Originally posted by David Poon:

It sounds like you're saying that the infraction should be different, depending on the speed of the player and the amount of communication. That seems… strange?

The communication element is in the definition of HCE "A player commits an error in the game that cannot be corrected by only publicly available information and does so without his or her opponent’s permission.“

Consider the two following scenarios:
- NAP looks at the two top cards, puts one on the bottom and the other one in his hand, then draws a second card.
- NAP looks at the two top cards, says ”one on top, one on the bottom", puts one card on top and one on the bottom and the second on top, and then draws two cards.

The first one can legitimately be seen as a weird way to resolve Anticipate up until the moment where the second card hits the hand instead of going to the bottom of the library. The second one is not. I am fine ruling the first one as HCE and the second one as GRV, not because AP allowed NAP to draw 2 (he didn't), but because clearly, something wrong was already happening when NAP put the cards back in the library, and AP could have stopped it at this point.

It occurs to me that this thread had mostly become a dialogue between the two of us, while we do not actually disagree that much. Maybe we should continue this conversation by mail?

If AP doesn't immediately realise the specifics of what Antiipate says to do, or if NAP does this block of actions very quickly, there isn't really an opportunity for AP to stop NAP from drawing the extra card.

"…does so without his or her opponent’s permission.“ means that if AP confirms the card draw, the infraction isn't HCE. This is different from AP just having an opportunity to stop NAP from drawing an extra card. It doesn't matter how much NAP communicates if AP doesn't communicate back.

Originally posted by Florian Horn:

It occurs to me that this thread had mostly become a dialogue between the two of us, while we do not actually disagree that much. Maybe we should continue this conversation by mail?

I was wondering the same thing, actually, but there are a couple other people in the thread that have posed questions which I don't feel have been definitively answered. I've also been wondering if someone else will come along with a third opinion.

Regardless, I've been enjoying our dialogue, as it's forced me to look at the details of this scenario much more closely than I otherwise would have!

My thought is that we should give AP the choice of which card is drawn off of anticipate, and return the excess card to the appropriate place in the library (the bottom is the scry puts 0 or one card on the bottom, shuffled away if the scry puts two on the bottom).

i.e.
(1)
scry both to top: AP chooses the extra card, and NAP puts it on the bottom of the library, where it should have gone. Optionally, we could put the card currently on top of the library to the bottom as well, but since a random card on top is the same as a random card on the bottom (leaving a new random card on top), I think it's unnecessary to make that movement. I would not allow NAP to look at the top card and have the option to keep it instead of another card in their hand, as this substantially offsets the downside of committing HCE.

(2)
scry one to top, one to bottom: AP chooses the extra card, and NAP puts it on the bottom of the library, along with the card that was scried there previously.

(3)
scry both to bottom: AP chooses the extra card, which is shuffled into the random portion of the library, preserving the bottom two cards that were scried there.

(4)
there is disagreement about where the cards were scried to: I could see an argument for either treating the entire library as unknown (again, excluding any previous, legal scries) or applying fix (3) and treating the bottom two (random) cards as the known ones. Of these two, I like treating the entire library as unknown better. So AP chooses the extra card, which is shuffled into the unknown portion of the library, treating the bottom cards as known only if they were previously known.