, none of the state based actions actually did anything at all. So, no loop.
    – Joshua
    Commented Mar 20, 2020 at 1:50
  • Actually, in the Lich's Mirror loop, at minimum, every iteration you have to shuffle your library. Sure, the game state may not have meaningfully changed, but actions were still performed and that's what matters here.
    – murgatroid99
    Commented Mar 20, 2020 at 2:29

The rule is necessary. Limiting purely mandatory loops to a certain number of repetitions is impractical, especially for paper Magic, and it is potentially game-distorting.

Magic the Gathering is Turing-complete. This means a few things. Primarily, it means you can simulate a Turing Machine with only tournament-legal cards and the official rules. It also means that the game is subject to the Halting Problem, which means that there exist legal game states when it's impossible to predict, even in principle, whether or not a loop in the game will run forever.

That means the game needs a hard and (literally) fast rule to decide when to end the game permaturely. In paper Magic, without digital computation, it's impractical to make players play out a loop a number of times (e.g. 100) and then decide whether the loop will go on. In digitally enforced Magic such as Arena, it's even harder, because algorithms can't handle infinities or handwaving very well. And how can you possibly determine the optimal number of repetitions? Is the optimal number 10, 100, 1000, or a billion? At which point can you be sure that the loop will go on forever? You can't.

Note that the term "game state", which is for example used in deciding whether or not a loop is fragmented and whether players may continue performing it, isn't even defined anywhere! On the one hand, for a game in which every single minute detail is nailed down in the rules, sometimes repeatedly, it is quite shocking to have such a fundamental term undefined. On the other hand, it shows intuitively that the game is as complex as the scientific paper cited above has proven. In some extreme circumstances, the game just has to stamp down and call it quits. Because of the existence of paper Magic, that point has to come as soon as possible.

That's why 721.4 is necessary.

  • 14
    I think the problem with citing Turing completeness here is that if you can't decide whether a loop will run forever or not, then you also can't decide whether rule 721.4 applies. Also, I think the point about the impracticality of playing out loops is a bit of a red herring: For almost all loops, it would be easy to create a shortcut for however many iterations.
    – murgatroid99
    Commented Mar 17, 2020 at 17:25
  • This is actually a misnomer as while it IS a sequence of mostly mandatory actions. It's not actually a loop (Repeating the same-same choices and actions over and over), it's much more complex and falls outside what is defined as actual Loops by the Tournament Rules committee.
    – L.P.
    Commented Mar 19, 2020 at 14:12
  • Also, is the game still even considered Turing-complete if in a format that doesn't have the access to the required cards (Such as standard or a cube).
    – L.P.
    Commented Mar 19, 2020 at 14:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .