I thought of a circumstance where this matters!
Say I've got the following board:
Creepy Doll
Hellkite Charger
Nature's Will
7+ mana
And the opponent has
2 Wall of Denial
Bubble Matrix
So, here's my situation. I get to attack as much as I want. I can't kill them in a single swing, because the Walls block my Doll and my Charger. However, each time the Doll hits, I get to flip a coin and see if the Wall dies. I win the flip once, and from there on out I can guarantee a kill. However, this is a loop with an identical game state, and I cannot tell you in advance exactly how many attacks it will take before the Doll kills the Wall and I get in for the kill.
Am I allowed to shortcut this and end the game?
Am I even allowed to repeat the loop until I punch through?
Under the rule as you described it, I am not allowed to do either, it seems, which is bizarre. If I don't flip correct the first time around the loop, I have to stop or get hit with Delay of Game.
EDIT: Argh, this doesn't work because Bubble Matrix stops the Doll's ability. Going to look for a different way to set this up...
OK, let's pretend that Bubble matrix doesn't stop this.
In this instance, I don't think I would give Slow Play. Here is why:
You could demonstrate one iteration, and if the flip does not go in your favour tell your opponent you will do it 10 more times. At this point you have shown a shortcut to an action in game.
I think that one key difference is that you can't shortcut shuffling your deck. BUT you can shortcut tapping/untapping cards.
If you find another example I'm more than happy to offer my thoughts.
-josh-
How can you shortcut this, though? Each time I attack, there is an independent fifty-fifty chance that I will kill the wall with the Doll's ability. If I tried to "shortcut" to ten attacks, there would be a %5 chance that both walls are alive, a slightly higher chance that just one would be, and a large chance they would both be dead as would the opponent.
So I don't think I can shortcut this random event any more than I could shortcut a shuffle. Both are random and therefore preclude your ability to look past the random choice. In both cases, I cannot tell you, the judge, exactly what the board state will be after X iterations with 100% certainty. I can get arbitrarily close to 100%, but I can't ever actually reach it.
What I was driving at is that the Rule as stated is silly, because it would prevent me from even acting out each attack repeatedly until I win! The only actual difference between this senario and the shuffle problem is one of scale; the shuffle problem simply takes longer to perform each action and has a much, much smaller chance of success each go around. But the Rule doesn't talk about any cut-off for when a random event is so remote that you get delay-of-game for pursuing it.
I think a better rule is one that incorporates the concept of being arbitrarily close to 100%; in other words, a limit. The Rule, as written, is like Zeno's paradox; sure you can shortcut your shuffle problem a googleplex times, but it still won't be a sure thing that you get Emrakul where you need him. It demands you realize an infinity of shuffles to guarantee you hit him. I think it's far better to implement a rule that RESOLVES the paradox; if you can prove that you can get arbitrarily close to 100% of the event occurring, you should be allowed to shortcut the arbitrarily large number of loops and just reach the end result.
