Facts:

• It's a card type for cards that can be n your library and in your hand.

Reasonable assumptions:

• It's a permanent type.
• It can be cast and will have a mana cost.

Wild guessing:

• It could be like a cross between planeswalkers and dungeons:
  Planeswalker-like interactivity: Opponents can attack/damage battles to prevent you from "winning" the battle.
  Dungeon-like benefit: You can do the same to advance through "victories"/"acts" and are rewarded with spell-like effects or maybe static effects (King of the Hill-style). Invest resources to gain incremental benefits.
• Probably once you are winning strongly enough, you might get some sort of "ultimate" reward, sacrificing the battle.

What I don't expect:

• Battles probably benefit you, if you "are victorious", but don't benefit your opponent for doing the same. Rather they can interact with the battle to keep you from gettign your final reward.

For example:

Battle for Supplies
Battle
(After you draw, if you have enough conquest counters, collect your reward and advance the battle marker. Sacrifice after final reward.)
[4] You gain 5 life.
[6] Draw a card.
[10] Return up to two target nonland permanent cards with mana value 3 or less from your graveyard to the battlefield.
[2]

RULES NOT WRITTEN ON THE CARD:

• The starting number of conquest counters on this battle is 2 (see bottom "[2]")
• You can "attack" any battle and decide whether your damage is used to add or remove conquest counters (to make multiplayer more interesting - otherwise you add to your own battle and remove from opponent's/other player's battle)
• number of counters added/removed could be based on either total "damage" dealt by your attackers, or by number of unblocked attackers (The current numbers are in my mind balanced around number of counters added equal to power/damage)
• Any opponent may block your creatures attacking your battle.
• You may block any creature of an opponent attacking your battle.
• As a SBA a battle with no conquest counters is buried.
• conquest counters are like loyalty, but the "battle phase" is tracked with a marker along the edge of the card (think chapters on a Saga, but using a "venture marker"-like game object that is not a counter)

Notes:

• This design is a bit complex since it uses both counters and a single marker for the battle phase, but this marker only has to move along a single route unlike a dungeon's "venture marker", so I consider it realistic to just move it along the side of the card.
• You never go back to a previous battle phase, so this is really just a Saga that you have to put some effort into.
• I use a Saga-like "after you draw"-condition to allow opponent's input before the first reward, but it's also likely that "end of turn"-trigger" is a better choice, especially if - like in this case, you have to put in an attack to even get the first reward.
• You get only one reward/advance only one battle phase per turn, so even if you make a big attack/battle contribution on the first turn, your opponent may still prevent you from advancing with a counter push.

Why this needs to be a new card type IMO:

• My idea is based around having a planeswalker-like interaction during combat. You don't just e. g. make an enchantment subtype that makes enchantments attackable

Why this might be in ONE:

• Works nicely with proliferate.

My .02/fan creation. I'm here all week.

I don't think there's any design space for a new permanent type. Anything that goes on the battlefield can be represented by an artifact or enchantment, with a new subtype if they want to hide some of the backend rules text

Nonpermanent spells are even worse. I thought there might be room for a multi-turn sorcery, like Hearthstone has messed around with a little bit, but that's better done as a Saga

I think the only option is for a new card that goes in the command zone. Battle as a new dungeon-like mechanic that starts in your deck could be interesting. It seems likely that battle evolved out of the cut skirmish mechanic from War of the Spark

I think it will be representing individual planes fighting the phyrexians. You and the opponent will be on one side or another of the war perhaps having to do a sidequest similar to a mix of an enchantments saga and dungeon mechanic to "win" the battle.

I think a cute idea is a "reverse planeswalker". Battles just a permanents on the battlefield and you can "attack" a batte you control to add counters to it, while your opponent can block.

Then at certain thresholds you get effects similar to Sagas; you just have to go through combat to advance the battle.

It's more simple than my above suggestion.

An interesting new tidbit about battles is the new arts we've seen that's probably for them. It's horizontally oriented, and makes me think of planar cards. Planes themselves are just really similar to global symmetric enchantments though, so no idea what to make of that.

Battles in MOM (some, most, all?) are confirmed to be double faced. So yeah, I already dislike them.

I guess this implies the back side is the "reward" for completing some goal on the front side.

My only hunch is that battles are something that changes combat, and that you (and possibly the opponent) assign creatures to, like Raging River.

We now know what battles are: Invasion of Zendikar

You play the battle, get an etb effect, and assign an opponent to defend. Any other player can then attack the battle. When its defeated, it transforms and you get the spell on the back

The file name has a .jpg suffix, but the MIME type is image/webp. So, I installed the webp package, but then, it cannot read from stdin, so I need to store a copy of the file and then convert it; however, output to stdout does work, so storing a copy of the converted file is not necessary.

The reminder text does not explain much, but it gives some idea of its working. I will hope to see the actual rules later, since then it would be explained better, probably. (I can see why they made a new type, though.) One thing that I notice is that it has a subtype, and the reminder text mentions the subtype.

We should switch to getting images from Scryfall instead of Gatherer, then I could just link to previewed cards :)

Reverse planeswalkers it is. This is very similar to a card type I created: You could "attack" the permanent you yourself control for a reward.

I always was unhappy that allowing your opponents to block made the card type weak in multiplayer. They solved that by introducing a "protector". That's nice.

There will be different variants of battles it seems. I'm looking forward to read up on which aspects of the card we got are inherent to all battles and which parts are tied to Siege specifically.

Now I found some more information about it, in the March of the Machine Mechanics article.

The use defense counters. That seems to be similar to loyalty counters, but it is a different kind of counters, being removed due to a different type.

It says "every player except a battle's protector may attack it". Does that include players on the same team as the protector? How does this work with rules dealing with "defending player"? (Currently the rules are that the active player decides which player is the defending player as a turn-based action during the begin combat step (possibly restricted to only the previous or next player on a different team, depending on the specific variant being played), unless all opponents are automatically defending players. How does that interact with battles? Does the defending player need to be the battle's protector to be able to attack it, or is that somehow not the case?)

If a Siege's controller changes, who is its protector? If the controller changes control during combat, is it removed from combat and no longer attacked? I would think so, if it is like existing rules, which says that a permanent whose controller changes is removed from combat, so I would guess that it would apply here, too.

It says "when the last defense counter is removed from a Siege battle, the battle is defeated and a triggered ability triggers". That suggests a (presumably intrinsic) triggered ability, which could be countered (or otherwise fail to resolve) and leave it with no counters, unless there is some other rule having to do with that somehow. (One alternative would be a state trigger, which would allow it to retrigger in such a circumstance.)

Can a Siege creature attack itself? (And, therefore if the blocking creature has positive power and the attacker does not have first strike (or the blocker does have first strike), it will damage the battle anyways, whether or not it is blocked)

My assumption on the trigger on defeat is that the battle will be put into the graveyard as an SBA, but the intrinsic triggered ability tracks that card.

It is already out there that the default behavior of battles without defense counters is to go to the graveyard, Sieges just add to that.

Yes, Tabak said that there's a state-based action that puts battles with no counters and no triggered ability on the stack into their owner's graveyard, just like sagas

He also tweeted that battles that are also creatures can't attack or block

Does anyone else find that second point annoying? Every other permanent type can attack if transformed into a creature...

No other permanent would be able to attack itself. Yeah, planeswalkers can block for themselves, but battles already mess with the logistics of the combat phase through the protector mechanic.

I understand why they did it, but for cases like the one you pointed out, I would still rather the ruling be that they can't attack themselves, not that they can't attack at all.

The state-based action mentioned makes sense. (My unofficial "generalized ongoing" rule would suppress it like it also does with Sagas, schemes, dungeons, and phenomena.)

I also find the second point annoying, too. I think that it would be cleaner (i.e. not needing so many exceptions) if that rule that battles that are also creatures cannot attack or block is omitted, and just allow it to attack and block, and even to attack itself. (It would be a creature so you can declare an attack with it, and it is a Siege that you are not the protector of therefore you can attack it, so it should be allowed to attack itself. I also dislike the rule about creatures not being allowed to be attached to other stuff.)

(If a Siege creature attacks itself (if it were allowed), there is still a strategy involved. For example, if it is a 5/5 creature, then regardless of the number of defense counters, I would expect that since the damage is now equal to its toughness so it will die before the triggered ability can resolve. If the protector has a creature with zero power that can block (e.g. many Walls), then it will not receive any damage. Alternatively, if it is a 4/5 creature with 4 defense counters, then blocking with creatures with 5 or more total power will cause it to die before the other side can be used, too. And yet still, battles are not normally creatures, so you will still need to play combinations of other cards. There is also blocking, which seems that it would be easier to use to your advantage, but still battles are not normally creatures and effects after blockers are declared can still affect the power and toughness of creatures, and abilities such as first strike, etc.)

zzo38, I don't know what your "generalized ongoing" rule is, so mentioning it adds nothing to your comment for me.

Battles attacking themselves makes no sense mechanically and little sense narratively.

Here is another issue with battles attacking/blocking: Now you can use it as a blocker to get rid of defense counters, so instead of investing resources you drain opponent's resources. It's just so plainly cirectly circumventing the "protector" concept that doesn't apply e. g. to planeswalkers that I can easily see why the exception is made here...

There are also logistics to be considered. Battles already have to track both a controller and a protector in multiplayer, that gets even more complicated if they enter combat.

It could be kept track of, by e.g. taking a paper and writing who is the protector. (Even if it is not a creature, writing such things can be helpful for this and other cards too.) (Actually, I might have preferred to not assign a "protector" and handle defense differently than that, but maybe there is not a better way.)

(Another question: If the battle's controller changes, does the protector change, or does it stay the same (even if it is not an opponent of the new controller), or something else?)

I did consider blocking, as I have mentioned. The possibility of combinations to allow it is not much more significant than combinations to damage it directly (or otherwise remove counters from it) anyways. And, if it dies without being able to cast the spell on the other side if it dies due to damage greater than or equal to its toughness, then there is that consideration too, as I have mentioned (that is more important to consider when it attacks than when it blocks, but even if it blocks or is damaged by something other than combat, that can still be relevant). (There are ways to avoid this too, such as by regeneration or indestructible.)

On Twitter, some people also say, it is unintuitive to disallow battles from attacking, and many other people also dislike this rule. (Some people say battles should not be allowed to attack themself, but I think that is unnecessary.) However, some people like such a rule. Matt Tabak has said "you work through a few scenarios with battles attacking and who’s the protector and who can block and you start thinking about future battles possibilities and you realize battles attacking is a very bad idea." Actually, I think that it isn't too difficult. In a game with more than two players, you still need to keep track of both the controller and the protector, even if it does not attack.

MTG Wiki lists the following rulings for battle:

1. Battles are susceptible to more than just combat damage. Some spells and abilities may specifically say that they cause damage to be dealt to battles. Also, any spell or ability that says "any target" can target a battle.

2. Battles enter the battlefield untapped.

3. They can be tapped, but they don’t naturally tap. Being tapped doesn’t have any mechanical relevance.

4. Battles that are Animated as creatures in addition to being a battle can't attack or block.

5. Battles that become an attacking or blocking creature drop out of combat.

6. Battle creatures that receive damage remove that many defense counters in addition to the damage being marked on the creature.

7. If lethal damage also removes the last defense counter, the battle creature will die before it can be exiled and the back face be cast.

I agree with 1, 2, 6, and 7 (and are what I had already expected); they "naturally" make sense within the rules. (I dislike the wording "any target", but the way that rule works is good.) I disagree with 4 and 5, and think that the rules would be cleaner to omit those special cases. I mostly agree with 3, although there would also be possibility of gaining abilities with (or of naturally having such abilities in some cases, even if none of the official cards do), in which case there would be a relevance of being tapped.

• "It could be kept track of, by e.g. taking a paper and writing who is the protector."

You must be new to game design. ;)

To re-state: The task of a designer is to put work into allowing things to work without the use putting a lot of work into it.

That's what the rule does. It takes away an option to make things run more smoothly.

"(Actually, I might have preferred to not assign a "protector" and handle defense differently than that, but maybe there is not a better way.)"

As someone who has spend a lot of time trying to figure this one out without protector-tech: I'm happy it's there. It's the solves the main contention point I had with my own variant. It's a blessing.

There are many things in Magic: the Gathering that you might need to keep track of by markers, papers, etc, including counters, some tokens, Lifelace, Magical Hack, choices made when playing a card, life points, effects that last over multiple turns, etc. (Actually, I think that just writing down who is the protector will likely be clearer even if it does not attack and block) (If each player is using different card sleeves, then you will not need to keep track of owner in this way since card sleeves can be used for this purpose instead. However, that is not always the case.)

About "protector", there are some potential complications with the rules, although presumably they would write them to handle such things (including the protector leaving the game, the controller changing, differences of multiplayer rules (depending on the specific kind of multiplayer and team rules being played, certain considerations may be needed), etc). (I have some ideas, but I will wait to see the official rules.)

I have written unofficial rules, and might continue to do so. (I will wait for the official rules for battles before attempting to write any unofficial rules relating to such a thing, though.)

I have downloaded the latest version of the official rules, which includes the rules for battles. I have updated by unofficial rules with the rules for battles. Some of them do not match the official rules (because I disliked some of them). Some of the official rules are good, and most of them have been kept as they are. For one thing, the protector of a battle with no subtypes is its controller (this is an official rule, and I think that this is good; I might make up some unofficial cards that use this).

Battles, specifically Sieges, are boring, anti-climactic, and anti-interactive.

Sieges demand a protecting opponent. Well, what happens if that player has no blockers when you play a Siege? Basically you get to transform the Siege on the same turn it enters the field with zero interaction or suspense. Whoopdedoo

That's why my set Raiders of Tsinbork reverts back to Scenes (which are essentially Planeswalkers.) In this regard, they create more tension and suspense between players. Opponents have a chance to respond to prevent them from flipping or going off. Scenes build up (whereas Battles erode down), this subtle difference is a better indicator of threshold when something exciting would happen (such as transformation or "ultimate".)