Today is Wednesday and that means it's time for another amazing edition of Knowledge Pool. This week, we'll be looking at a situation that can be commonly encountered in Standard and Sealed events now that the Scry mechanic has made a reappearance in Theros.

This scenario is rated BRONZE, which means it's directed toward L1s and L1 candidates. If you don't fit in those categories, please give others some time to post their thoughts before adding yours.

Big Girls Don't Scry - BRONZE

Alice is playing in a Standard Grand Prix event. She has a Thassa, God of the Sea and a dozen assorted lands in play. She begins her turn by Scrying for Thassa's beginning of upkeep ability. When she goes to her deck, a 2nd card is stuck to the top card. Alice looks and realizes that she has 2 cards and immediately calls for a judge. You arrive at the table. Alice and her opponent state that the cards are still in the same order as she immediately put the cards face down on the table, and also explain that Thassa was played on Turn 3. How do you handle this situation?

You can also check out this scenario and many more like it at the Knowledge Pool Blog! http://blogs.magicjudges.org/knowledgepool/?p=907

Edited George FitzGerald (Nov. 6, 2013 08:55:36 AM)

Well for starters, Alice here is pretty clearly Looking at Extra Cards. Regardless of the fix we apply to the game, she will be receiving a Warning for LEC.

Now, normally the fix would be to take the extra card (the one stuck underneath, not the top card) and shuffle it into the randomized portion of the deck, then continue the scry. In this case, however, we might not know how much of the deck is randomized. Because Alice has “a dozen assorted lands in play” and the Thassa was played turn 3, she's presumably been scrying with Thassa for a number of turns. If Alice knows the exact number of cards that she's sent to the bottom of the library, that'll make the situation much easier; we would remove that many cards from the bottom of the library, set aside the top card of the library, shuffle the extra card into the remaining portion, then return the removed cards to the bottom and continue the scry with the set-aside top card of the library.

However, if Alice doesn't know how many cards she's sent to the bottom, and neither does her opponent, then we have a trickier situation. My gut reaction is to say that if Alice no longer knows the identity of a card on the bottom of her library, then we should be able to consider it random; in that case, if they can agree that Alice sent at least X many cards to the bottom, then we could remove that many instead of the whole amount, and apply the fix as usual.

However, that would result in an incorrect gamestate, wherein cards that had previously been in a set position relative to the bottom of the library are now randomized. Alice could gain an advantage here, as she'd be more likely to draw something that she previously scried to the bottom. Additionally, Alice cannot prove that she doesn't know the identity of the cards on the bottom of her library. I believe that because of this, the technically correct answer would be to leave the extra card where it is, rather than attempting a partial fix. That results in Alice gaining information she wouldn't otherwise have, though, which could change her decision on the scry.

In this situation, we would have to weigh which is more disruptive; altering the game state by shuffling previously scried cards into the random portion, or allowing Alice the additional information by leaving the extra card where it is. I would argue that the second option would be significantly more disruptive. Giving Alice information about the next card in her library could strongly influence her decisions for the scry action and for the rest of the turn. With that in mind, I would attempt the first fix; determine how many cards both players can agree were sent to the bottom, then shuffle the extra card into the other portion and continue the scry from there.

Edited Beau (Nov. 6, 2013 09:51:38 AM)

Matt has the gist of the assessment I would make. However, I think I might go slightly further, and once establishing how many cards the players can agree were sent to the bottom, I might actually leave one more card in its original position. So if they can agree that scry has sent away 4 cards, I would actually separate the bottom 5 before shuffling.

I believe that leaving an otherwise randomized, unidentified card in a static library position is less disruptive than potentially risking a shuffle moving a nonrandom, technically identifiable (even if the player who put it there cannot recall the identity in the moment) card from the bottom back into a position where it may be drawn.

Edit: To be clear, I would only take the extra card if the players were unsure how many cards have been manipulated, but agree that it was “at least X”. If they are confident that there have only been 4 cards shipped to the bottom, then we just leave those 4 out of the shuffle.

Edited Rebecca Lawrence (Nov. 7, 2013 09:31:12 AM)

I also agree on the Warning for Looking at Extra Cards. As for the remedy, I think the major decision comes from the Additional Remedy of shuffling the extra card into the randomized portion of the deck. If it comes down to the two players agreeing on how many cards have moved to the bottom as a result of Thassa's, or any other scry effects for that matter, I would think using that number to separate cards from the bottom of the randomized library would be sufficient.

However if the players do not agree on the number of cards that were shipped to the bottom, does this become a question of whether Alice would be getting an advantage or be hurt by having the wrong number of cards included in the shuffle. For example, her opponent could say only two cards were shipped when the actual number was six, meaning that 4 cards that Alice presumably wanted at the bottom are now randomized back into her library.

If the goal is to absolutely maintain the state of the cards that could have been sent to the bottom when randomizing the deck, would a better solution be to separate the maximum potential cards that were sent, in this case that would be nine turns of scrys (barring any missed land drops,) plus any other scry effects that Alice used in her graveyard and/or exile.

I'm heading in Kevin's direction as well. I'd also want to look through the graveyard or ask about anything else that could have manipulated the library. (other scry cards)?

While the IPG talks about contacting the head judge for deviation of penalties or for rollbacks, this doesn't really fall into either situation. I think I'd still contact the head judge to keep consistency.

My only other note would be that by “not-shuffling” too many cards, we're also shrinking the space the improperly-known card can be found it, which gives her an advantage.

But that's one versus a bunch, so this is less disruptive to the game state than reshuffling the whole thing. I'd ask to treat some potentially cards an known and shuffle accordingly.

Everyone here has stated that they would confirm how many cards Alice has put on the bottom of her deck with the scry ability. However, shouldn't we also ask Alice and her opponent how many cards on top of Alice's deck are also known to her via scrying before we shuffle the extra card into the unknown portion of her deck?

Eric Pare…

A good observation. This would come from checking for other cards with Scry
as the others have mentioned. However, it's only likely to have an impact
if there are multiple Scry 2 cards. It has less of an impact with Scry 1
cards since it's hard to keep that card on top of the library through
multiple turns.

-George FitzGerald
L2, Sarasota, FL

Originally posted by George FitzGerald:

Eric Pare…

A good observation. This would come from checking for other cards with Scry
as the others have mentioned. However, it's only likely to have an impact
if there are multiple Scry 2 cards. It has less of an impact with Scry 1
cards since it's hard to keep that card on top of the library through
multiple turns.

-George FitzGerald
L2, Sarasota, FL

Horizon Scholar, Omenspeaker, and Prognostic Sphinx are scry >1 cards that can possibly be in Alice's deck. Although they're not mentioned in the scenario, we should check the battlefield and Alice's graveyard for such cards to help us determine what's been scryed and left on top.

We need to ask Alice if she did scry 2 or more on her previous turn and left at least two cards on top because if she did, then the identity of the card that's stuck to the bottom of the top card should already be known to her and so we shouldn't have to apply the additional remedy for L@EC and shuffle it away. She still committed an infraction in this case because she saw the face of a card in a hidden zone when an instruction didn't allow it.

Originally posted by Eric Paré:

they're not mentioned in the scenario

and we prefer to stick to the scenario, as introducing additional elements really introduces a different scenario.

I completely agree with the LEC penalty, and that the best remedy will be to randomize all but the top card, and the bottom X cards. However, we might be able to do a better job of determining what X should be.

I would ask Alice if she remembers the first card that she pushed to the bottom. If she does, I would have her tell it to me away from the table, as well as how many copies of that card were pushed. Assuming that card can be found near where the players describe it should be, I would use that as the cutoff.

This admittedly gives Alice a slight opportunity to gain advantage (for example, if she pushed an Aetherling to the bottom early, she might get you to reshuffle it by naming the second card pushed), but I estimate that potential to be low enough to ignore.

Warning for LEC, see if we can get an agreement for number of cards on the bottom, etc etc etc.

In the case that we can't figure out how many cards on the bottom are known, we can figure out the maximum number of cards known and randomize all but those. Here's how we do it:

1) Figure out if Alice mulliganned and how many times. This should be pretty easy. Let the number of times Alice mulliganned be m.

2) Count the number of cards in Alice's revealed zones plus her hand, and then the size of the subset of those cards with the words “draw a card” on them. We'll call the number of cards c and the number of cards with “draw a card” d. We will count “draw 2 cards” twice in d, likewise “draw 3 cards” thrice, etc. The number of turns that have passed, which we will call t, is equal to c - d - (7 - m) (+1 if Alice was on the play)

3) The number of turns Thassa has been in play is t - 3.

Therefore we know the maximum number of cards Thassa has put on the bottom, which is t - 3. Take all but the bottom t - 3 cards and the top 1 card and randomize them.

This is easy assuming Thassa is the only Scry card Alice has played. If Alice has played other Scry cards, I'd probably just leave well enough alone and leave the second card where it is, because Alice gaining additional information is probably less disruptive than shuffling some random number of cards from the bottom that have been Scried away.

The other thing I feel needs to be considered is that she may not have “thrown back” very many cards, if any. Maybe she was always happy with her topdeck. I'd ask the question, “Roughly how often do you feel you didn't like the card you scry'd with Thassa?”, to see if a general idea of how many cards went back.

“Oh, I almost always put the card on the bottom” will yield slightly different shuffling actions than “Oh, I almost never put the card back”

Lyle: It seems like that procedure would take a lot of time and manhandling of the players' cards to complete. Do you think they would be happy with that fix?

While spelling it all out that way makes it sound very complicated, in
practice it's the same sort of card count we do to see if someone has drawn
extra cards. The only real wrinkle I see is accounting for the opening
hand. Otherwise, it's the same.

Justin: sure, but there are a number of differences here. Most importantly, DEC has much higher potential for abuse, so being certain is important. Here, we are just trying to fix up the game so the players can be on their way.