In a recent blog post by Kevin Desprez, it was mentioned that if the players forget about Kalitas's replacement effect, you should not move the creature card into exile even though it appears to fall under one of the listed partial fixes.

Originally posted by Kevin Desprez:

When one infraction leads to several different consequences on the game state, it becomes arguable that fixing part of the infraction brings the game state any closer to what it should have been. Therefore, unless the 2/2 token was actually put on the battlefield, you should not partial fix this.

I'm having some trouble understanding the reasoning behind this decision. To me it seems that the two actions of making a zombie token and exiling the creature card are unrelated, aside from the fact that they are both caused by the same ability at the same time. In my opinion, fixing one of them still leads to a “more correct” game state then fixing neither.

Let's consider the situation if Kalitas instead had two separate abilities: “If a nontoken creature an opponent controls would die, instead exile that card” and “whenever a creature card is put into exile from the battlefield, put a 2/2 black Zombie creature token onto the battlefield”. In this case, would we still take into account an “unrelated” trigger that should have triggered when determining whether to apply the partial fix? If so, how far do we go with this? Do we take into account triggered abilities of other objects that should have triggered?

Edited Isaac King (June 21, 2016 01:51:57 AM)

I don't know the reasoning that caused this particular decision, but it seems to me that resolving only half of an action is dangerous for the integrity of the game. We can estimate that the consequences here are not dire, but an important part of the philosophy of the IPG is to have consistent fixes, rather than ad hoc “best way to repair a game”.

Originally posted by Isaac King:

To me it seems that the two actions of making a zombie token and exiling the creature card are unrelated, aside from the fact that they are both caused by the same ability at the same time.

That's quite a important relation, though.

Originally posted by Isaac King:

Let's consider the situation if Kalitas instead had two separate abilities: “If a nontoken creature an opponent controls would die, instead exile that card” and “whenever a creature card is put into exile from the battlefield, put a 2/2 black Zombie creature token onto the battlefield”.

The difference between this version of Kalitas and the Traitor of Ghet is more than just wording. In particular, it triggers whenever a creature is exiled, not only when it is exiled because of Kalitas' replacement effect (and there is no easy way to mimic the Traitor of Ghet's ability with a trigger). Notice that, if one forgot to exile a creature with this version, the fix would most likely result in exiling the card and putting a Zombie in play (assuming that Kalitas is still on the battlefield when the fix is made).

Originally posted by Isaac King:

In this case, would we still take into account an “unrelated” trigger that should have triggered when determining whether to apply the partial fix?

No.

Originally posted by Isaac King:

Do we take into account triggered abilities of other objects that should have triggered?

If the other object is still here, it triggers now. If it is not, it doesn't trigger.

Edited Florian Horn (June 21, 2016 09:24:05 AM)

Originally posted by Florian Horn:

I don't know the reasoning that caused this particular decision, but it seems to me that resolving only half of an action is dangerous for the integrity of the game. We can estimate that the consequences here are not dire, but an important part of the philosophy of the IPG is to have consistent fixes, rather than ad hoc “best way to repair a game”.

I agree that we should have consistent fixes, but it seems to me that it is inconsistent to not apply the partial fix here. There isn't anything inherently disruptive about exiling the card now, so why avoid it just because we should have gotten a zombie? Why do we partial fix a Rest in Peace, but not Kalitas?

Originally posted by Florian Horn:

That's quite an important relation, though.

Why?

Originally posted by Florian Horn:

The difference between this version of Kalitas and the Traitor of Ghet is more than just wording. In particular, it triggers whenever a creature is exiled, not only when it is exiled because of Kalitas' replacement effect (and there is no easy way to mimic the Traitor of Ghet's ability with a trigger).

That's not really relevant to the question. I could have used the trigger condition “Whenever a creature card an opponent controls is put into exile as a result of Kalitas's ability”, but I didn't see the point in making it more complex. The point of this example is to have a card that functions effectively the same way, but uses a separate triggered ability instead of the same replacement effect.

Originally posted by Florian Horn:

Notice that, if one forgot to exile a creature with this version, the fix would most likely result in exiling the card and putting a Zombie in play (assuming that Kalitas is still on the battlefield when the fix is made).

I don't believe it would. We move the creature card from the graveyard to exile, not from the battlefield to exile. When we apply a partial fix, we are actually moving the card now, we aren't going back to the point of error and fixing it then.

Edited Isaac King (June 21, 2016 12:13:10 PM)

Originally posted by Isaac King:

I agree that we should have consistent fixes, but it seems to me that it is inconsistent to not apply the partial fix here. There isn't anything inherently disruptive about exiling the card now, so why avoid it just because we should have gotten a zombie?

The issue is not with Kalitas but with every other Magic card that could ever be printed.

I'm sure there exist, or could exist, cards where partially applying a partial fix to a complex ability causes serious damage to the game state. Policy can't possibly define what those cards are or how to tell the difference between them and Kalitas. So, we set up policy to avoid forcing us to break the game–and having a slightly less optimal fix for Kalitas is the cost of that decision.