[/cets_callout_box]
It’s a small update this time around, so I’m going to use this space to talk about a few additional things at the bottom that aren’t policy changes, but are useful guidance. In the meantime, updates!
We’ve Been Here Before
By far the biggest change in terms of text volume is a new Communication section in the Magic Tournament Rules dedicated to loops. They’d come up again because of Teferi, and I was talking to Magic Rules Manager Eli Shiffrin about how we could make them better.
“It’s really easy. We just want to do this, and this, and this.”
“That all makes sense, but it’s not formal logic, so it won’t work in the Comprehensive Rules.”
“It’d work in the MTR. Loops are just a kind of shortcut, and we don’t have to be so strict there.”
“Go for it.”
It turns out that expressing how to handle loops is much harder to do in formal Magic rules than it is if you can wave your hands and use plain English. So there’s a new section in the MTR that supersedes the simpler ones in the CR when there’s a tournament.
From a game perspective, not a lot changes. Loops that span multiple turns are addressed, and judges have flexibility to identify that there’s a loop even if there’s irrelevant things changing. Everything is expressed in terms of asking the players questions and then applying the results, and hopefully it’s much easier to follow.
We’ve Been Here Before
By far the biggest change in terms of text volume is a new Communication section in the Magic Tournament Rules dedicated to loops. They’d come up again because of Teferi, and I was talking to Magic Rules Manager Eli Shiffrin about how we could make them better.
“It’s really easy. We just want to do this, and this, and this.”
“That all makes sense, but it’s not formal logic, so it won’t work in the Comprehensive Rules.”
“It’d work in the MTR. Loops are just a kind of shortcut, and we don’t have to be so strict there.”
“Go for it.”
It turns out that expressing how to handle loops is much harder to do in formal Magic rules than it is if you can wave your hands and use plain English. So there’s a new section in the MTR that supersedes the simpler ones in the CR when there’s a tournament.
From a game perspective, not a lot changes. Loops that span multiple turns are addressed, and judges have flexibility to identify that there’s a loop even if there’s irrelevant things changing. Everything is expressed in terms of asking the players questions and then applying the results, and hopefully it’s much easier to follow.
We’ve Been Here BeM19 Draft
Normally, we tell players to take out the token and the land from booster packs. But M19 introduces a wrinkle where there’s a mix of basic and draftable non-basic lands (and a checklist card) in the same slot! So we’ve called out that slot as being special; for most sets you still remove the basics, but if draftable cards appear in the basic land slot, keep it all in. That way, everyone will have the same number of cards to draft.
This does have the downside of making the checklist draftable, even though it isn’t playable. Plus, there’s only one card listed on it, so make sure to remind new players that drafting the checklist doesn’t mean that they have a Nicol Bolas in their draft pool and they should just pass it along until the end.
Upkeep, Draw, Untap
We’ve all seen it: a player draws a card for the turn, then untaps, or casts a spell with a stack of land that’s still tapped from the previous turn. It’s all obvious, and usually handled through Out-of-Order-Sequencing or a bit of handwaving. Now the IPG provides a stronger endorsement that this is OK in the form of a partial fix that untaps the things that were supposed to be untapped this turn. We don’t expect judges to need this much as OoOS is a better option in general, but it’s here when you need it.
Quick Hits
- We’d established that penalties reset at the Day 1 cut even with a round to go last update, but what REL was that “not really Day 2” round being played at? This update addresses that (Professional).
- Any time you change a decklist after the tournament has begun, a Game Loss should be issued. That includes when you can’t find replacement cards for marked ones. That’s explicitly called out as an upgrade now.
- Normally, if you have too many copies of a card that’s in both your deck and sideboard, it gets upgraded. But, if you discover it while all the copies (both deck and sideboard) are in the library, it’s not upgraded.
- CPV previously looked for an opponent to take an action before it could be called. Clearly choosing to not take an action also qualifies (though I think the threshold there is a little higher).
- We cleaned up the language about issuing Game Losses in both the introduction and Decklist sections. This is mostly nonfunctional, though it does emphasize that it’s pretty rare to wait on a Decklist penalty nowadays. It used to be a lot more relevant when we counted lists more aggressively.
Triggers
One thing that’s come up a bit recently is a debate about Detrimental Triggers and defining them, and I want try to give a little insight into the philosophy here. In particular, I want to address the idea that self-mill triggers are always detrimental, even though some cards – when evaluated in full card context – use them for beneficial reasons.
There isn’t an iron-clad edict from Wizards that self-mill must always be considered detrimental. It has the same weight as anything else that’s regarded that way. It’s like “When this enters the battlefield, sacrifice a creature”. That’s a detrimental ability, but I think we’d all agree that “When you sacrifice a creature, you win the game” on the same card would make it not a detrimental trigger.
However, when the trigger changes happened, we had Innistrad, and we got a lot more pushback on self-mill than all the other mechanics, since it was often a bonus in that format. That meant there was a lot more “no seriously, self-mill is detrimental” posts than on most other mechanics, and a lot more pushback on arguable cases. That has led to a perception that milling was more stringent than the others. It does default to detrimental, but it’s no different from other triggers. You need to look at the overall context of the card (*not* the game state) to make that determination, and this is true for all abilities, not just self-mill.
Slow Play
Finally, a nudge.
Slow Play is one of the toughest calls a judge has to make. It’s a pure judgment call, and the player affected will almost always argue, talking about how they have lots of time on the clock, or this is a crucial turn, or that they didn’t take that long. This is partly because the perception of time when you’re tanking is different, and what may seem like not a lot of time to think is a huge amount of clock time.
How aggressive judges have been about calling slow play has ebbed and flowed over the years. We’ll make an effort to be more aggressive, that’ll happen for a bit, and then we’ll all collectively slide back a bit. It’s time for another one of those efforts. I’m here to tell you today that you (and I include myself here) aren’t calling slow play enough and need to make a conscious effort to do so.
I can’t give you a neat trick that will make your slow play calling game better; what works for one person won’t work at all for someone else. I personally find that if I’m watching a game and start to get bored, that’s a big warning sign, but that’ll vary a lot depending how much you play Magic and how good you are at analyzing options. Whatever your personal heuristic, it takes effort and practice to make it part of your judging repertoire, and a willingness to interrupt a player’s train of thought to get the game moving again. It’s not easy, but we’re not doing it enough right now, and everyone should make a conscious effort over the next few months to be aware of the clock and how long players are taking.
Thanks
Thanks to everyone who offered suggestions this time around. In particular, a shout-out to Isaac King, who’s been providing great suggestions for improved wordings. If you see a sentence change that you think is much clearer, you probably have him to thank for it.
Enjoy M19!
If we can can accept loops explained in plain English and apply them without formal logic, does this eliminate the problem with Four Horsemen’s “loop?”
No. Horsemen is still nondeterministic.
This seems silly to me. It’s only non-deterministic in formal terms. In plain English we can easily describe what is otherwise practically a loop with certainty that is constantly approaching 100%. I recognize that this loop is not literally 100% determined but if we’re using practical thought, plain English, and setting formal logic aside, why is 99.99% repeating not sufficient?
It’s the fact that even on an infinite timeline it is statistically possible to never get the ordering that you want. MASSIVELY UNLIKELY but still possible, and possible is good enough.There are ways to “fix” the deck to not be nondeterministic in this way now with some new cards though.
I’m so happy to see that loops are moving to the MTR. That section in the rules has always been rather vague and left certain situations uncovered, and it’s more of a tournament issue anyway.
Can you clarify the reasoning behind the new exception to the Deck Problem upgrade? Why do we not upgrade if all copies are in the random portion of the library, but we do upgrade if some copies are there and some are still in the sideboard?
It’s not that some are still in the sideboard, it’s that if you have too many in your deck, but they’re all still in your deck (you notice while fetching, for example) we can fix it without problems.
The example for this upgrade is still saying that ” there are three copies of
Shock in the library”, which would now be not an upgrade.
I’ll make the example clearer next time around. Implication is that one of them is discovered as it leaves the library.
As usual, my document diffs can be found here:
MTR: https://draftable.com/compare/HfqHNQBRFWZK
IPG: https://draftable.com/compare/rXayGEkHvOGb
Does “involved in maintaining the loop” mean “making an active decision to maintain the loop”?
You had previously posted that Duskwatch Recruiter’s “look at top 3, …, put on bottom in any order” plus infinite activations was legal to loop to stack a deck, given that the deck was not a multiple of 3 in size. That appears to have changed as it “requires a decision tree.” Is this interpretation correct?
I think this is still OK (if annoying, and they’d better do it fast!). You’re able to give the number of iterations needed, and you’re not taking different actions depending on the outcome of something in the loop, which is what the decision-tree restriction is there to prevent. It’s just “I do this thing X times and here’s the end state”
Was it taken into consideration for the updated ‘forgot to desideboard’ upgrade path that the potential for abuse is fairly huge in eternal formats such as legacy and vintage where a player can use a brainstorm and fetchland to get rid of the sideboard card, thus giving him an out to not call a judge ?
Investigations seem hard to handle when the player calls a judge right when he’s fetching through his library and claims he just noticed it, even if he just played a brainstorm or it isn’t the first time he looks through his library.
It was considered. It’s not great, but at minimum the player hasn’t been able to benefit from the excess of that card, even if they did find a way to avoid the game loss.
What if only one player is involved in maintaining a loop across turns (as in this situation: https://twitter.com/CubeApril/status/958037833033371649 )? Do they still have to end the loop eventually? (Meaning that in the linked example the Lantern player has to stop using Academy Ruins and lose the game.)
That’s correct.
I can’t be the only one who thinks the day 1 wipe is a mistake right? This just gives the go ahead to cheaters to make “mistakes” again on day 2 because day 1s “mistakes” can’t catch up to them.
It hasn’t been a problem in the time it’s been legal. The only change here is the REL the final round of Day 1 is run at.
It looks as though the new loop rules bar the common pattern of “repeat the loop until {X},” even when the maximum number of iterations required is finite. For example, use Devoted Druid and Vizier of Remedies to generate infinite mana, then use Duskwatch Recruiter to dig for Walking Ballista. Generating infinite mana is fine, but the problem comes in the second part. I can’t predict how far in my library I will need to go to find Walking Ballista, and the new rules explicitly forbid conditional actions.
Is this the desired outcome? Should we require players to look through their libraries 3 cards at a time until they find the Walking Ballista?
Repeat the loop until X has technically never been allowed, but here you can just repeat for your deck to get it. There’s a fixed number of iterations.
> “If two or more players are involved in maintaining a loop across turns…”
What constitutes involvement?
If I have an empty library and a Lich’s Mastery and my opponent has me in the Teferi lock, does me saying “go” count as involvement?
Involvement means doing something to keep the loop going. Saying “go” is just a turn mechanic, not something we can say not to do any more.
Regarding loops. this post:
https://blogs.magicjudges.org/blog/2015/09/11/magic-judge-monthly-august-2015/
Says that a library can be sorted by “looping” petals of insight 410,758 times. Is that still a legal loop? (It seems to me that any simple description of sorting the library in a loop will involve conditionals so it’s not deterministic.)
Supposing it’s still a legal loop, if an omnitell player says “I cast petals of insight 410,758 times to sort my library” can that player’s opponent break the loop at – say – the 205,379 th casting?
Your last question is why this is still a horrible scenario which, fortunately, doesn’t come up much. It’s currently legal, maybe it shouldn’t be. I’ll ask.
I don’t quite understand how to evaluate the “detrimentalness” of a trigger. From what I’ve been told this far we should only look at the trigger itself, not any other parts of the card (and of course not the game state). But the last sentence in the part about triggers here reads like that is not the case. Is the Nyx Weaver trigger supposed to be detrimental? What about a card with a linked ability like Bomat Courier?
You can take the context of the card outside of the game state into consideration. Don’t go overboard, though.
“However, when the trigger changes happened, we had Innistrad, and we got a lot more pushback on self-mill than all the other mechanics, since it was often a bonus in that format.”
And in every format since then that had such cards (self-mill triggers, no ability on the card itself that made them beneficial):
M15 – Necromancer’s Assistant: a minor GB graveyard theme with Undergrowth Scavenger as the biggest payoff
Theros – Forsaken Drifters: a major GB graveyard theme with Nemesis of Mortals, Graverobber Spider, Font of Return
Origins – Screeching Skaab (reprint): spell mastery, as well as cards like Deep-Sea Terror, Skaab Goliath, Undead Servant
Tarkir – Sultai Skullkeeper: delve
Shadows – Laboratory Brute, Crow of Dark Tidings, Wailing Ghoul: delirium
In practice it’s always a beneficial ability.
Sooo … Does the above statement regarding draft procedure with M19 supersede what I read in this week’s Cranial Insertion (http://www.cranialinsertion.com/article/2713) or is it just woefully incomplete? I mean, this (possibly/likely?) being handled differently between GP Day 2s and everywhere else is quite an enormous deal, right?
Their last sentence looks wrong to me (there’ll be non-basics in GP drafts), but the rest lines up.
Toby, another scenario came into play when we discussed those changes in an article posted on the Brazilian community. The short history:
Ananias casts Brain Freeze with 10 billion storm copies (they used Naru Meha and Release to the Wind to do so) targetting Normand, but Normand has an Emrakul, the Aeons Torn in his library.
Our fellow judges question if it is possible to shortcut to some reachable number of storm copies iterations since randomizing the library 3 or 10000 times doesn’t make much of a difference, but also question that relying on probability and decision trees can’t be shortcuted.There is also some judges who propose a draw since we can’t really do so many triggers and shuffles in a timely manner.
What are we supposed to do if a scenario like this is presented?
No amount of rules is going to fix this; it’s not even a looping question, really, since it’s a game-fixed number. It doesn’t rely on probability or decision trees.
I think your best bet is to work backwards. Assuming 60 card decks, there can’t be more than 20 Brain Freezes on the stack after the n-1th shuffle. So execute 20 iterations and leave it at that.
The new loop section references a thing that I’ve maintained for a while needs to be more clearly spelled out. In 4.4 it describes that a non-deterministic loop must be stopped “If at any point during the process a previous game state (or one identical in all relevant ways) has been reached”
An actual rules definition of “all relevant ways” is desired, or also a rules definition of what exactly “Advancing the game state” means in a practical sense. The way we treat a game state looking at it and the way the -game- treats a game state differ wildly, so reconciling those would be nice.
There is no way to list all the relevant ways, so we give judges the discretion here. It’s usually not that hard to figure out.