Ever since the “It's Plain to See” Knowledge Pool scenario was released in March, I have had questions about how we resolve GPE - GRVs for incorrectly paying costs for spells and abilities. I'd like to discuss the philosophy behind the current ruling, and boldly suggest a possible partial fix.

The problem arises whenever someone makes an error in paying a cost – mana cost, tapping creatures, sacrificing permanents, etc. – and continues to resolve the spell or ability without noticing the error. If such a spell or ability involved something like a scry, card draw, or other game action that would be difficult to simply backup, this issue becomes much more clear.

Imagine this scenario:
Aaron has three untapped lands: a Taiga, a Polluted Delta, and a basic island. Aaron casts Brainstorm, tapping his Taiga to pay for the spell. He draws three cards and puts two back on the top of his library. He then sacrifices his Polluted Delta and picks up his library to search for a mountain. At this point, he notices his error and calls for a judge. The judge rules out cheating, issues the GRV penalty, and then begins to apply a fix.

That fix is pretty ugly, isn't it? We would have to:
a. undo the Polluted Delta activation (recreating the library from before he picked it up)
b. undo the casting of Brainstorm (…seriously, gross.)
OR
c. leave the gamestate as is.

Ew.

My question is this: Why do we have to backup the casting of a spell or activation of an ability when the casting/activation was entirely legal except for the paying of costs?

Wouldn't it be so much easier if we could just untap Aaron's Savannah and tap his basic island? No backup through a shuffle effect plus recreating the Brainstorm set and then replacing it into the library. AND, we don't have to leave the game-state as it is. It seems so intuitive…but nowhere does the IPG allow us to do this.

I'd like to propose a partial fix for this error:
If a player cast a spell or ability without paying its costs correctly, the player could legally cast the spell or activate the ability, and the error was noticed in a reasonable amount of time, undo all costs that were paid (untapping any lands, returning sacrificed creatures to the battlefield, restoring life that was paid, etc.), and then have the player pay the costs for the spell or ability correctly.

This partial fix would not apply to situations where
a) the player could not legally cast the card (AP casts a sorcery on NAP's turn)
b) the player could not pay for the cost of the spell or ability when it was cast (Example A: Ariel casts Cryptic Command with only two islands and two mountains untapped… Example B: Allen is at 1 life and casts Gitaxian Probe with no untapped lands*)
c) the error is not a GRV (Aaron -above- casts and resolves Brainstorm before tapping any lands to pay for the spell before his opponent has time to catch his error)
d) so much time has passed that applying this fix would be too disruptive to the game (Aaron -above- casts Brainstorm as above, fetches for a mountain, and then casts Izzet Charm to draw two and discard two, before the error is noticed)

What do you think? Do you think this partial fix could work? How could we revise it to make the wording more clear or address any anticipated issues with it?

I look forward to your ideas!

*This example brings up a great issue. What if Allen were at 2 life and cast Gitaxian Probe with no untapped lands? Yes, he can legally cast the spell, so the partial fix would apply here, but doing so would cause him to lose the game! With this partial fix, we would still hold Allen to the casting of the spell, resulting in him losing the game, just like we do if a player tries to fetch when he/she is at 1 life.

Thank you for posting this! Like Riki said yesterday, I have been looking forward to it as well!

As for whether or not I believe the partial fix works, I think it makes sense and is well thought out. As laid out in the current Knowledge Pool question, (attempting to activate Cryptbreaker's first ability (tap three zombies you control, draw a card) but instead tapping 1B and Cryptbreaker (payment of the second abiliity), caught in a reasonable time (immediately after it happened)), what is the detriment to backing this up to the point where the announcement of the ability is on the stack when there is a legal option to pay for the ability?

For your exceptions, if the player controlled two zombies and attempted to activate the ability, then it's HCE (Drew a card that shouldn't have been able to have been drawn), based off a GRV (paying the wrong cost to activate an ability). Things get a little muddy at that point - the root cause of the HCE is GRV so GRV is the penalty, but would the IPG allow for resolution of having drawn the card in error (i.e. the Thoughtseize fix)?

I think an exception might have to include a clause regarding when a block of actions are taken simultaneously, such as, player taps three lands of various mana to cost two spells, but causes a GRV with the casting of one of the spells, resulting in a back up - which land(s) are you untapping? Ex: Legacy, your opponent has Ethersworn Canonist, and playing quickly, tapping Underground Sea and two volcanic islands, you cast Delver of Secrets followed up by Young Pyromancer. Opponent calls a judge, you've broken a game rule by casting the second spell. Which one gets backed up, and which land(s) get untapped? In this particular case, if Delver hit the table first, then Young Pyromancer would be the spell that couldn't be cast. Which two lands do you untap though?

An exception might have to be something like this (although, in theory, I think this might already be present in current precedent).

e) Anytime that backing up would be ambiguous (i.e. where the specific unique resource used to cast a spell or activate an ability cannot be identified), the game state should be left as is.

Edited Bernie Hoelschen (June 15, 2017 08:29:18 AM)

+1 to being interested in how this conversation plays out :)

I am, generally, a fan of this fix because it is typically the fix most players would employ if a judge weren't involved.

The first issue that pops into my head would be removing the opportunity for the opponent to react once the correct cost is paid. Consider your above scenario - what if the opponent let brainstorm resolve because they saw an untapped island and were thinking about the possibility of countermagic?

We use the new partial fix, and now NAP is grumpy because “if they had tapped properly and I knew they had no blue sources I would have made a different choice”.

I'm not sure if that window for advantage in the fix you suggest is larger or smaller than current policy, but I'm still uncomfortable with removing the opportunity for the opponent to consider a new play once the error has been corrected.

Originally posted by Jon Goud:

We use the new partial fix, and now NAP is grumpy because “if they had tapped properly and I knew they had no blue sources I would have made a different choice”.

I'm not sure if that window for advantage in the fix you suggest is larger or smaller than current policy, but I'm still uncomfortable with removing the opportunity for the opponent to consider a new play once the error has been corrected.

Personally, I'm not sure if this consideration is any more unreasonable than some of the other partial fixes we already have. Suppose AP casts brainstorm as their only card in hand, draws three, and then only puts one back. NAP sees AP has 2 cards in hand and tries to play around Force of Will. That consideration shouldn't prevent us from applying the correct partial fix of putting an extra card back now, even though NAP did make decisions based on the incorrect number of cards in hand. Or instead consider virtually every time the “declare blocker order” partial fix is invoked, when the order has suddenly become relevant because one player made a play based on a particular understanding of what would happen later in combat.

I'm reminded of the principle from Embracing the Rewind that backups aren't a question of fairness, but of moving the game from an incorrect state to a correct one. It strikes me that the same consideration would be true of partial fixes.

April, first off, thank you very much for posting this follow up discussion.

Your Gitaxian Probe example hits close to my primary criticism of this proposed partial fix and that is pain lands. If a player cast Wrath of God with WUUU and their only other white source is an Adarkar Wastes, your fix means tapping the Wastes for mana and 1 damage. In a small percentages of scenarios this results in the player's immediate death. There are also situations where the result is a kind of delayed death, say if the error in noticed on the subsequent turn after a burn spell puts the player to 1. I also think it is unclear what your line of “too disruptive” would be for exception d) in these cases where a pain land is involved.

Life totals are one of the most important pieces of information in making game decisions, which is why there is the emphasis on keeping track and communicating properly. Decisions like how to attack and block can vary wildly based on one life point difference, which is why we don't want partial fixes that can result in such changes.

I'm also not seeing how to deal with non-mana abilities, fetch lands and Deathrite Shaman being the most played examples I can think of. Would you force players to use these abilities to satisfy the partial fix? Not sure if that fits the criteria of a) not being able to legally cast.

The vast majority of cases are ones where your proposed partial fix doesn't matter much. They tap a different land, and nothing else changes. There are a few cases where a player can now do something different because of the different land. And then there are the disaster scenarios. There are certainly disaster scenarios the other way, but I tend to fall back on “You did this to yourselves.” The players both had a chance to catch the error. They are responsible for the game state together. Introducing this type of partial fix creates weird scenarios where our intervention and decisions can start to really affect games, like the pain land scenarios. I like to minimize those interactions.

Isnt it possible to ask the player if he would cast the spell again, so we could just use the partial fix?
Most of the time the player will cast the spell again anyway and it would help reduce the disadvantages of some fixes(like putting random card back from hand and such).

Besides if that is really the only issue we can find, we easily could add a “with only minor disruption”-clause to it like we have for card in incorrect zones.
I like this fix as well. My biggest issue is that the response of the opponent could be different because e.g. counter mana was open, but wouldn't be with the correct mana. However I believe that this is only minor and the suggested partial fix gives a better result compared to all those cases we can't backup

Originally posted by Riki Hayashi:

April, first off, thank you very much for posting this follow up discussion.

I'm also not seeing how to deal with non-mana abilities, fetch lands and Deathrite Shaman being the most played examples I can think of. Would you force players to use these abilities to satisfy the partial fix? Not sure if that fits the criteria of a) not being able to legally cast.

The vast majority of cases are ones where your proposed partial fix doesn't matter much. They tap a different land, and nothing else changes. There are a few cases where a player can now do something different because of the different land. And then there are the disaster scenarios. There are certainly disaster scenarios the other way, but I tend to fall back on “You did this to yourselves.” The players both had a chance to catch the error. They are responsible for the game state together. Introducing this type of partial fix creates weird scenarios where our intervention and decisions can start to really affect games, like the pain land scenarios. I like to minimize those interactions.

I think that Riki raises some significant concerns with possible problems in allowing a partial fix of this nature, but I want to specifically address the question of non-mana ability payments.

Specifically, we could just limit this fix to mana abilities. This doesn't resolve your concern regarding players accidentally killing themselves with pain lands, but it does alleviate a large swath of potential issues and questions.

In thinking about it while typing, we can even tailor our partial fix to be more narrow and restrict it to just mana abilities that don't affect life totals and that alleviates the vast majority of cases where there are problems.

Unfortunately it also removed a lot of utility from this partial fix and makes me question whether such a carve out is worth the space in the IPG… It certainly is a thorny issue, but I think the biggest question related to “where's the line” is how narrowly do we have to tailor the partial fix in order to avoid regularly blowing ourselves up, and then whether such a narrow partial fix is worth the complexity cost to the IPG. The main other question I see that needs to be resolved is what is our tolerance for the potential of disaster scenarios? Are we willing to tolerate a 1% implosion rate in exchange for a much cleaner fix the other 99% of the time? How about 5%?

I think this discussion quickly becomes a broader philosophical issue of what are we willing to sacrifice in order to cover the majority of scenarios more cleanly. A discussion that I think is well worth having.

Originally posted by Bryan Henning:

In thinking about it while typing, we can even tailor our partial fix to be more narrow and restrict it to just mana abilities that don't affect life totals and that alleviates the vast majority of cases where there are problems.

But then what about cards like City of Brass, where the life loss isn't part of the mana ability? Or Chromatic Star, when you're close to running out of cards in your library? Adding on more and more caveats to a partial fix has diminishing returns after a certain point. Having the fix only apply to mana abilities is clean and simple, but anything further than that starts to bog down the wording and leads to errors and inconsistencies when the fix is performed.

Having a line similar to the one in the zone-change partial fix

and it can be moved with only minor disruption to the current state of the game

would allow the judge to make a judgement call based on the exact situation, and in my opinion would be a much better option.

Originally posted by Isaac King:

Having a line similar to the one in the zone-change partial fix

and it can be moved with only minor disruption to the current state of the game

would allow the judge to make a judgement call based on the exact situation, and in my opinion would be a much better option.

While following this discussion over the weekend, I, too, came to the conclusion that this would be a very simple way to address many of the above concerns. So I've revised the proposed partial fix to read:

If a player cast a spell or ability without paying its costs correctly, the player could legally cast the spell or activate the ability, the error was noticed in a reasonable amount of time, and the fix can be applied with only minor disruption to the current game state, undo all costs that were paid (untapping any lands, returning sacrificed creatures to the battlefield, restoring life that was paid, etc.), and then have the player pay the costs for the spell or ability correctly.

Activating abilities such as those of City of Brass, Scalding Tarn, Llanowar Wastes, or Deathrite Shaman would preclude the use of this partial fix because such actions would be too disruptive to the game state. Sacrificing creatures and paying life would also no longer fall under this partial fix. If these costs would be necessary for the player to cast/activate the spell or ability, the partial fix does not apply and a simple backup would be considered.

I recognize that this does significantly narrow the applicability of the partial fix, but when it would apply (such as in the “It's Plain to See” scenario) it would be very helpful. What do you think?

Originally posted by April Keeler:

Activating abilities such as those of City of Brass, Scalding Tarn, Llanowar Wastes, or Deathrite Shaman would preclude the use of this partial fix because such actions would be too disruptive to the game state. Sacrificing creatures and paying life would also no longer fall under this partial fix.

In my opinion, this is where we reach the core of the problem. I would love to have such a partial fix as I see all the benefits, but this is just too subjective. Either we would have to explicitly codify what counts as “too disruptive”, which is obviously impossible and impractical, or we get a huge variety in personal opinions and hence a lack of consistency between tournaments and judges.

Originally posted by Markus Dietrich:

My biggest issue is that the response of the opponent could be different because e.g. counter mana was open, but wouldn't be with the correct mana. However I believe that this is only minor and the suggested partial fix gives a better result compared to all those cases we can't backup

+1 to this. Most of the time it won't matter, but when it does, it matters a lot, at least in the perception of the player who was not at fault.

That being said, I'm very happy with this thread and all great suggestions made here, and I would love to see more options for partial fixing common problems with minor disruption!

Re: subjectivity: We already have guidelines and frequent discussions on
what's too disruptive for a backup; why not the same for this?

2017-06-19 12:50 GMT-04:00 Dustin De Leeuw <

Originally posted by Dominik Chłobowski:

Re: subjectivity: We already have guidelines and frequent discussions on
what's too disruptive for a backup; why not the same for this?

Part of the appeal of default fixes is that, at large events, floor judges can just perform them without involving another judge, because they're very simple. “You forgot to draw? Okay, draw a card. Game is fixed.” They're also not optional - floor judges don't need to determine whether or not they'd be problematic because the IPG says to just do them if they would fix the problem on their own.

The recent addition of “missed zone change” goes a little past this line, but comes up pretty rarely, and I'd recommend against using it as a model - none of the other partial fixes come with a disruptiveness check, and I'd rather see us avoid adding new ones..

For mana paid, there are a lot of implications, and they're not always obvious. If a player casts a Lightning Bolt with a Plains, and their only red source, a Steam Vents, is also their third blue source, we're locking out of playing Cryptic Command or even representing it. If we rewind, we're giving the player the ability to choose which is more important, and to send a signal to his or her opponent, rather than making it ourselves.

Codifying this as a partial fix means it takes precedence over evaluating a rewind, and most cases of incorrect mana can more appropriately be fixed with a rewind, especially if a player catches them immediately.

I think a lot of judges want this at first, because it seems intuitive and it's exactly what you'd do if you were just playing and there was no judge available. The problem, like several people have mentioned, is that once you make it a partial fix, you have to decide where to draw the line on it. Can you force people to tap out when they want to leave mana open? Can you force them to take damage that puts them in burn range? Can you force them to sacrifice a creature to Ashnod's Altar?

While on the subject of potential fixes though, I'd love to see something to the effect of “If a player announces an action and then doesn't take the correct action, do a simple backup to the point of the announcement if possible.”
This would give us a way to fix obvious simple errors without either making the game state worse than it already was or holding a player to something that they clearly did not intend to do.

Originally posted by Dustin De Leeuw:

In my opinion, this is where we reach the core of the problem. I would love to have such a partial fix as I see all the benefits, but this is just too subjective. Either we would have to explicitly codify what counts as “too disruptive”, which is obviously impossible and impractical, or we get a huge variety in personal opinions and hence a lack of consistency between tournaments and judges.

I agree on that statement, but I'd like to push the discussion a bit more far: valuing consistency is great when we're in an acquisition phase, and it has its own advantages, but once we'll move into a solidification phase, I'd like us to value more empowering judges to deliver the best ruling available to a precise situation, even if that means being a bit loser on consistency. That would mean a less rigid (and codified) IPG, and more room for this kind of subjective fixes.

- Emilien