I have Melira, Sylvok Outcast, Viscera Seer, and some third creature with Persist in play (for the sake of this discussion, let's not use Redcap). I demonstrate that I can loop the ability of Viscera Seer any number of times at will, and state that I want to repeat it 100 times, guaranteeing that I have seen the entire ordered contents of my deck, and then some.

To what extent for the interest of time/shortcutting am I allowed to simply pick up my entire deck and simply “cut” so that a particular card or sequence thereof is on top of my deck? (Obviously I'm not allowed to reorder the cards, because it's technically Scry 1, but cutting should maintain the proper order for this purpose).

Just curious where, if ever, we draw the line on this kind of technical precision and whether we'd be requiring the player to remember the card sequences mentally instead, since the ability doesn't technically entitle me to see more than one at a time.

This question comes up from time to time. The answer is as follows:

Can you tell me EXACTLY how many times you will repeat the loop? (And by EXACTLY, I mean without having to resort to high-level math on sort orders, logic, etc etc.)

If the answer is no, which it is in this case, you can't do it.

You MUST be able to specify a number of times you will repeat a “boundless” loop and you MUST be able to specify the exact ending game state. If you can't do this, you must manually execute the loop and do so in a timely manner.

For more discussion on loops, see Toby's blog here:
http://blogs.magicjudges.org/telliott/2012/11/02/horsemyths/

I think that I would disagree with David.

Scry 1 is a lot different to “Shuffle and hope that I get a correct order of my library”.

I would ask the opponent if they intend to interrupt the shortcut at any point, and if they say no, allow the player to pick up the library find the card they want at the top and put the cards above it on the bottom in the same order.

It really is no different to looking at every top card and putting on the bottom, but quicker.

Originally posted by David Hibbs:

This question comes up from time to time. The answer is as follows:

Can you tell me EXACTLY how many times you will repeat the loop? (And by EXACTLY, I mean without having to resort to high-level math on sort orders, logic, etc etc.)

If the answer is no, which it is in this case, you can't do it.

You MUST be able to specify a number of times you will repeat a “boundless” loop and you MUST be able to specify the exact ending game state. If you can't do this, you must manually execute the loop and do so in a timely manner.

For more discussion on loops, see Toby's blog here:
http://blogs.magicjudges.org/telliott/2012/11/02/horsemyths/

I think that this is too strict a criterion to apply in this case. The player needs to be able to specify a maximum number of iterations to achieve his desired end state, not an exact number of iterations. The problem with the Four Horsemen deck is that there is no maximum number of iterations which absolutely guarantees that the player will achieve his end state.

In this case, Nathaniel can say, “I have about 50 cards in my library. I will loop through the entire library once so that I can see the order of all the cards, and then I will loop through again to cut to the desired card. This will take no more than 100 iterations in the worst case.”

Without any advanced math, an ordinary tournament player can see how this will work. There is no need to go through the physical motions of Scry 1 a hundred times. If the opponent objects, I would ask the opponent to explain why (possibly away from the table). If the opponent wants to interrupt in some way, of course that is fine, but I can't really see another reasonable objection to this shortcut.

In fact, you can say you Scry 1 exactly 2x your library size if you want
because the last however many Scry 1s can just leave the card on top. No
need to even worry about “maximum” vs “exact” in this particular case.

Originally posted by Oren Firestein:

David Hibbs
This question comes up from time to time. The answer is as follows:

Can you tell me EXACTLY how many times you will repeat the loop? (And by EXACTLY, I mean without having to resort to high-level math on sort orders, logic, etc etc.)

If the answer is no, which it is in this case, you can't do it.

You MUST be able to specify a number of times you will repeat a “boundless” loop and you MUST be able to specify the exact ending game state. If you can't do this, you must manually execute the loop and do so in a timely manner.

For more discussion on loops, see Toby's blog here:
http://blogs.magicjudges.org/telliott/2012/11/02/horsemyths/

I think that this is too strict a criterion to apply in this case. The player needs to be able to specify a maximum number of iterations to achieve his desired end state, not an exact number of iterations. The problem with the Four Horsemen deck is that there is no maximum number of iterations which absolutely guarantees that the player will achieve his end state.

In this case, Nathaniel can say, “I have about 50 cards in my library. I will loop through the entire library once so that I can see the order of all the cards, and then I will loop through again to cut to the desired card. This will take no more than 100 iterations in the worst case.”

Without any advanced math, an ordinary tournament player can see how this will work. There is no need to go through the physical motions of Scry 1 a hundred times. If the opponent objects, I would ask the opponent to explain why (possibly away from the table). If the opponent wants to interrupt in some way, of course that is fine, but I can't really see another reasonable objection to this shortcut.

In fact, in this particular case, it is possible to specify the exact number of iterations needed–you just do it in two steps. First, you count the number of cards in your deck, and Scry that many times, saying that you will put the card on the bottom each time, thus allowing you to look through your entire deck. Then you count the number of cards down from the top your desired sequence is, and Scry that many times, again placing cards on the bottom. Any overshoot (say because we also want to gain tons of life with Kitchen Finks while we're at it) can be handled by replacing the card on top each time.

Or, basically, what Joshua said.

Originally posted by Joshua Feingold:

In fact, you can say you Scry 1 exactly 2x your library size if you want
because the last however many Scry 1s can just leave the card on top. No
need to even worry about “maximum” vs “exact” in this particular case.

Is this the kind of technical precision required, then?

Originally posted by Nathaniel Lawrence:

Is this the kind of technical precision required, then?

I don't really see any reason to require it, no.

If someone has an infinite life combo that gains the life in increments of 2 and wants to change their life total to a billion we don't insist that they first calculate exactly how many iterations would be required to reach that total from their current total, we just let them change it to a billion (and one, if their life total was odd to start with) and only worry about the exact number of iterations necessary if there's a reason to do so. The important thing is that the number of iterations is finite; it needs to be knowable, but it doesn't need to be known.

Let the player look through their library and cut it. It has the exact same end result and saves everyone a great deal of time and trouble.

Thanks for input, everyone! In my mind this did seem different than the usual “boundless loop” problem, but I wanted to be sure that we weren't violating other expectations of player skill or execution to allow this kind of shortcut.

Two worries here. The first one is that, in this particular case, the player has to execute their loop at least a number of times that will cause the deck to return to it's original configuration. Since the game state doesn't change in relation to the loop during this process, why arn't we considering Slow Play here?

Secondly, when we normally apply the loop rules, the player executing the loop does not get any additional information which will inform how and when they decide to end the loop. In this case, they do, because if the player wants to cut to a certain card and they have multiple copies of said card in the deck, they may make a decision based on the relative order of the rest of the cards around those copies. I understand that, when executing this loop the player gets access to that information, but they only get access to it one card at a time, not all the cards at once. Note that allowing a player to take notes about the order of their library would likely be considered excessive. Arguably, it will require more skill to be able to decide which copy of the card to cut to when you only see the cards one by one, therefore access to this information is relevant to the result of the game. Should we be comfortable granting this to the player? I don't think I am.

To me, the loop shortcut rules are designed to let the player get on with the process of doing something the rules clearly allow him to do (in a finite number of iterations!), but would be physically tedious to go through the motions of. It seems clear that the rules do in fact allow a player to see his whole deck and cut to a particular card, and the player can do so within ninety billion iterations (or whatever other ridiculously high number they can name), ergo we should let them shortcut to just doing it.

Darcy, I disagree in that the game state does not change during looping through your deck. I think it is a significant change in game state to know the exact position of every card in your library resulting in a more advanced game state for the comboing player. Since all the actions of the loop by itself are legal and he can say how often he wants to loop I would let him shortcut since I can't really see any infraction in there. As I understand it, the loop rules are there to prevent repeating for a unknown number of cycles when the resulting game state is random (see milling versus a single Emrakul until he is the last card in library). But here while scrying there is no randomness in the resulting game state since it is clearly advanced.

Edited Christian Genz (Aug. 27, 2013 09:56:10 AM)

I'd like to apologize for my previous comments on this loop. My answer was terse and unclear–and to make things worse, I clearly misread and put more into the question than was actually being asked.

Moving onward to what is a hopefully more coherent answer…

I don't want the player to simply pick up their deck and perform a “cut” operation. I feel this is likely to give more information than should be gained by a loop shortcut. I feel that this goes beyond the intent of using shortcuts.

With that said, I have seen this loop happen as a player looks for their win condition. Here are my thoughts on when this happens:

First, I do not require repetition of the sac/trigger/return mechanics. That process is pretty clear, and a long as the opponent doesn't want to interrupt–go for it. I think to make this part happen is a bit… well, excessive.

Second, I prefer the player look at the top card one by one instead of just picking up the deck. They can keep the cards in hand and not move them to the bottom, but by not looking ahead they don't gain information beyond having found their card. This process is pretty quick; they can find what they are looking for and then put the cards from above it on the bottom.

As another option, I've also had a player in this situation just tell their opponent “Scry for Murderous Redcap, kill you?” and have their opponent concede. I like this shortcut better! :) Of course, with the possibility of bluffing, concession is not always an expected response. The player can always ask, though. :)

On Tue Aug 27 14:39, David Hibbs wrote:
> I don't want the player to simply pick up their deck and perform a “cut” operation. I feel this is likely to give more information than should be gained by a loop shortcut. I feel that this goes beyond the intent of using shortcuts.
>
> With that said, I have seen this loop happen as a player looks for their win condition. Here are my thoughts on when this happens:
>
> First, I do not require repetition of the sac/trigger/return mechanics. That process is pretty clear, and a long as the opponent doesn't want to interrupt–go for it. I think to make this part happen is a bit… well, excessive.
>
> Second, I prefer the player look at the top card one by one instead of just picking up the deck. They can keep the cards in hand and not move them to the bottom, but by not looking ahead they don't gain information beyond having found their card. This process is pretty quick; they can find what they are looking for and then put the cards from above it on the bottom.

I definitely think this is requiring too much of the player. As a further example, lets say I can scry 2 any number of times. With this I can sort my deck into any order in exactly 2*(n^2) iterations. This is clearly shortcuttable. I can state an exact number of iterations and the end position of all objects in the loop. In fact, _not_ doing so probably counts as slow play. While it's tempting in your case to say “iterating scrying one is quick, they should do it”, here actually performing the bubble sort on your library will not complete in reasonable time.

In the ‘cut’ case, if you're worried that they don't know how far down in the library the card they want is, then they can do two separate shortcuts. One is ‘scry 1 (size of library) times always putting them on the bottom’ - which allows them to view the order of their library. Then they can do ‘scry 1 (distance to card I want) times always putting them on the bottom’. This doesn't even use the trick of switching to ‘do nothing’ scrys to make up the rest of the number we proposed to do.

We allow a player to propose to advance the game to a known state all the time. ‘go’ and ‘can I attack’ are some of the more common ones. We have a restriction that you can only advance the game a known, finite number of steps, but here they are doing that. We shouldn't require them to play it out for the sake of playing it out. Indeed, the slow play guidelines say that they must not play it out if they can shortcut (when it's a loop).

Matt

I like your write-up, Matthew, and I agree.

Also, Colleen, you got an important concept - we want to save time for “real” game play.

As David says (in his 2nd post), just telling the opponent what's about to happen - and explaining how, if needed - is an ideal shortcut. (Again, assuming it's not a Four Horseman or “until Emrakul is the only card” bit of nonsense!)

And with that, I think we've covered all the salient points. At the risk of showing an unnatural fear of loop questions (LOL), I'm going to wrap this one up.