Matter Reshaper dies.
Player announces the ability, then draws without revealing.
What would you do?

It does appear to be textbook GPE-HCE, so have the opponent reveal their hand, and shuffle one card of the opponent's choice into the unknown part of the library, as per the Additional Remedy. While on first thought, this might appear overly punitive, it's quite possible the card was a permanent with a CMC of 3 or less that the player didn't want to put on the battlefield immediately (such as a Pacifism.)

If that top card had been previously revealed though, this would not fall under HCE.

The issue with Matter Reshaper is that even if the card is a cheap permanent, the player is not obliged to put it into play. If you flip a 2 drop you can still choose to put it into hand.

The only error is not revealing the card. The reveal isn't needed to prove that the action was legal. The fix I believe is “Reveal the hand” and carry on with a Warning. The number of “Excess or Unverified Cards” being zero in this case. This also happily fits the philosophy of “giving the opponent sufficient knowledge and ability to correct the error so that it cannot generate advantage.”

I think the card that covers Francesco's situation would be Coiling Oracle

Edited Jose Miguel Sanchez Navarro (Jan. 28, 2016 08:28:10 PM)

Oh yeah, missed the ‘may’ on Matter Reshaper. That does change things.

Originally posted by Mark Mc Govern:

The only error is not revealing the card. The reveal isn't needed to prove that the action was legal. The fix I believe is “Reveal the hand” and carry on with a Warning. The number of “Excess or Unverified Cards” being zero in this case. This also happily fits the philosophy of “giving the opponent sufficient knowledge and ability to correct the error so that it cannot generate advantage.”

Mark, I like your reasoning.

I support Mark's solution and reasoning. This is HCE because the error
cannot be fixed with publicly available information - we don't know which
card in hand to reveal - but there's no possibility that this card
shouldn't have been drawn - we weren't supposed to verify any
characteristics of this card in order to draw it. Intuitively, the simplest
fix is to reveal the entire hand, and HCE supports that solution.

I also agree with Dan.

However let's say you use a Nissa's +1, and do not reveal, it probably should be fixed by shuffling a card from the hand in the library. I already had an issue about how this issue was handled previously and this case is still bugging me.

I have been feeling that the fixes though the different iterations of the IPG since last year left this case on a sad fix solution but I think that is mainly an issue of templating of the ability.

Originally posted by Dan Collins:

I support Mark's solution and reasoning. This is HCE because the error
cannot be fixed with publicly available information - we don't know which
card in hand to reveal - but there's no possibility that this card
shouldn't have been drawn - we weren't supposed to verify any
characteristics of this card in order to draw it. Intuitively, the simplest
fix is to reveal the entire hand, and HCE supports that solution.

If it were the only card in hand, is it correctable with only publicly available information then? Seems like it would be hard to argue that we don't know which card was supposed to be revealed. Does that make it a GRV then with a backup or leave-as-is solution, depending on what has happened since the error?

Yes it is fixable easily like the case of a mystical tutor on top of the deck.

This seems to indicate the new fix for a failure to reveal with Keranos, God of Storms in play would be “reveal the hand” and you get no triggers, where previously it would have been “play on.”

What's going to happen when we see a hand of all land or all spells?

Originally posted by Marc DeArmond:

What's going to happen when we see a hand of all land or all spells?

Nothing. The player failed to reveal a card, trigger never fired.

Originally posted by Marc DeArmond:

This seems to indicate the new fix for a failure to reveal with Keranos, God of Storms in play would be “reveal the hand” and you get no triggers, where previously it would have been “play on.”

What's going to happen when we see a hand of all land or all spells?

Maybe I will be not 100% technically correct, bear with me anyway:
we decompose Keranos ability in a static ability (“reveal”) and two triggered (“whenever you reveal…”).
You drew without revealing? Well:
- with regard to the static ability, HCE, we want you to reveal, but we don't know which card you drew, so reveal your whole hand (no cards are selected and removed).
- with regard to the triggered abilities, well, you missed them.

Seems to be correct. I'm just noting overall that whenever we have a card that was supposed to be revealed and then drawn we are now revealing the entire hand to compensate for the failure to reveal.

This means for issues of drawing without revealing in cases like Courser of Kruphix and Divination, we'd now be revealing the hand while previously this would not have been the case.

The error probably “cannot be corrected by only publicly available information”, and - as far as we can tell from the original scenario - this was done “without {the} opponent's permission”.

So, the player reveals their hand, and the opponent selects zero cards - there are no “excess or unverified” cards in this example. Since no cards are being returned to the library, no shuffle is required.

Jacob: if there's only one card in hand, then “the zone in question contains only a single card”, in which case a simple correction is “reveal that card, please”, and this is a GRV.

Nissa, Sage Animist's +1 ability was brought up frequently in the discussions that crafted the new policy. This ability gets kind of weird, depending on whether there's zero, one, or nothing but lands in hand - and we certainly can't have a fix that depends on that, even though we are revealing the hand and thus eliminating that concern.
And, the IPG text actually handles this:

Originally posted by IPG:

…opponent selects a number of cards equal to the number of excess or unverified cards. Those cards are returned to their original zone.

Nissa's ability requires that you verify that the card should go to your hand, instead of the battlefield; since that can't be corrected by only publicly available information, we apply this remedy - reveal the hand, the opponent chooses one card, that card is shuffled into the random portion of the library.

For things like Keranos' “reveal the the first card drawn each turn” instruction, the error fits the HCE criteria, and there are zero cards that are either “excess or unverified”. And, as noted, the player didn't reveal a land or non-land “this way”, so neither trigger fires. (I suppose a player with 2 land and 1 non-land might argue for drawing 2 and dealing 3… uhh, NO.)

d:^D