How would you solve the following situation during an event run at competitive REL? You are the Head Judge, and suspect no cheating.

Player A puts Primeval Titan into play with Through the Breach, and attacks. They remember both triggered abilities. They pass the turn after combat, forgetting the delayed triggered ability from Through the Breach. Player B untaps, draws a card, then taps two lands and says “bolt the titan twice”, showing two Lightning Bolts from their hand. At this point, Player A realizes their mistake and calls for a judge.

I give Player B the option of whether to resolve the trigger the next time a player would get priority or at the beginning of the next phase. Player B had plenty of opportunity to notice the trigger, I'm not going to let him take back his Lightning Bolts.

I tend to operate under the belief that players can interrupt themselves as they do things. (No, intentional manipulations are still a Very Bad Thing <tm>.) So if a player has tapped a few Mountains and is holding two cards in hand, it's clear that they interrupted themselves. The original post says “at this point” and “showing two Lightning Bolts,” so I'd consider that self-interrupted. A player self-interrupting can reverse and modify their actions, so I'd let the player stop the incomplete action and resolve the problem.

Now if the player had played more correctly and resolved the first Bolt (or at least clearly indicated the casting was done) and was moving on to the second Bolt, then we're not interrupting that first one, and the ruling is therefore different. (Interruptions are only for the currently involved actions.)

Imagine a moment where the player says, “Bolt the Titan,” looks at his or her hand for a Bolt and then stops to say, “Actually, that's dead,” without revealing anything. Common sense says we don't force the guy to show a Bolt and then tap a Mountain. I find that giving players the maximum flexibility (but not infinite flexibility) is a line that is easy to understand, easy to teach, and pretty well accepted among the players.

If he hadn't finished the process of casting the Lightning Bolt, I'd agree. But having tapped two Mountains and revealed the two Bolts makes me pretty sure that he finished casting them. If Lightning Bolt were a sorcery, my answer would be different, but being an instant he could have just cast them both before either one resolved.

The entirety of the interruption is based on one word: “showing.” If a player is in the process of doing a thing, the thing is not does. And since we're talking about two discrete game process (cast Bolt, cast Bolt), muddling the two together clearly indicates non-clarity. To be precise, no player reveals both spells they are playing because two spell are never played simultaneously (even with Fuse).

If the scenario separated the two cast or if the scenario clearly indicated that the people had actually completed the muddled process (“after putting both Bolt on the table” or “after putting both Bolt in the bin”), then yes, the argument would be that he is done with casting them. Those versions of events are not what was posted, and I suspect Håkon Gulbrandsen's original intent is to understand where the line in between one interpretation and the other.

History Time (a.k.a. Remembering the Mistakes of the Past):
Let's imagine walking up to a player who has stopped a line of play while holding a pair of cards and telling him or her that they can't stop. What happens to the world of Magic? More casual players get hurt, angry, frustrated, and taken advantage of. More sharkish players turn into hawks (flying, finned, deathbirds now… how lovely), pounce on their opponents for the tiniest things, and debate (but not argue, they're hawksharks, so they get to the line but do go over) with judges every point on minutia. And judge? We become little Napoleons running around squeezing the humanity out of the game because there is no room for the reality of paper Magic played by people, just robot rules enforcement.

It turns out, the Judge Program went through this era before. It used to be that touching a land meant tapping a land, and once a land was tapped it could not be rewound. This is the same time when putting a spell in the graveyard meant that the spell had resolved, and if you were searching for a thing in your library, you had to stop since the spell had resolved (“the Harrow rule,” if you will). This was a universally despised time.

From that time a few ideas developed in the judge world:
1) Players are human
Players are not perfect, impartial robots who process the rules like so many lines of code. They are imprecise in their operation, but they also can function in an imprecise world.
2) Judges are human
Attempting to be as impartially robotic as possible leads to processes and results which violate the humanity of both players and judges, breaks the notion of what humans think of as fair, and severely breaks the notion of what humans think of as fun.

From that realization, some new practices developed:
1A) Tapping and untapping a land
This action neither generates infinite mana nor causes the world to end. It's ok to let players fidget with lands and other stuff while trying to sort things out. We prefer they be clear (“not blocking yet” or “just thinking for a moment”), and we do not allow fishing for reaction (tap, wait, look at opponent, tap other).
1B) Catching oneself
A player taps W3. They play Wrath of God (costs WW2). He or she immediately stops and untaps an Adarkar Wastes and taps a Plains. No blood, no foul. The same is true for announcing spell X but revealing Y (or saying “Island” but putting a Plains into play–come one, who doesn't play Island first???).
1C) Players aren't master criminals
No, we do not need to DQ this guy for a 56-card decklist. No, cheating is not the default belief but one that requires investigation. Most players most of the time are normal(-ish) people; sometimes they make mistakes.
2A) Judges have judgment
When in doubt and for 95% of the time, the situation is simple, and the documents should be followed. But that doesn't mean judges are just human versions of MTGO nor that we should attempt to be. Sometimes part of our judgment will include “what makes sense here ~from the player's perspective~?” If we make a ruling and ~everyone~ is confused and upset, then we should probably look at the situation again.
2B) Judges who are too loose or too tight cause problems
Like many things in life, moderation is key. A strict judge alienates players by sucking the fun out of the game. A loose judge alienates players by making the environment too unpredictable and open for exploitation.
2C) Being pedantic is always wrong
I'll admit, I originally wrote “almost always” wrong, but then I realized the pedantics out there would latch on to the “almost” (because they are, you know, pedantic) and then continue to be awful. Why such a harsh stance about pedantic folk? First, let me be clear that being precise is a Good Thing <tm>; there’s a fundamental difference between precision and pedantry. Someone who is pedantic has “slavish attention to rules.” The rules are all that matters, and such a view deletes humanity and judgment from the process of playing Magic. A judge who goes around citing rules but has no concept of how humans interact or what they reasonably expect in a scenario drives a hard wedge between the players and the game. They are the Dementors of fun-sucking.

The reason I almost wrote “almost” is that pedantry and precision often come to the same conclusion. However, precise judges also include realities of the physical game rather than only the words in a terrifyingly large rulebook. When precision approaches pedantry, the Magic Community and the Judge Program are best served if we redouble our consideration for the fact we’re dealing with humans in a real world. Then, whatever our ruling may be, we can address the players as people rather than robotic devices for shifting of cardboard.

Strictly speaking, Isaac is correct, in that what should happen is the delayed trigger goes on the stack and player B loses their bolts. However, I feel like this advantages Player A too much and am not comfortable with this fix, so I would deviate by rewinding to the point at which the Lightning Bolts were not yet cast and putting the trigger on the stack then. I recognize this is a deviation, but I think it is correct in this situation.

@John: Noteworthy is that it was player A, not B, who noticed the trigger, while it was player B who had the Lightning Bolts. Player B did not self-interrupt, rather, it was player A who interrupted him. I don't think your argument works in this case, although I do agree that Isaac's solution, while I believe it is correct, is too draconian.

Edited Lyle Waldman (Nov. 12, 2016 04:03:29 PM)

I can see how my initial post could be worded more precisely. For clarification, Player B held out the Lightning Bolts towards the Primeval Titan, and was awaiting a response when Player A realized he had missed the delayed triggered ability.

Also, as Lyle commented above, Player A (the controller of the Titan) was the one who realized the error.

(I know it is confusing with “Player A” and “Player B”, but AP/NAP didn't work out in this situation)

The “and awaiting response” makes the modified scenario very clear on whether the player (B) was interrupting himself during an action. Thanks!

It's worth pointing out that according to the standard shortcuts defined by the MTR, Player B has not cast two bolts; rather, he has proposed the shortcut of casting a bolt, having it resolve, and then casting the second bolt. Does that change anyone's answer at all?