Originally posted by Peter Richmond:

For argument's sake, in 1.4 of the IPG for backups, one of the actions stated is that “a shuffle is reversed by shuffling again.” Assume that we agree that this is a GRV and we are backing up. Is this or is this not supported by this portion of the IPG?

Without debating the particular definitions of these things, a cut is not necessarily a shuffle. Thus this section of the IPG would not apply.

I would shuffle the deck. I feel that leaving these five cards in a row is more damaging to the game state than shuffling the library.

I wouldn't shuffle the deck as part of the fix to that GRV. Honestly, I think the best thing we can do is leaving everything as it is now. Assuming an ordinary cut (at least 10-15 cards in both piles), those cards supposed to be on bottom won't appear again.

Hopefully, a fetchland will set everything fine :)

Edited Joaquín Pérez (Feb. 25, 2015 12:01:27 AM)

Originally posted by Florian Horn:

I would shuffle the deck. I feel that leaving these five cards in a row is more damaging to the game state than shuffling the library.

Strongly disagree, for the reasons in my previous post. Basically, if you shuffle, the player is far more likely to draw a card they “shouldn't draw” (because it should be on the bottom). If you leave the library as it is, those five cards aren't going to be drawn (unless the game goes on for 18 turns), and the player is still drawing from a portion which is effectively randomised.

Game rules allow players to sometimes know the top or bottom few cards of the library. It doesn't allow them to imperfectly(roughly) know the position of cards in the middle of the deck.

Shuffling the deck allows for the most fair option since we are not able to put the correct cards on the bottom.

I wouldn't get caught up in them drawing a card that they “shouldn't”. The percentages that they would draw a card out of those 5 after a shuffle is quite small.

The philosophy 9 deck shuffling is that the whole thing is randomized . The deck was just badly randomized, it should turn into a fully randomized deck. We randomize be shuffling. While a deviation, as its not explicitly stated in the ipg, I believe it fits within the philosophy 9 decks being randomized.

Originally posted by Nick Rutkowski:

Game rules allow players to sometimes know the top or bottom few cards of the library. It doesn't allow them to imperfectly(roughly) know the position of cards in the middle of the deck.

This is an interesting point, but ultimately I think you've missed the mark. Game rules neither allow nor disallow a player to know the position of any card in their library - the game rules only tell us that libraries should be properly randomised at certain times, and that we should follow the instructions written on cards. So yeah, it's wayyyy more common for a player to know what the top or bottom few cards of their library are than it is for them to know something about another position in the library, but Unexpectedly Absent, Oust, and Long Term Plans do exist. It's certainly weird but there's no direct rule against it.

So what I'm interested in is a] I'm drawing a functionally random card in my upcoming draw steps (more precisely “my library minus 5, randomised”); and b] a segment of 5 known cards, which I can't interact with, are not going to be drawn in upcoming draw steps (which is knowledge I actually should have, in the uncorrupted gamestate).

Originally posted by Nick Rutkowski:

The percentages that they would draw a card out of those 5 after a shuffle is quite small.

Have you done the calculations? Because I'd happily bet that this is incorrect - it's virtually guaranteed to turn out to be a birthday paradox.

Edit: I just did some quick number crunching, and if we assume a library of 40 cards (seems reasonable for the point in the game where Dig Through Time is being cast) i.e. 35 random + 5 “bottom” cards, then we shuffle, the probability of drawing at least one of the 5 cards that was on the bottom after N draws looks like this -

1 0.125
2 0.2371794872
3 0.3375506073
4 0.4270707955
5 0.5066442961

I definitely wouldn't call that “small.”

Basically it seems overwhelmingly to me that unless we anticipate a number of draws over 15 before there's either a natural shuffle effect or end to the game, not shuffling leaves us with a game that plays out much more closely to the uncorrupted gamestate than shuffling does. If we can never get back to A (uncorrupt gamestate), and our choices are stick with B (the currently corrupt gamestate) or move once more to C (a different, but still corrupt gamestate), why would we move to C if it's even further away from A?

Edited Alex Roebuck (Feb. 25, 2015 05:01:31 AM)

Originally posted by Alex Roebuck:

Basically it seems overwhelmingly to me that unless we anticipate a number of draws over 15 before there's either a natural shuffle effect or end to the game, not shuffling leaves us with a game that plays out much more closely to the uncorrupted gamestate than shuffling does. If we can never get back to A (uncorrupt gamestate), and our choices are stick with B (the currently corrupt gamestate) or move once more to C (a different, but still corrupt gamestate), why would we move to C if it's even further away from A?

I think that a randomized library is preferable to having the (possibly) known five card pile. The potential advantage from knowing the next 4 draws when hit that pocket if fairly large, and while it's unlikely that they'll get there in a standard event, if they don't hit any fetches, it is quite possible. Jeskai Ascendancy can really turbo through cards, and knowing the next streak is amazing when deciding on the optional loot trigger. The other decks that could easily get there (and play this card) are various control builds, which can play cards like Divination and Jace's Ingenuity as well as various scry effects to get to that point in just a few turns, and likely have ways to guarantee the game lasts that long.

Now, if this happens in a legacy event, we have a much greater chance of reaching the stacked portion of the deck. Decks like solidarity actually pay close attention to the order of cards put on the bottom with DTT because they can, and often do, draw their way to the portion of the deck that has been stacked, through the use of piles of cantrips and large draw effects, some of which also manipulate the bottom of the library. It's quite possible that that bottom 15 cards of the library had been legally known prior to the accidental cut, due to cards such as Opt, Impulse, and other Digs, and the pocket could be reached in just a few cards from the top after the cut.

Edited Dan Milavitz (Feb. 25, 2015 07:52:02 AM)

To shuffle or not to shuffle, that is the question.

Right now if the players memory serve them well, they have knowledge about the order of their deck. Then the error in the rules occur when the opponent cuts the deck. Assuming we all agree this is an ordinary GRV, the question becomes what is the fix? If we attempt to fix the deck what we're really doing is trying to fix one problem by creating another problem.

You and the opponent don't know the cards that player (and only that player) were entitled to see. You don't know if they're stammering “uh, yeah, I think it was four Opulent Palaces, and a Temple of Malady” because it was, or because that's the most ideal thing to get out of the way. The idea to verify that those cards were clumped together sounds nice, but what if they're lying? Suddenly we catch them in a lie and they say “oh, I guess I forgot.” What if they get lucky, and those 5 cards ARE together?! What if regardless of all of that we just revealed to the opponent 5 cards they haven't seen all game in what looks like a match against Blue Black control due to missing all their green sources?

I don't think we should fix this problem by creating other problems. Lets let the cards stay where they are because nothing that I can recall in the policy instructs us to do otherwise. I'm not opposed to trying to fix this if there is some way we can assuredly fix this (the cards were cut and returned facing the opposite direction; the piles of the deck are touching, but both players know and agree that the cards are touching slightly off base of each other at the point of the cut…), but I'm not fine with letting players manipulate information that not everyone was entitled to in order to fix the problem.

Not everything is going to be perfect when mistakes happen, and that's life.

Edit: I'd like to add that I don't agree with shuffling because I don't agree with backing up, which is what we would do when we shuffle to fix the shuffle (the cut). Leaving the game in the current state is not a substantially worse solution (to me). The cards are in some known and desired order, perhaps beneficial or detrimental for the affected player, in some unknown place in the deck. Shuffling the deck to randomize the deck because it was accidentally randomize when it shouldn't have been is where my whole “fixing problems with problems” stance comes into play.

Edited David Greene (Feb. 25, 2015 10:05:11 AM)

I'm behind shuffling them. I don't think that you need to try and fix the outcome of the Dig Through Time effect (5 known cards on the bottom), but you need to fix the cut (which is essentially a type of shuffle, although the poorest one of them all).

For what it's worth the DTT player could have send 5 good or 5 bad cards to the bottom, having these shuffled now completely is way better than a possible game that gets dragged on and the DTT player starts noticing half-way through his deck that he is about to draw the sequence of 5.

I'm not sure I understand the motivation of people trying to shuffle.

Dig Through Time puts 5 cards on the bottom of the player's library. Importantly, it puts those cards not on top. Following a cut, even if the cards aren't on the bottom, they are still not on top. Given that DTT decks also play fetch lands, it is fairly likely that the deck will be naturally shuffled before that cluster of cards is hit.

That means you would be deviating in an attempt to fix the game, but creating a different (and, in the short term, worse) problem in the process. This just doesn't compute. If you are going to deviate on a fix (and don't do that) then it should be to make the game strictly closer to the correct configuration than it was before.

I am also completely comfortable asking the DTT player which 5 cards he put on the bottom (away from the table) and taking his word for it, if that means I get to do a clean rewind. Or if the deck is still askew, un-cutting it to that point. This is not something the DTT player could have ever planned, and the angle shot of naming 5 “bad” cards and hoping they all appear in a cluster so the judge will un-cut them away is pretty unlikely to work. If the investigation fails to yield a clean un-cut position, I'm leaving the game as-is.

The shuffle is just for the hell of it. The top (N-5) cards are supposed to be in a random, unknown order. If AP has a one-of in his deck, or a single foil or alternate art card, whatever that was in those five cards he put on bottom, then if we don't shuffle he can eventually reach a point where he draws a card, and he knows his next four draws. And that could be good or bad for either player, but that's not for us to determine. And while that is closer to what the deck should be, it should be at the bottom, not 15 cards down in a 40-card deck. And it's not unreasonable at all for a deck playing Dig Through Time to be able to get there in just a few turns (nor is it unreasonable for there to be a shuffle effect anyways, but it's not our job to predict that scenario.)

Since we have a deck that is not completely randomized, and is improperly configured because of the cut, the simplest solution to me is to shuffle it. Now it's completely random. Yes, You can't assume the five cards that were thrown away are all bad. In any seven-card pile, you can easily have three or four cards you want. That you can only choose two does not mean you only want the two you chose. I'd rather just shuffle it, make it completely random, politely tell the player life sucks sometimes (as current policy leads to us doing on occasion anyways) and move on.

And as for asking the player what five cards were on the bottom… “Yeah Judge. Island, island, plains, plains and some throw-away card I don't care about. Thanks, Judge. Can you hang around in case I uh ”accidentally“ shuffle my hand into my deck and need to get it back?” No thanks.

Edited Sean Stackhouse (Feb. 25, 2015 04:47:42 PM)

Originally posted by Sean Stackhouse:

And as for asking the player what five cards were on the bottom… “Yeah Judge. Island, island, plains, plains and some throw-away card I don't care about. Thanks, Judge. Can you hang around in case I uh ”accidentally“ shuffle my hand into my deck and need to get it back?” No thanks.

I believe what Joshua was suggesting is to only implement that fix if the cards listed where all next to each other as they should be.

Originally posted by Sean Stackhouse:

The shuffle is just for the hell of it.

Sorry, Sean, but that's possibly the worst rationale I've read yet.

Judges, please, PLEASE - do not ever apply an infraction or remedy “just for the hell of it”. Follow policy to the best of your ability; when the situation is just weird enough to cause these kinds of problems, then be sure to follow the guiding principles of philosophy as best you can.

I think I've let this one ramble on too long, so:
1) DO NOT SHUFFLE the deck to fix this.
2) it's OK, if you get reasonable answers from the players, to undo the cut (or come very close to it).
2a) Nico, about how much of a cut did you do? Right about there? OK, thanks.
2b) Alex (away from the table if he prefers), which card was on the very bottom? do you recall the other cards?
2c) Are there other cards that have been ordered - Scry, another Dig, etc?
3) if you find you can't reconstruct the proper place to “un-cut” the deck, then leave it as is.
4) DO NOT SHUFFLE.

I think Josh Feingold gave us a very good analysis, trying to keep people from getting lost in the weeds.

d:^D

Well, since there's an official procedure I guess leaving the deck as it is, is the way I'll have to go. I will 100% never actually ask the player to tell me what the cards were and try to reconstruct the deck that way. If the opponent can't verify the cards, I'm not going to trust his opponent and ask him to do the same.