The Hidden Corners of HCE

The new IPG has been out for a week now, and people seem really happy with the new Hidden Card Error (HCE) infraction. We’ve gotten lots of kudos on how intuitive the infraction is and how easy it’s been to explain to judges.

In particular, by being philosophy-driven, it means that most situations in which it applies should be quite intuitive and, conversely, a situation where HCE seems horribly wrong is a good sign that it probably is! Seriously, if you find yourself in a situation where it looks like applying the HCE remedy would do ridiculous things to the game, issue a GRV and talk about the situation with other judges later.

That being said, there are bits of confusion here and there, and I thought it’d be worth talking about them, both to provide some consistency in how they’re handled and clarity about the infraction and what it means. Understanding these scenarios should help you see the borders of HCE and how it’s intended to work.

Remember that we’re going to descend deep into the corners here. The first few are kind of normal, but it starts to get crazier. That’s OK! Policy always starts by making sure that the common cases are handled well, and we’ll refine it to deal with the corners better as they come up.

Dark Confidant

Dark Confidant has been a source of pain since it got printed. The Missed Trigger rules reduced the annoyance substantially, but there were still a few corners of pain. That’s why we put a Dark Confidant example into HCE explicitly, because it now handles those corners.

Note that the example refers to “resolving” a Dark Confidant trigger. Simply drawing a card for your turn isn’t resolving Dark Confidant incorrectly; it’s missing Dark Confidant entirely. Resolving Dark Confidant involves something along the lines of announcing “Resolve Bob”, then putting a card straight into your hand, or starting your turn and drawing two cards.

The reason we chose this example was because it’s important to understand that HCE may leave players “down” a card, and this is acceptable due to the nature of their error. If I remember my Dark Confidant trigger and fail to reveal, I lose the card, even though there’s no scenario in which I wasn’t supposed to have that card. Losing the card is the penalty for the error.

Rummaging Goblin

People have wrestled with the “If a pending ability on the stack would result in a legal overall outcome” sentence. It’s best not to dive too deeply into what it means to be on the stack versus the general idea that there’s a sequence of related actions going on, and if one of those upcoming actions is going to fix the situation, that’ll do.

So, if I activate a Rummaging Goblin and draw before discarding, there’s clearly a draw action coming up, even if technically all the bits that need to be paid to put it on the stack aren’t there. This part of the rule says that Tap-Draw-Discard instead of Tap-Discard-Draw gets you a warning and we move on. Likewise for Horizon Spellbomb resolved out of order, or Serum Visions done backwards. Think of it as a slight loosening of Out-of-Order Sequencing (which it isn’t quite) to make everything work.

Sensei’s Divining Top

There’s still an example in LEC for a player looking at the top 3 cards of their library for a Sensei’s Divining Top that’s no longer in play. I don’t think it much matters. The result – shuffle those cards into the library – is the same either way.

Fundamentally, HCE is about giving the opponent agency. And in the case of someone looking at the top 3 cards of their library, there’s none to be given – the shuffle itself fixes the infraction. So I don’t think there’s any need to involve the opponent here.

The Forbidden Look

I’ve written about the forbidden look before. A player mulligans, scries, then announces they’re going to mulligan again. The guidance here – they’re going to mulligan down another card, most likely – hasn’t changed, but the path has, technically. Honestly, it’s a bit handwavy, but here’s the process:

The player draw their hand of six. They scry. Then they announce a mulligan and shuffle their hand into their library before anyone can stop them. Maybe they draw a new hand, maybe they don’t.

Looks like the infraction was shuffling cards into their library that they weren’t supposed to! The HCE fix is to show the library to the opponent and let them choose the number of cards that are supposed to be in the hand – in this case, five, since that’s where they’re mulliganning to.

Of course, odds are the opponent will just short circuit that and mulligan to four right away. And that’s fine, since they have that option.

(A strict reading of the infraction might be interpreted as the opponent chooses the six card hand because that’s what got shuffled in. But that doesn’t pass the smell test here, since it would be totally irrelevant.)


Apparently there’s been a rash of situations where a player has put their hand down on a Morph card, and picked it up while doing so. My goodness! Face-down on the battlefield philosophically qualifies as a hidden location.

You reveal the hand to the opponent and they choose a card to put onto the battlefield face-down. It may not have morph, but we never had an assurance it had morph in the first place, and manifest has allowed scenarios where a face-down card without morph is possible. Of course, we’re not going to penalize the player for having a face down card that doesn’t have morph!

It’s worth noting that if this happens with a manifested card, the card that returns to the battlefield should still have a manifest ability attached to it.

Pyxis of Pandemonium

So I just exiled two cards into my Pyxis pile. Whoops.

This is one of those situations where the HCE solution is obviously insane as there’s a philosophy mismatch even while technically meeting the standards. Welcome to deep in the corners! Functionally, all those cards are identical, since nobody has knowledge of them, and there’s no need to mess with that, even though the zone is technically hidden. Just shuffle one back into the library at random.

Sylvan Library

That’s not a real card.

What? It’s in Gatherer.

People get really upset when I say this, but the truth is that we don’t think about Sylvan Library when we’re writing policy. That card doesn’t work in the framework of real-life Magic, and the best thing to do is just let judges improvise a solution as best they can.

That being said, it turns out that HCE actually provides some help here. If they’ve gotten themselves into a situation where the Sylvan Library cards have ended up mixed in with the rest, ask the player how much life they wish to pay. Then, let the opponent remove whatever extra cards are left and shuffle them back into the library.

Speaking of shuffling cards back in, what’s with not putting them back on top if the cards were known by the player before being drawn?

There’s a certain appeal to being able to make things get back to where they were, and it’s something that we’ll likely explore moving forwards. As with the Dark Confidant question, it’s OK to not restore the game to match exactly. The penalty for making an error of this type is losing a card to your deck, and that’s consistent.

31 thoughts on “The Hidden Corners of HCE

  1. Hi!

    So, the HCE category doesn’t provide any guidance or any further examples to deal with IDSG. For example, the new version of the IPG doesn’t mention a person taking a mulligan after they decided to keep as a case of HCE. This article now provides some guidance, but for those that are seen the IPG for their first time, they don’t know what used to be IDSG and now should be handled as HCE.

    I believe it would be clearer if it had more information about those former IDSG cases.

  2. Could you expound on the Morph case a little. From the article I don’t understand the situation. As I read it a player picks up their own morph on the battlefield but doesn’t do anything with it (e.g. put it back to their hand). What’s the mistake that we’re fixing with HCE here?

  3. About the Rummaging Gobling part, so the filosophy here is:
    “Don’t worry about that mistake, complete the procedure, don’t do it again and you have a penalty for doing that.”

    But now im just imaginating a corner situation, or maybe im just bluffing, last round before top, the players are playing to enter the top 8, last game, the oponent is in 2 lifes, last turn of game, the player who makes Rummaging have 1 card in hand, and he “makes the mistake”.
    The presure of situation gives you the idea that is unintentional and he actually has not commited other errors during the tournament.
    So, you apply the HCE?

    Correct me if I made a mistake during this example.

  4. I love the clarification for how to remedy HCE when a player shuffles his hand into his library. However, I admit that I (as well as other judges who came asking me about it) did not know how to handle this before this clarification (I even posted in your last blog about it). Isn’t this worth mentioning in the documents, even though it is corner case-ish?

      1. I see. I just thought it was such a different way to fix in comparison to others that it’d be completely non-intuitive in comparison to the rest of the rules, hence the need for a clarification.

  5. OK, I didn’t get anywhere with this question on IRC so we’ll try here:

    Is there a philosophical reason (other than “that’s the way we’ve always done it”) why, if a player gets an HCE, we allow the opponent to “reverse” the error through the “thoughtsieze fix”, but don’t (ever) allow the player to try to resolve the original ability again?

    Take Idyllic Tutor, because I’m bad at coming up with relevant cards for examples. I Idyllic Tutor and put a card in my hand without revealing it. Mr. Judge comes over, shows my opponent my hand. My opponent chooses the Opalescence in my hand (obviously) and it gets shuffled into my library. The damage from my error is appropriately mitigated. We have effectively rewound the effects of Idyllic Tutor – why don’t we then try to resolve the tutor properly? Especially if the error is caught immediately, it seems *more* damaging to the game state for the judge to “deprive” me of my chance to search for a card.

    I’m not a fan of “you screwed up, your penalty is losing your tutor effect” because it isn’t consistent. In some situations it is inconsequential, in others it is a game loss. What’s more, this isn’t the penalty, it’s the fix – the Warning is the penalty.

    1. For the same reason the penalty used to be a Game Loss. It needs to act as a deterrent for something that can be hard to catch in many situations.

    2. Think about it for a moment. Other wise if I have Opalescence in hand already I can use Idylic Tutor to tutor for any card (preferably one not as scary as Opalescence) and get just a warning?

    3. What if they already had the Opalescence in their hand and found something else? What if they have two Enchantments in their hand?

    4. There seems to be this misconception that the only penalties we should be issuing are Game Losses or Warnings, and that we should issue any real penalties inside the game. This doesn’t feel true at all, because every game has in-game penalties for infractions at every level of play.

      Football – Loss of yardage
      Baseball – Plays being called dead or runners being called out.

      Even in esports, you have in game penalties. In League of Legends, if your team shows up late, or makes a last minute substitution, or players trade champions within the last 20 seconds of Champion Select, the team loses one of their three bans.

      Even in Pokemon, certain DDLPs and the Marked Cards upgrade aren’t Game Losses, they allow the players opponent to take 1-3 of their Prize Cards, depending on the severity.

      We’ve kind of pigeon-holed judging MtG into penalties being “do nothing, end the game, or disqualify them” and I honestly believe there are many more lighter touches we can be doing. Yes, it makes the game slightly less “organic,” but for the purposes of tournaments, the IPG really is an extension of the MTR.

  6. How does HCE apply if a player accidentally shuffles their hand into their library in the middle of a game (not connected to mulligan decisions)?

    1. Reveal library, opponent gives you a hand.

      This’ll come up almost never. It’s just an interesting quirk.

      1. How about in situations where the player, after shuffling their library, places it on top of their hand?

        Do we just treat this as if the cards had been shuffled in, as per policy that a card which has been placed on top of a players hand is an unidentifiable part of the hand?

  7. Question about the “complete set of cards that contains the unrecoverable information” phrase.

    In Atlanta this last weekend I had a player start resolving Collected Company, pick up 3 cards with his left hand, then 4 with his right hand. Opponent stopped him immediately, but he still saw the 4 cards. Both players agree the 3 and 4 never touched or came close to touching.

    Is “the complete set of cards” all 7 or just the 4 with the extra card?

  8. i am not too sure abt
    “a player has put their hand down on a Morph card, and picked it up while doing so. ”

    does it mean
    player cast morph, put the card, and has his hand left the card, and then pick it up
    player cast morph, put the card, while his hand is still touching the card, and then pick it up

    1. A player casts a morph, then later puts their hand down on top of that morph. When they pick up their hand again, they pick up the morph as one of the cards.

  9. What if…

    Anton resolves a See the Unwritten, but ends up pulling/revealing nine cards instead of 8.

    Does the infraction depend on how Anton goes about resolving StU?

    Method A: Anton flips one card at a time faceup to all players and spectators and miscounts or fumbles and does a ninth card by mistake. The ninth card is readily identifiable. This seems like L@EC as the infraction.

    Method B: Anton pulls cards facedown in a pile and two clump or he miscounts, picks them up and looks at them, then calls judge because he realizes he has one too many. He hasn’t revealed what the cards are yet. Anton may or may not have manipulated the cards before realizing, say separating them into creatures and non-creatures. Is this HCE? It seems very similar to example E under HCE.

    Does the infraction depend on how Anton resolves the spell?

    1. Somewhat, yes. If the last card is kept separate, it’s not part of the overall set. Once that card touches the others, just like with drawing, it’s considered to be part of a random group. Until that point it’s separate and distinct.

      Similarly, if Anton pulls up five cards, then picks up the next four, but doesn’t mix the two piles, the remedy can be applied to the pile of four cards.

      1. “Once that card touches the others, just like with drawing, it’s considered to be part of a random group.”

        If the cards are being revealed one at a time, doesn’t the following sentence apply? :

        “If cards are placed into a public zone their order is known and the infraction can be handled as a Game Rule Violation. “

      2. If they’re being revealed, then we’re OK. But if not, then the cards are considered identical and we don’t know which is which.

  10. We had a fun discussion in the US – North Slack about a fun situation:

    Ana casts Ugin’s Insight, with her only non-land permanent being a Nissa, Sage Animist. Insight resolves and she scryes 3, putting one on the bottom, then draws 3 cards. At this point Nacho calls a judge, since Ana shouldn’t have scryed at all.

    Is this an HCE? How do we fix it? What if Ana had a 2 drop, but still scryed 3? Then we’re pretty sure we’re well into HCE territory, but still unsure of the fix. In both situations, it seems like doing nothing or shuffling the one card that was shipped to the bottom into the unknown portion of the library is the best course of action.

    We also came to the conclusion that some language needs so cleaning up. Specifically the part about not repeating bad actions. The current IPG reads like if we catch the error while still scrying, we take the extra card(s) and shuffle, then finish scrying, but if we catch the error after the scry happens, we take the extra card(s), shuffle, but keep any that were legally scryed where they are. This is almost certainly wrong, but I think could use some clarification.

    So, our conclusions:
    1: What do we do when a extra-sized scry or something similar happens, followed by another action that isn’t fixed by a simple back up before we catch the error? Nothing? (Examples include scrying too many with Ugin’s Insight, then drawing, or maybe Digging extra cards, and not catching it till we put the two in the hand.)

    2: We like the philosophy, but some of the language could be a lot clearer.

    1. If the set of cards that had a problem no longer exists, it’s still HCE, but there’s no remedy to be performed. Language will definitely be cleaned up for the next release!

  11. About the morph upgrade.
    We now have very little GLs in the IPG, tending to a more friendly competitive events.
    But in this case, why the morph case is still a GL ? I am not sure it is “worst” than the other cases we delete ? Or is this because it is difficult to fix properly ?

Comments are closed.