Discussion between murgatroid99 and Zags

Zags

00:00

but casting first is at best no different than casting second and at worst a disadvantage

murgatroid99

OK, different one: active player has a bear. non-active player has a tapped bear. Both players are at 2 life. Both players can cast either murder or Gods Willing with the mana they have up. The player with priority loses

no, I guess that doesn't quite work

Zags

there's a reason i posted the question

priority ends up being really subtle in application

murgatroid99

It's a little hard to think of this on the fly

Zags

I've been thinking about it all morning

I'd love more examples of where it matters

It's got some far reaching implications into timing rules for games in general

quick break; back in 20

murgatroid99

00:28

OK, I thought of one: active player has a 4/2 black creatures, defending player has a 2/2. Both are at 2 life. Active player can cast either Murder or activate Heartwood Shard's ability. Non-active player can cast either Fog or Gods Willing.

After declare attackers: AP casts murder=>NAP casts gods willing. AP activates Shards=>NAP casts fog. AP does nothing=>NAP casts fog=>AP casts murder.

If NAP had priority: NAP casts Fog => AP casts Murder, NAP casts Gods Willing => AP casts Murder, NAP does nothing => (choice tree for AP, but removing the last choice).

Actually, in the case of NAP casts Gods Willing, AP could take either action and win

Alternatively, AP does nothing => NAP passes => declare blockers => AP activates shards => NAP casts fog

By passing priority, the active player would potentially be giving up Murder as an option

Zags

01:44

let's assume this is taking place in the declare blockers step

so a double pass goes straight to combat damage and both creatures die

from AP perspective, fog cancels trample, protection cancels murder

murgatroid99

I specifically said declare attackers for a reason

Murder only lets AP win if they cast it before declare blockers

Zags

hmm

yes

murgatroid99

And neither player knows what the other has

Zags

but if NAP plays fog, it only saves them for a turn

but is guaranteed to save them for a turn

murgatroid99

No, because AP's only creature is tapped, and NAP has a 2/2 attacker

Zags

01:51

ah, hence murder as a response

so basically, each has a card that could beat a card of the other player

and whoever plays first loses

murgatroid99

right, but NAP's cards are more responsive

If both pass, and don't play anything, NAP has the opportunity to untap and draw a win card

after both creatures die to combat

Zags

well, from NAP's perspective, murder beats fog and trample beats protection

murgatroid99

Well, yes, as long as they're responding

Zags

the double pass doesn't exactly nullify murder as an option though

the attacker will still be blocked but the defender will have lost their creature

murgatroid99

Once it goes to declare blockers, Murder no longer allows the attacking creature to directly damage the defending player. It removes it as a win option

Zags

01:54

only for the turn

murgatroid99

And then, we're back to the situation where NAP has the opportunity to untap and do something

Zags

sure

but in order for the double pass

both players have to decide to pass

murgatroid99

It's in NAP's interest to pass

Zags

either way (AP first or NAP first), the AP has the chance to act

murgatroid99

If AP passes

Zags

01:55

but in that case, if priority worked with NAP first, NAP would pass

because then NAP could respond to AP acting, or get the desired double pass

murgatroid99

the point is not that the outcome would be different if NAP had priority (though I did that case). The point is that it's in AP's interest to act in the declare attackers step with insufficient information because they have priority

Zags

but i don't think it's the priority that matters

it's in AP's interest to act in the declare attackers step because waiting till the declare blockers step removes one of their options

meanwhile, NAP is sitting there holding reactions and consequently wants to use them as late as possible

murgatroid99

And because they can't let NAP take an action first without losing that option

AP doesn't know what NAP has in their hand.

Zags

but how does that interact with AP needing to act first?

we have to assume something about what AP knows

either AP knows the full situation, in which case he doesn't act because any action would get countered

AP knows one of the responses (lets say it was revealed by a tutor), in which case he thinks playing one of his two cards is safe and doesn't need to wait

AP knows about neither of the responses, in which case what he chose to do would depend a lot on his play style

or AP knows about the possibility of the responses, in which case we have to construct a rather complicated probability model

murgatroid99

I'm talking about the case where AP doesn't know. That's the simplest model for hidden information. In that case there is a best move: Murder

Zags

02:03

Sure

but in that case, NAP's best move would be to pass during the declare attackers step

(if NAP went first)

(obviously respond to murder with protection)

so if NAP goes first, NAP passes

then AP plays murder

NAP responds with pro black

if AP goes first, AP plays murder

NAP responds with pro black

same course of events

murgatroid99

I'll revise my earlier statement: the important thing here is that AP has to make a choice because the priority rules don't let them wait around for NAP to do something; they force AP to be proactive with incomplete information

Zags

Let's simplify the example

AP: 2/2 trample w/ Giant Growth in hand

NAP: 2/2 w/ Lightning Bolt in hand

AP has attacked, NAP has blocked

it's declare blockers

2 life vs 2 life

nope

jk

lightning bolt to the head

murgatroid99

OK, I can agree that priority is entirely irrelevant in that case, because NAP could have Bolted 5 steps ago

Or 10, more realistically

Zags

use Electrostatic Bolt

can only target creatures

if AP is first to go

your claim is that they would use growth for the win

(or at least try to)

murgatroid99

of course

Zags

02:11

and be foiled by bolt in response

murgatroid99

yes

Zags

but if NAP were first to go, they would try to use bolt for the win and be foiled by growth

murgatroid99

yes

Zags

it's a compelling application of the priority rules

but surprisingly hard to create a situation where it happens

murgatroid99

Actually, let's look at a recently active question that illustrates the simplest example

2

Q: If both players are killed at the same time, do they both die?

I play Magic: The Gathering and I just had a really strange game with my opponent. We were both about to die (both sides with 2 life) when we both set down instants that damaged each other (at the same time). I played the red instant Shock, and she played the red instant Staggershock. It was my t...

magic-the-gathering

Effectively, each player has low life and can cast Lightning Bolt.

Zags

02:15

unlikely situation

why wouldn't the player that drew bolt first use it right away?

murgatroid99

I was wondering about that myself

It is possible, though

Zags

Skyscribing, i presume

murgatroid99

Player 1 has 3 life and Lightning Bolt. Player 2 has 4 life, activates Yawgmoth's Bargin once, draws a Lightning Bolt

Or that

Zags

we're still in "rare and bizarre situations" territory

murgatroid99

@Zags remember, in aggregate, rare situations happen all the time

Zags

02:18

yea, but basically in all of this situations, whoever jumps first loses

in order for priority to matter, we need to have both players equally eager to jump

and even so, if someone decides to skip priority, no one else would complain

ex. if NAP erroneously said, "Ha ha, I lightning bolt you for the win"

AP would not be like "You can't do that yet; I haven't passed priority"

murgatroid99

This is the point where I point out the choice asymmetry created by priority rules, and then we restart this conversation from an earlier point.

So, I think I'm going to be done

You've made your point, and I've made mine

And I have other things to do

Zags

it's interesting to think about

i'm not sure if I have a solid formalism for it yet

Transcript for

Discussion between murgatroid99 and Zags