Hello everyone!

I'd like to discuss with you about the Knowledge Pool Scenario that you can find here.

Alice is playing against Nicky in round one of a Standard PPTQ. Alice controls two untapped Plains and two untapped Mountains. Alice taps both mountains to cast Veteran Motorist and announces she will scry 2. Nicky confirms. Alice looks at the top two cards of her library and places both cards on the bottom of her library. A spectator watching the match notices Alice forgot to tap one of her plains when she cast Veteran Motorist and calls for a judge.

The official answer is that we can decide to do a backup or leave everything as it is.
If we choose for a backup, the two scried cards are put back on the top of the library in a random order and the Motorist is put back in his owner's hand.

I believe that here we have two errors, not just one: a wrong mana payment and a wrong “Scry” action. The second one depends on the first one: Alice looked at those cards only because she misplayed the creature spell.
Consequently I think that we can treat this mistake as a GRV+LEC: we can fix the second by shuffling the two cards back in the library and than we can backup the GRV putting the Motorist in Alice's hand.
Considering that the root cause is the GRV, we can issue only one Warning.

I found an article that speaks exaclty about a similar situation:

"Albert plays a Temple of Malady, however he already played a Forest earlier in the turn. He scrys and leaves the card on top.
Here we have two infractions, a GRV of playing multiple lands in a turn and a LEC when he scrys. When we’re fixing this situation we need to apply fixes for both of these offences. Clearly the land needs to go back to hand (a rewind, the fix for GRV) and we need to shuffle the random portion of the library (the fix for LEC). However, seeing the extra card was clearly caused by the original infraction of playing an extra land. Therefore we give a warning for GRV and not for LEC. This explains why we can apply an LEC fix when we’re giving a GRV warning. Intuitively it’s the correct way to fix a GRV which has resulted in seeing cards they should not, but it may not be obvious how that follows from the IPG."

I honestly can't find differences between this scenario and the KP one, and I don't think that the philosophy has changed in a way that it doesn't allow us anymore to deal with these problems in this way.

What do you think about that?
What am I missing?
Thanks in advance for your answers and your help! :D

Edited Jacopo Strati (March 28, 2017 12:02:21 PM)

I should note that the article referenced by Jacopo dates back to August 2015. Since that time there has been quite a few updates in policy.

Don't have the time to elaborate further (being at work), but I suggest you check Toby's blog archive, you may find your answer there ? :)

Hi Jacopo - I want to make sure I'm understanding what you're asking here.

You seem to be indicating that the opinion of the KP team that a backup isn't warranted in the scenario conflicts with the article where a backup was allowed due to what is arguably a very similar situation. Is that correct? I believe the article you linked more represents justifying and handling of multiple infractions, not so much of ‘because it was done in this case, it should be done in all cases’. As noted above, it's also from a couple of years ago - this doesn't, by default, invalidate the information, but it is possible that there have been subsequent policy blog articles which touch on this subject more recently and with a different stance (perhaps even including reasons why to lean towards a particular decision regarding whether or not to backup).

In my opinion, I agree with the Knowledge Base team's belief that a backup isn't warranted, specifically because both backing up and not backing up seem to offer some level of advantage to the player that played incorrectly, with not backing up seeming more ‘fair’ for both players involved.

In the scenario outlined by the Knowledge Base scenario, not backing up gives the active player knowledge of what's on the bottom of their library and leaves a creature in play with incorrect mana tapped. Knowing the location of those two cards increases the odds that the top card or cards in their library are potentially cards they may be looking for, so they've gained an advantage of knowledge, but the cost is that the board state is not backed up and they're committed to the play they previously chose.

Allowing for a backup would basically give the active player another chance to find what they're looking for (since a backup in this case would allow for the scry'd cards to be placed back on top of the library, randomizing the unknown portion of the library, returning Veteran Motorist to AP's hand and untapping mana). While this removes the knowledge of what's on the bottom of the library, the player has the choice to replay the Veteran Motorist and scry, or hold up - a decision that could have been made based on seeing either of the cards that were seen, which means that it would potentially be made with knowledge from looking at extra cards (though, admittedly, I find it moderately unlikely that seeing the cards that were scry'ed would change the decision to play Veteran Motorist, the potential is there).

Edited Bernie Hoelschen (March 28, 2017 12:24:58 PM)

Originally posted by David Lachance-Poitras:

I should note that the article referenced by Jacopo dates back to August 2015. Since that time there has been quite a few updates in policy.

Don't have the time to elaborate further (being at work), but I suggest you check Toby's blog archive, you may find your answer there ? :)

I checked all the policy update bullettins Toby made, but sadly I was unsuccessful in finding references to this change. :(

HI Bernie :D

Originally posted by Bernie Hoelschen:

You seem to be indicating that the opinion of the KP team that a backup isn't warranted in the scenario conflicts with the article where a backup was allowed due to what is arguably a very similar situation. Is that correct? I believe the article you linked more represents justifying and handling of multiple infractions, not so much of ‘because it was done in this case, it should be done in all cases’. As noted above, it's also from a couple of years ago - this doesn't, by default, invalidate the information, but it is possible that there have been subsequent policy blog articles which touch on this subject more recently and with a different stance (perhaps even including reasons why to lean towards a particular decision regarding whether or not to backup).

Yes, I'm having difficulties in understanding where the differences are between the KP scenario and the Scryland one.
What I see is that, in both situations, the offender gained informations he shouldn't have had due to a previous GRV (in KP scenario we have a misplayed creature spell, in the article one we have land that shouldn't be on the battlefield).
Why the Scry effect is considerd to be a result of a GRV in one scenario and not in the other one?
Policies seems not to be changed on this detail. Maybe I'm missing something?

Originally posted by Bernie Hoelschen:

In the scenario outlined by the Knowledge Base scenario, not backing up gives the active player knowledge of what's on the bottom of their library and leaves a creature in play with incorrect mana tapped. Knowing the location of those two cards increases the odds that the top card or cards in their library are potentially cards they may be looking for, so they've gained an advantage of knowledge, but the cost is that the board state is not backed up and they're committed to the play they previously chose.

Allowing for a backup would basically give the active player another chance to find what they're looking for (since a backup in this case would allow for the scry'd cards to be placed back on top of the library, randomizing the unknown portion of the library, returning Veteran Motorist to AP's hand and untapping mana). While this removes the knowledge of what's on the bottom of the library, the player has the choice to replay the Veteran Motorist and scry, or hold up - a decision that could have been made based on seeing either of the cards that were seen, which means that it would potentially be made with knowledge from looking at extra cards (though, admittedly, I find it moderately unlikely that seeing the cards that were scry'ed would change the decision to play Veteran Motorist, the potential is there).

If there's no way to shuffle those two cards back in the library, then I totally agree with the KP team official answer. :)

Edited Jacopo Strati (March 28, 2017 12:37:38 PM)

Originally posted by Jacopo Strati:

HI Bernie :D

Yes, I'm having difficulties in understanding where the differences are between the KP scenario and the Scryland one. What I see is that, in both situations, the offender gained informations he shouldn't have had due to a previous GRV (in KP scenario we have a misplayed creature spell, in the article one we have land that shouldn't be on the battlefield). Why the Scry effect is considered to be a result of a GRV in one scenario and not in the other one? Policies seems not to be changed on this detail. Maybe I'm missing something?

If there's no way to shuffle those two cards back in the library, then I totally agree with the KP team official answer. :)

In regards to the first part - the scenarios aren't different because of the cause of the LEC - I don't think anyone sees them as different based on that, but I could absolutely be wrong. The scenarios are different because of where the board states end up. In both cases, the LEC is caused by a GRV. In the KP scenario, the backup requires untapping mana and returning a creature to hand that could be played again this turn. In the article's scenario, the scry land can't be played again this turn without some effect that allows for playing an additional land or lands.

I think I'm starting to see where you're coming from in regards to the second question. It seems as though you're wanting to perform the LEC fix for the scry, but NOT back up the GRV itself (leave the creature on the field with the incorrect mana tapped and randomize the library for the LEC). The trouble with this is, what if other cards on the bottom of the library prior to the scry were known?

The goal, as I understand it, is to ‘fix’ the situation with the least disruption possible, while at the same time mitigating potential situations that cause an advantage. As such, I feel like leaving the game state as is appears to be the most consistent action here.

Originally posted by Bernie Hoelschen:

It seems as though you're wanting to perform the LEC fix for the scry, but NOT back up the GRV itself

No, this is not my intent. Maybe I wasn't able to explain me in the best way possible. :D
I was saying that I consider those two scried cards a LEC and a consequence of the previous mistake (paying the Motorist with the wrong mana - GRV). Alice wasn't entitled to see those cards because the effect that granted her this right was misplayed. So she is actually looking at extra cards, imho.
We have, then, two infractions with the same root cause: GRV + LEC. We are going to issue just one Warning but, to fix the problem, we can apply both LEC and GRV fixes.
So, going backward, we can:
1) fix the LEC by shuffling those two cards in the deck
2) backup the misplayed Motorist, untapping the lands and putting it back in Alice's hand.

This is what I was suggesting, which is supported by the article I shared… if it's still applicable.
:)

I want to chime in here.

Honestly, I think the multiple infractions article, as quoted, is incorrect based on the current understanding of the IPG fixes. I can't find any indication that the pertinent section on backing up has been changed since the article would have been written, but applying an additional fix in that manner seems out-of-keeping with how other parts of the IPG handle such things. Additionally, the AIPG has the following note in the section on backing up:

If they are legally drawn after the error, they are legally known by one of the players. Since exactly which card was drawn is probably lost, we are going to put back a card at random, and we don’t want to shuffle them away. This policy has an odd interaction with scry (which could also be legally known) which is likely unintended and will hopefully be fixed in the next policy update.

bolded for emphasis.

If this were a situation in which cards were drawn instead of scried, it would be clear that we don't apply the HCE fix in addition to backing up for this type of infraction. It doesn't make much sense to apply a different standard for a philosophically similar infraction. In both cases (drawing vs. scrying) the cards become know to one player as the result of legally resolving a spell or ability that has been illegally played. In both situations, the root cause of the infraction is the GRV, and so in neither case would any other infraction apply on top of the GRV. This means that we should restrict ourselves to additional remedies called for in the GRV remedies.

Further, philosophically the judge community seems to be in a place where we're not inclined to give players a “free shuffle” in situations where they aren't actually seeing cards they aren't supposed to see. This thread, for instance, treats an HCE as such. In short, if we are going to back up through or otherwise “fix” a scry or card draw, we want to try to have the player end up looking through the same cards that they looked at the first time, to minimize any advantage that might have been gained from the application of the backup (like getting a scry, shuffle, scry to dig for an answer).

All that to say, maybe someone should look into writing an update to the multiple infractions article.

Originally posted by Andrew Keeler:

If they are legally drawn after the error, they are legally known by one of the players. Since exactly which card was drawn is probably lost, we are going to put back a card at random, and we don’t want to shuffle them away. This policy has an odd interaction with scry (which could also be legally known) which is likely unintended and will hopefully be fixed in the next policy update.

It say that the interaction with Scry is unintended. This paragraph obliges us to preserve also scried cards that could be easly identified by both players and, then, shuffled back into library. That‘s why it‘s described as an odd interaction, I think.
Otherwise, doing a backup after a Scry effect would be always impossible. :\
If I cast Preordain with R and I realize it when I'm already Scrying 2, do I have to finish to resolve it anyway?
Isn't this abusable?
Drawing a card in this way is quite different imho: a drawn card is lost among the others in the hand, so there's no way to know which one is the extra one. Having a GRV before the error anyway (made with public information), doesn't allow us to handle it as an HCE: the oppont has some responsability in this mistake, so letting him look at offender‘s hand and letting him choose a card to shuffle away would be too disruptive. So we pick up a random card and we put it back onto the library.
My two cents. :)

Edited Jacopo Strati (March 29, 2017 04:33:33 AM)

I wrote the article quoted. In my opinion, the state before the error was that the library was random, the lands were untapped and the creature was in hand. If we rewind, we should rewind to that state, not to the state where the creature is back in hand and the player knows the top 2 cards of their library.

Now, I certainly think that policy is unclear around issues like this (which is why I wrote the article), but I will note that Toby reviewed and approved it before it was posted.

You say there's an advantage where they get to scry, don't like them, shuffle and scry again. I think that's using knowledge post-hoc of what they'll find. The alternative scenario is that two cards they want are shuffled away where you're letting them keep them on top, not play the motorist, then scry again after they've drawn them. I don't think we can have it both ways here. In any case, at the point they would have to decide to cheat by mis-tapping then the player doesn't know whether it will be good or bad for him to have them shuffled away.

Thanks for responding Matthew. Your thinking makes a lot of sense. I'm beginning to recognize why this interaction is called out as awkward in the AIPG, and I hope that the future update and clarification happen soon.

Originally posted by Matthew Johnson:

I wrote the article quoted. In my opinion, the state before the error was that the library was random, the lands were untapped and the creature was in hand. If we rewind, we should rewind to that state, not to the state where the creature is back in hand and the player knows the top 2 cards of their library.

Now, I certainly think that policy is unclear around issues like this (which is why I wrote the article), but I will note that Toby reviewed and approved it before it was posted.

You say there's an advantage where they get to scry, don't like them, shuffle and scry again. I think that's using knowledge post-hoc of what they'll find. The alternative scenario is that two cards they want are shuffled away where you're letting them keep them on top, not play the motorist, then scry again after they've drawn them. I don't think we can have it both ways here. In any case, at the point they would have to decide to cheat by mis-tapping then the player doesn't know whether it will be good or bad for him to have them shuffled away.

Thanks for your answer Matthew. :)