I already posted this as a rules question, but as the moderator correctly put it, this is not a rules question but a “how do we handle it question.”
http://apps.magicjudges.org/forum/topic/33634/

The question:
I had an interesting interaction with my EDH deck.

I make infinite (10^50 to be exact) Golem tokens withPrecursor Golem (using Nim Deathmantle, Krark-clan Ironworks,Dross Scorpion and Grim Monolith). I then target Precursor Golem with Chaos Warp.

What happens? I am not sure how many non-permanents there are in the deck, but it is maybe 5-10 of the 60+ cards in the library. Pretty soon I will be hitting those cards over and over, without having found all my permanents.

What happens in this scenario? Imagine we are at a competetive Legacy event.
- Will I be forced to break the chain of events by sacrificing all remaining golems to Ironworks? What if I sacrifice it to itself prior casting Chaos Warp?
- Will I have a limited amount of tries to find new permanents once they start drying up, i.e. the same non-permanents start coming up over and over?
- Will the game be a draw because this takes forever? Again: Any difference if I sacrifice the Ironworks prior playing Chaos Warp?

Did this happen at an event? Is it likely to happen in the near future?

Policy is designed to handle things that regularly occur in events in a fair and consistent manner. I wouldn't worry about this coming up and handling it differently than another judge might.

A far more likely scenario to come up is posed in this week's Knowledge Pool - I think you'll find that you'll advance your policy knowledge more by participating in that discussion.

Even of it is a corner case I believe it is good to discuss such situations to improve our understanding of the rules/IPG/MTR/other documents. This is my interpretation:

From the rules view everything is fine, you just resolve each of the triggers even if this might be 10^10000 or more.
Okay, so let's look at the situation from the IPG perspective: What might hinder the execution of every trigger? Slow Play. If we execute a lot of triggers without advancing the board state we surely waste a lot of time. What can save us from that problem? Shortcuts. However, you can't give an time where you got all your permanents out of the deck. No matter how much tries you have, it can always be the case that you don't hit all your permanents. So no shortcuts.
If you can break the chain I would make you do it after you didn't hit something I would make you hit something. If you don't have that I would rule that you can shortcut into not finding anything for the remaining triggers and have to do that to not waste time (or you will get an updated Slow Play ending the game quite soon). This seems a little bit out of the box because you can't normally “find nothing” for a reveal of the top like you can for the search, but I'm willing to make that deviation here to ensure your game will end in timely fashion. I won't allow the tournament be delayed by you resolving such a number of triggers every round (end of time wouldn't stop you because everything happens in one turn and extra turns will never end)

That just sounds like Four Horsemen with extra steps.

http://blogs.magicjudges.org/telliott/2012/11/02/horsemyths/

I thought about that as well, but then I realised that the problem does not arise from a loop so it seems a little bit different, because there might not be a normal game decisions that stops the loop.

In this scenario like much others… If you can't tell how many times you do something (in this case, 10^50 times do chaos warp) and the EXACTLY final game state (100%, not 99.999999% chances), the scenario becomes:
-Slow play for do 10^50 actions
or
-Something posible or not… but without usefull solution.

It's similar to the scenario with 2K Goblin Test Pilot tokens activing their hability in response… we can think about the chance to everyone to die or not… but since it's a random event, we can't do “this happens 100%”

Here is the same… Why can't you show a non-permanent card 10^50 times? the posibility exists…
What's the chance of get 27 consecutive heads in coin-toss?? I do that once! Of course, only once in a very large time playing Pokemon… but that 27H-coin-toss gives me a tournament…

Random is random… we can't handle it with rules

(all this was my opinion)

Originally posted by Iván R. Molia:

In this scenario like much others… If you can't tell how many times you do something (in this case, 10^50 times do chaos warp) and the EXACTLY final game state (100%, not 99.999999% chances), the scenario becomes:
-Slow play for do 10^50 actions
or
-Something possible or not… but without useful solution.

Here is the same… Why can't you show a non-permanent card 10^50 times? the posibility exists…
What's the chance of get 27 consecutive heads in coin-toss?? I do that once! Of course, only once in a very large time playing Pokemon… but that 27H-coin-toss gives me a tournament…

Random is random… we can't handle it with rules

(all this was my opinion)

I am somewhat of the same opinion. But I have a problem with “Random is Random”. This will become somewhat mathematical know but easy maths and I try to be short about it.

Let´s consider the coin-toss example from a mathematical perspective: We have a chance of 1/2 or 0.5 to have the Coin come up with tails.
If we want to fix that for 27 cointosses we get 0.5^27 which is: 7.450580596923828125 × 10^-9 or 0.00000000745…. which is extremely low. Slightly lower than the usual Lotto case (6 out of 49) with 6.436011595890 × 10^-8 which is 0.00000006436.
So the 27 cointosses are pretty much a tenth in likelyhood to winning the lottery (6 out of 49). So pretty impressive you got that. :D But hitting this probability is still nothing compared to the given example in which case I would work with actuarial or statistical expectation value.
(BTW: having hit 26 times tails already the chance to hit it again is 0.5 again^^)

Ok…now let´s examine the proposed case:
Proposed are 5-10 cards that can get hit in a 60 card library. Let us blow up the example a bit and take a look at the even more unlikely scenario of having only 1 card left and 100 cards still in the library. (to make the math easier). At some time only 1 permanent that can be put into play will be left in the library. And making it still a 100 card library makes hitting that one even less likely than only having 60 cards. So everything is fine.

Now…hitting exactly the card we want is 0.01 and we have 10^50 tries. So to make it easier we make it 2 cards. 1 with the probability of 0.01 and one with the probability of 0.99 (so like the coin but the heads comes up only once in 100 tosses on average) Using the same idea as before we have 0.99^(10^50) so this makes it roughly a 0.(47 zeros)1. So the chance of NOT hitting our last permanent has a probability of 0.0000….0001 (with 47 zeros) an thus incredibly unlikely (having an actuarial expectation of 0) and thus a “just put all permanent into play” is fine for Regular REL.

Now to come to competitive REL. Here we have the same problem as the horseman combo. Between every try the deck has to be shuffled and every time you do not hit the permanent you have to shuffle again without advancing the board state. (This can be discussed…you lose a golem token, so the board state actually changes) This would result in slow play and thus if you can you should brake up the combo or deal with the slow play problem.

In conclusion:
So in competitive REL I would handle the combo in the same way as the horseman combo is handled.
In Regular REL just put the permanents onto the deck and ignore the rest of the Chaos Warps

This would be my ruling. I am happy to be corrected :)

As others have said, we already have a way to handle this at Competitive, so I'll restrict my thoughts to Regular.

Honestly, my first impulse is to just tell the players to handle it however they see fit. If one of the players wants to be a troll and make you go through the motions every single time, then that's the price you pay for putting this combo into your deck in the first place. If all the players agree that you can just have all your permanents in play, then there's no reason for a judge to be involved at all.

As far as actually resolving the stack, there are some shortcuts you can employ. For example, once you have no eligible hits left in your deck, shortcutting the rest of the stack is fine (you might want to ask a judge to verify that there are no eligible hits left if one of the other players doesn't trust you for some reason).

You can also employ the “mind's desire” shortcut (not an ‘O’fficial shortcut of course) that says that removing the top card of a randomized library leaves a randomized library. As a practical matter, this means that every time you hit with a chaos warp, just treat the rest of the deck as already randomized. Obviously this wouldn't apply if the top card isn't a hit (since it stays on top) or if a card is actually shuffled in, but this would cut down on a lot of the shuffling.

TL,DR: you kind of did this to yourself, but even if you are forced to go through the motions there are some shortcuts you can employ to streamline to process.

Still on comp rel, I don't understand what “handling like horsemen” means if you don't have a way to deal with the tokens before the copies of the spell resolve to fizzle them(which is the most intresting case in my opinion) Am I missing something here? Because with horsemen the player can break the loop by legal game actions, which doesn't seem to be the case in this scenario.

Edited Markus Dietrich (Feb. 24, 2017 09:13:23 PM)

Agreed that this post should be moved to Regular REL. You did say it came up with your EDH deck, so until there is a reason to suspect something like this appearing at a “competitive Legacy event”, it's not fruitful to discuss it at Comp REL.

Working with Regular REL, basically whatever you, your opponent, and your judge (if applicable) agree on is going to be what happens.

For a practical solution, why don't you sort your library into X permanents and Y non-permanents? Generate a random number between 1 and (X+Y); if you roll X or lower, choose a random permanent to put on the battlefield, decrease X by 1, and repeat; if you roll X+1 or higher, just repeat.

Thank you for so many interesting answers. I want to point out that the whole point of this question is to highlight the holes in the rules regarding how these situations are handled. The scenatio as such is just an example of “yes, this can actually happen. How do we handle such cases in competetive?”

This means that:
- no we cannot move this to Regular REL forum. Its not like my playgroup would stop me from putting my permantens into play.
- Likelyhood calculations does not give much. I can generate enough tokens so that the probability of failing is arbitrarily low.
- “You chose to make that play, you handle it” isn't an accepted answer (please let me know if you ever heard of such a ruling!)
- “Has this happened” is not the thing here - Focus on “It can happen”.

Interesting discussion points:
- I generated all tokens last turn, my opponent has Moat and destroyed my Ironworks during his turn. How can I be punished for playing my Chaos Warp? Or for playing my deck?

- As stated, the likelihood of success can be as high as I want it, but not fully 100%. Likelihood needed to legally rule you father of a child is ridicoulusly lower (I heard numbers of both 95 and 98% conficence interval.

- At which point do I need to stop trying, say that is the ruling? I.e. how many times can I fail finding something? When I start there is about 60 cards left in the deck, there is maximum 10 non-permanents so to begin with I will almost always hit. Is the “retry counter” reset each time I hit something, i.e. is that to be considered progressing the board state?


Finally, thanks all for your thoghts! This is a tough subject for sure, and I wish that the rules allowed players to do what the game makes possible. Especially since it is obvious that the intended game state will occur, was it only allowed. The policy that this would be to tough for magic judges to handle is obviously falulty.

Edited Olle Liljefeldt (Feb. 26, 2017 02:59:26 PM)

This is the Competitive REL forum, questions & discussions should be about things that happen in Competitive REL so introducing Regular REL answers is off topic. Suggesting that the discussion move to Regular REL can be handled by reporting the original post, not adding to off topic discussion.

As it stands, the tournament rules cover things that are likely to happen at competitive REL events thankfully I don't think we will see this coming up.

To expand on Mark's point- The IPG and MTR aren't trying to be comprehensive. There are far too many situations that can come up in real life for a written resource to have any chance of covering everything.

“Holes” in the policy documents are quite abundant. In fact, most of the posts in this forum are about that very thing- if the question were covered in the written documents, it wouldn't need to be asked here. The policy documents are there to tell us how to consistently handle a few common situations, and to teach us the underlying philosophy so that we can exercise our judgement when something unexpected comes up.