Alex has a combo on the board which basically amounts to being able to activate an ability that says “Target player puts the top 2 cards of his or her library into his or her graveyard” as many times as he wishes. Nathan has 51 cards in his library. Upon using this ability to put Nathan's entire library into his graveyard, Alex discovers that Nathan has one copy of Emrakul, the Aeons Torn in his deck. Since Emrakul shuffles itself and the rest of the graveyard into the library, and Alex can mill 2 cards an arbitrarily large number of times, Alex wishes to shortcut to the eventual iteration where after shuffling, Emrakul is on the bottom of Nathan's library, and Alex mills the top 50 cards without hitting it. Nathan argues this shortcut, and a judge is called. What is the ruling here?

“It is also slow play if a player continues to execute a loop without being able to provide an exact number of iterations and the expected resulting game state.”

Hi Michael,

Matteo Callagari and myself co-wrote an article on Slow Play not too long
ago that addresses this same exact scenario.

You can read the section about loops at the following link.

http://blogs.magicjudges.org/articles/2012/12/25/slow-play/#loops

And for anyone else, I think the rest of the article is pretty awesome as
well and you should give it a read. ;)

-George FitzGerald
L2, Sarasota, FL

Toby, not surprising, is spot on here. This ruling came up a number of months ago at an SCG Legacy tournament. I believe there was a deck called the 4 horseman that used Basalt Monolith and Mesmeric Orb to fill his gy and reanimate a fatty with Dread Return. The deck used a single Emrakul to loop.

Speaking in a purely mathematical sense, we can know the desired iteration will happen eventually; however, since a specific number cannot be given to provide that iteration, it does not qualify to be an infinite loop shortcut.
MTR 4.2 – Tournament Shortcuts

The player piloting the deck ended up with a game loss in round 3 for a repeat slow play infraction. You can see Toby's comment about it here.
http://starcitygames.com/events/coverage/deck_tech_four_horsemen_with_j.html

I'll also point out, this question actually forms somewhat of an non-deterministic loop - as in, I can't provide an exact number of times this question (with various cards involved) has been raised.

That historical fact, however, is news to a lot of judges reading these forums; I shouldn't expect everyone to know that “we already answered that! back in 2005!!” - in fact, there's probably some judges who were active back then who have already forgotten that iteration of this loopy question. :)

But, the short version is “you can't do that”. The policy quote that Toby provides was a result of people arguing that they should be able to do that, and our effort at codifying why it's bad to let you try.

If your opponent accepts your proposed shortcut, there's no problem. If they're competitive and at all informed re: history and policy, they'll say “no”, and then we will, too.

Hopefully, that all makes sense?

So these kinds of combos cannot be played without incurring repeated Slow Play penalties, resulting in GL'ing them from the tournament?

No, they can be played, and opponents can scoop to this. However, if the opponent doesn't scoop, they have to stop the loop, play a land, and hope to hardcast that Emrakul to win. Smart players may include other ways to win / gain an advantage over their opponent to prevent them from winning until they can create a more advantageous loop.

In Legacy this ruling is relevant, since graveyard order matters in that format. In Modern/Standard where graveyard order does not matter, is it legal to say “The game state is exactly where it is now, except there is no Emrakul in your graveyard and your graveyard is in <insert arbitrary deterministically constant sorted order here> (for example ”sorted alphabetically“, ”sorted by card type“, ”sorted by print date of the card“, ”sorted by the median value of the non-whitespace characters in the artist's name, using ASCII representation“)” and have that be the game state? If not, why not?

Incidentally, I would really like the DCI to ban graveyard-order matters cards in Eternal formats, with a rationale akin to the one used for manual dexterity and subgame cards (i.e. “this mechanic is stupid and we never should have made it, so we're un-making it now”)

Edited Lyle Waldman (July 9, 2013 08:03:25 AM)

How many iterations will it take to get to that state?

Paul Smith

paul@pollyandpaul.co.uk

Because there is no finite number of iterations guaranteed to produce the desired result. We only shortcut finite numbers of operations, which pretty much means no probabilistic outcomes.

Originally posted by Joshua Feingold:

Because there is no finite number of iterations guaranteed to produce the desired result. We only shortcut finite numbers of operations, which pretty much means no probabilistic outcomes.

What's the philosophy behind this? We all know that it will happen eventually, and we can deterministically define such a game state where it happens, so what's at issue here?

I'll post the link again then quote it for those who can't be bothered to click it.

http://blogs.magicjudges.org/articles/2012/12/25/slow-play/#loops

“A common argument that is made is that the certainty of the outcome of the « loop » can be proved thanks to a mathematical demonstration. This is sometimes the case. However, the R&D clearly stated that the outcome of a game of Magic should be decided by cards being played and not by a conjecture that requires a degree in Mathematics.

If you were a player who has never studied maths, would you be happy to hear from the judge that you lost, even though your opponent cannot demonstrate it with certainty but only with a couple mathematical formulas ?”

The rationale is that (most) people come to tournaments to play Magic, not to play a game where they have to engage in mathematical proofs to figure out whether their opponent can kill them.

Originally posted by Lyle Waldman:

We all know

I suspect that's the issue. I know. You know. But not everyone knows. An understanding of the maths behind probability and statistics shouldn't be a requirement to play Magic.

Plus, it's nice to have a neatly defined line - it's black and white. If we allow grey, then the next question is “how much grey?” How do we then decide how close to a probability of 1 is enough? Taking this to a logical (yet obviously silly) conclusion: consider the Modern Eggs deck before Second Sunrise was banned - it occasionally fizzled out without any interference from the opponent. What one player started demonstrating the loop and said: “I'm going to do the eggs loop and I'm 75% likely to kill you this turn, so shall we just go to game 2?”

Magic's loop rules require both a set number of iterations and a set
outcome at the end of those iterations.
Also, because there's randomness involved here, there's a chance, however
vanishingly small, that Emrakul never ends up on the bottom of the library.
On Jul 9, 2013 10:48 AM, “Lyle Waldman” <