Nissa controls Leyline of Sanctity and Ansgar casts Thought Scour targeting her. After milling the first card (a card was not yet drawn either) they both realize that targeting Nissa was illegal. If the judge should decide to perform a backup, where should the card end up (strictly by the IPG)? In the random part of the library, or on top of the library? Does the fix change in the following cases? In all situations the judge decided to do a backup.

• Both cards were already milled
  
• Both cards were already milled and the order is no longer known (because the players disagree for instance)
  
• Both cards were already milled and the order is no longer known and the card was already drawn
  

Originally posted by IPG:

To perform a backup, each individual action since the point of the error is reversed, starting with the most recent ones and working backwards. Every action must be reversed; no parts of the sequence should be omitted or reordered.

The question arose because people couldn't agree whether reversing means to put the card back on top of the library, or into the random part of the library (assuming it was a random card before).

Moved to Competitive REL.

I will assume that the top 2 cards of Nissa's library were previously unknown; if so, and we back up to a point where suddenly Nissa (and Ansgar) know the top two cards, we did a back up that lead to a substantially worse situation than just leaving everything as it is; if you don't want to shuffle the library, you shouldn't back up at all and just continue resolving the Thought Scour, no matter how far it already went resolving (mill 1, mill 2, mill 2 + draw).

I hope this doesn't seem like the best solution to you either; backing up seems easy enough to do, and will lead to a substantially better game state than not backing up. If you take the IPG extremely literally, and put the cards back on top of the library, all of a sudden extra cards have become known. Thankfully, we have a great fix for that, as described in Looking at Extra Cards: shuffle the random portion of the library.

Note that this L@EC was caused by the fix for the GRV, so we should not penalise the players for it. I agree that the IPG doesn't explicitly spell out that we sometimes use the L@EC fix after applying another fix, and this may need some clarification in a future edition. But when you look at the philosophy instead of the literal words of the IPG, we want to back up to the situation prior to the error: no cards in Nissa's graveyard, and no cards known in Nissa's library (and also na land tapped for Ansgar and Thoughtscour back in the hand).

If cards on top of Nissa's library were known (due to scry, fateseal, Brainstorm, Jace TMS), we obviously don't randomise when backing up.

Originally posted by Dustin De Leeuw:

But when you look at the philosophy instead of the literal words of the IPG, we want to back up to the situation prior to the error: no cards in Nissa's graveyard, and no cards known in Nissa's library (and also na land tapped for Ansgar and Thoughtscour back in the hand).

I agree with everything you just said and also agree that this is the best fix (which is not literally supported by the IPG right now). But, the philosophy part of the IPG also says:

Originally posted by IPG:

If a judge makes a ruling that is consistent with quoted text, then the complaints of a player shift from accusation of unfairness by the judge to accusations of unfair policy. Deviations from these procedures may raise accusations against the judge from the player(s) involved, or from those who hear about it. These procedures do not, and should not, take into account the game being played, the current situation that the game is in, or who will benefit strategically from the procedure associated with a penalty. While it is tempting to try to “fix” game situations, the danger of missing a subtle detail or showing favoritism to a player (even unintentionally) makes it a bad idea.

And I am pretty sure that, the “we apply other fixes, after fixing the root cause, in order to fix things that happed thereafter and get the final fix we want” falls under this category as long as it is not explicitly part of the IPG and should be handled with extreme care.

Edited Konrad Eibl (July 18, 2017 11:59:07 AM)

Originally posted by Konrad Eibl:

I hope this doesn't seem like the best solution to you either; backing up seems easy enough to do, and will lead to a substantially better game state than not backing up. If you take the IPG extremely literally, and put the cards back on top of the library, all of a sudden extra cards have become known. Thankfully, we have a great fix for that, as described in Looking at Extra Cards: shuffle the random portion of the library.

I totally agree with you Dustin. :)

By the way, I think that IPG now supports this “random is random” feature in the “Backup” paragraph.
I rememebr that Toby, in the Aether Revolt update bulletin, wrote this sentence:
"As part of backing up, cards that subsequently become known to both players can be shuffled in, as they can be fully identified. Only cards that one player knows the identity of (usually because they drew them) are returned to the original location.“

The only problem is that this ”feature“ has been added in the ”reversing a shuffle“ part of the paragraph, so it seems to be doable only if we have to undo one or more shuffle while performing a backup.
” Shuffles are reversed by a single shuffle of the random portion of the
library after the rest of the backup is complete. A card that became legally known to only one
player after the error was committed is not considered random and is returned to the appropriate
location after the shuffle has been completed. "

This can create some confusion in my opinion: if we have to deal with this kind of situations, the document seems to says something different from the actual philosophy.

Originally posted by Dustin De Leeuw:

Note that this L@EC was caused by the fix for the GRV, so we should not penalise the players for it.

I agree with the second part - when the L@EC remedy is an appropriate part of the fix, then we should include it, along with the rest of the backup; the IPG also agrees that we only assess the more serious of the simultaneous (or “same root cause”) infractions.

The first Toby-quote that Jacopo shared reinforces this. I've often tried to find a good way to word it, and get it into the IPG (still working on it)(not there yet) - but yes, you can - and sometimes, should - incorporate the L@EC remedy with a normal GRV backup.

However, I disagree about the cause being the fix - the error itself (targeting and resolving illegally) was the cause of the card(s) becoming seen, by all players.

Quick aside, here: this would not be what we're calling a “simple backup”. This would still be a regular backup, and all the cautions and restrictions outlined in the IPG would apply to this backup.

I also want to emphasize something else important:

Originally posted by Konrad Eibl:

should be handled with extreme care

While I agree with Dustin's proposed remedy, and the infraction is definitely GPE:GRV, and Failure to Maintain Game State for the opponent, Konrad is correct that we should always be aware of the need for extreme care when considering any backup - even a simple one. (Granted, the whole idea of a “simple” backup is that it shouldn't require any special care, but reality is special…)

d:^D