ATMA is a rules-light no prep Tabletop Role Playing Game (TTRPG) that requires significant improvisational work by all players in order to figure out how anything fits together. The mechanics kind-of work but leave you making up a lot of rules on the fly because they don’t cover a lot of common things.

Like most TTRPGs it’s based around the idea that you’ll go places, bad things will happen, and there’s a good chance you’ll have to fight bad guys to resolve the problem. If you want a good quick overview of the setting and how this game works, I recommend you check out this short 5 minute video, because this post is more about the details than the high-level stuff.

Disclaimer: after writing this whole thing I just feel drained and don’t have the energy to spend another hour editing it. My apologies for any repetition, rambling, or run-on sentences.

Some Perspective

Before I get into this, I feel I need to share the perspective I’m coming from. I refer to myself as a “frustration driven developer”. This true for both software, and game development. What’s notable is that the things that really frustrate me are the things that could have been really good, but were left in a state that just… isn’t.

It should also be noted that I’m a total nerd when it comes to game mechanics. I love learning new ones, and really understanding the tools available for RPG design.

ATMA has great bones, but it is missing a lot of important tools, and has many internal contradictions. This frustrates me greatly, but again only because it could be great.

First, Something Good

The game absolutely lives up to its claims of being a “no-prep” RPG. The setting is fun. It’s nice to just have the game tell you “stuff is happening here” and having the next scene be thought up by the game is also nice. The characters have nifty backstories. It’s cool to have art to give you an idea of your character and the characters they encounter.

It’s nice that each box is usable by itself, and that none of the boxes contain the same cards. It should be pointed out that the “Interlude” box is in no way an “interlude” between Act 1 and Act 2, or anything else. It’s just a tiny version of the Act 1 game with less cards that can be combined with Act 1’s cards and / or Act 2’s.

Is this a fighting game or not?

I made the assertion to Dachary (my wife) that this game was a “combat focused game” that expected you to go from scene to scene and usually fighting to resolve problems. While playing we managed to resolve one scene without any combat. We also learned that there were characters who were completely useless in combat situations.

Afterwards, she countered that it couldn’t have been intended as a “combat focused game”, because there are multiple characters who have no combat abilities. Unfortunately the same argument applies in the other direction because there are characters who are exclusively combat focused.

Our first scene was a cooking contest, and the possible goals were to win the contest, or stop an extortion attempt. The combat focused characters are going to suck at the former and the latter has a decent probability of erupting into violence.

This implies you could pick non-combat characters and choose non-combat goals. However, the “Story” card defines your overall goal and doesn’t let you pick. Ours was “Sabotage the weaponized army of [robots]”. That could theoretically be accomplished without combat, but that seems very unlikely unless you’re telling the story of a spy spending months or years to get into position.

Everything’s contradictory

Some of the characters are 100% combat focused. They have zero non-combat moves beyond the ones everyone has (Realize, Coerce, Survive). Some characters are useless in combat situations. They have zero combat moves beyond the ones everyone has (Rumble, Barrage).

The GM never rolls. NPCs never roll to attack. Instead the GM spends a token to justify the action, and players roll to “Survive”. Cypher System clearly demonstrates that games with “player-facing” rolls like this can work well. However, there’s no mechanic, or even guidance, in ATMA that supports NPCs in conflict.

Many of the cards clearly show that NPCs can be allies. However, in situations where a Player would roll to see if they succeed the NPCs have to spend a token. BUT tokens are the only currency the GM has to complicate the lives of the characters (make the story more interesting) and spending tokens for an ally to help the players directly undermines this ability.

The game clearly wants you to be able to get into a fight, and conveys that some characters are big and powerful and some are small and weak, via both art, character abilities, and different characters having different amounts of health. Mechanically however everyone does exactly the same amount of Harm when attacking. A titan with fist that look to be the size of Volkswagen Beetles can successfully punch a skinny human teen with no defenses and do 1 harm. The skinny human teen with no weapons or combat abilities can punch a massive robot and do 1 harm. The combat mechanics around damage completely undermine the narrative fiction.

You can take harm, but there’s no mechanism to heal. The quote below is the literal entirety of the healing rules.

If [the characters] need medicine, food, or supplies to heal, offer hints about where these might be found in the scene.

Missing tools

Sticking with the problem of healing for a moment. We have no idea what the effects of “medicine”, “food”, or “supplies” to heal are. It’s a sci-fi game so we could assume some “med pack” thing that inexplicably heals things, but how many hearts does that heal? Why would “food” heal a character who was stabbed? How do “supplies” help? Can any of this be used during a fight?

As mentioned before, there are no-tools for handling NPC conflict of any sort, and no tools for NPC allies to do anything. It’s clear that NPCs can be allies, but if an NPC has special abilities it requires the GM to spend a token to activate, which again undermines their ability to make things more “interesting”.

Drawing NPC cards randomly requires either a GM with master level improv skills OR a mini brainstorming session where the table tries to come up with some plausible explanation for why this character just appeared and what they want.

Most GM-less games address this with something like an “oracle”. This is usually two tables that you roll on. One is a “descriptor” and one a “subject”. You ask a question and roll on the table. For example “why did this food delivery person suddenly enter the scene?” I just rolled on one of my oracles, and it said “threatening location”. So, maybe they’re a robber who is just disguised as a food delivery person to gain entry. It still requires a little improv, but it’s not like the game’s saying “Fuck if I know. You figure it out.”

ATMA requires you to constantly do real work to figure out everything. In “Theatre of the mind” games like this you need to establish a shared narrative space that makes sense to everyone. The physical locations of each individual in the scene may not be important, but everyone needs to share the same understanding of what’s going on, and what the character motivations seem to be. It also has to be something they can all “buy in to”. If you’re thinking “This makes no sense. They’d never do that.” then you’re going to have a really hard time engaging with the game.

Combat is also missing a ton of important information. For example: I had a special ability where I could punch my huge fists into the ground and cause an earthquake. This was visually depicted as a huge rent in the ground shooting away from me in a straight line. So… could I effect multiple NPCs with this? Did they have to be in a line. The canon weilding bear’s ability specifically states multiple characters are effected. Mine mentioned nothing like this. We had to make a ruling on the spot. We decided “multiple, but in a line”.

Dachary’s character’s hair seemed like maybe it was prehensile, but it wasn’t stated. We ruled that it was. It was drawn as standard western female hair length but the cards said it could be used as a rope. This implies it could get longer. How much longer? These questions actually came up in play and had to be decided or kicked down the road to be discussed again later. We ruled that it was long enough to use as a rope across a passageway, and punted on the question of how much longer it could get.

By itself these aren’t big problems. They’re easy to rule on the fly. The problem is that you’re constantly having to do this. Imagine playing D&D but literally every spell and special ability is missing critical information like that. Not only are you constantly designing the game as you go, but you’re defining house rules about every ability than you now need to keep track of for the remainder of the game.

What happens if someone at your table wants to use that character again in a game months later? You probably didn’t write all those home-brewed rulings down. You’re going to have to do the work all over again, and sometimes you’ll rule differently and then someone’s going to get confused about what’s possible because things worked differently last time than this time.

Nonsensical cards

The “Props” and NPC cards are especially problematic.

NPCs were were kinda neat, but most of them made absolutely no sense once a scene was in motion. Imagine you’re playing out the fight in the beginning of Star Wars. The Empire’s forces have boarded the ship, laser bolts firing everywhere, they’re about to capture the princess, and suddenly Donald Duck walks into the room wearing only a bath towel. Or, maybe, a volcano erupts, and threatens a nearby town, that you’re obviously not in because you’re on a star ship. I’m not exaggerating. It’s that level of random w.t.f.

We ended up reading ahead and picking NPCs manually because most of them made no sense in the scene. When we eliminated them and needed to introduce a new NPC we just said “uh… I guess another one of those appears?” because none of the remaining cards made any sense given what had transpired.

“Props” were mostly useless. One was a mug of spicy chai. They either needed to be worked in as a scene was revealed. Like “oh, there’s um… a spicy chai vendor in this outdoor plaza”. Or else it was like “You just got out of a fight, and the room is filling with lava, but as you look around you notice a mug of spicy chai.” Both require effort to come up with a justification for, and both leave the players going “What?!” or “uh… ok? I don’t know what to do with that information.”

Invalidating Player Abilities

“GM Intrusions” is a term I’m borrowing from Cypher System, because ATMA has no term for this mechanic but uses it. In Cypher System (also FATE, and others) the GM offers the player some resource they can use in exchange for introducing a complication. In ATMA the GM can spend a token to say “yes i know you rolled successfully, but I’m invoking something that partially or fully invalidates that success” Tokens are a GM only resource so the player can’t do anything with it, and they have no means of refusing the intrusion.

I was the managing most of the cards in our game so I just ignored this because taking away player agency is bullshit.

In addition to this, the lack of GM rolls meant that the GM only has to spend a token to get around any kind of obstacle. For example, I had gravity spikes that people “fell towards”. I lined them up to create a barrier to hold back the bad guys. The GM never rolls, and spends a token to have the NPCs do anything and literally can’t fail if there’s no player roll to counter it. So, as a player you throw down your gravity spikes to slow the enemy and the GM says “ok, I spend a token on my turn and jump over or push through them”. What’s the effing point of giving me powers if the GM can just spend a token to ignore its effect?

The token economy is broken

It’s like, really broken. The instructions tell the GM to try and not keep more than three tokens. So, you’d be constantly spending them to introduce bad guys in a fight, or environmental complications. Tokens are generated for the GM on every roll that isn’t a critical success. So, they’ve got a LOT of tokens coming in that need to be spent.

In a fight this works if the number of bad guys is roughly equivalent to the number of Player Characters. If we have 4 characters on each side of a fight then the GM will spend ~4 tokens each round, and roughly 4 tokens will be generated each round.

We were playing it as a two player game, which is officially supported. Every scene is started with 3, 4, or 5 tokens + whatever you had from the last scene. Every round of combat we generate ~2 tokens. Let’s assume the GM starts the scene with 8 tokens. They either they need to bring out a lot of adversaries to spend them or introduce a lot of “props” or “twists”. You don’t want to introduce more than one twist per scene, and the props are mostly nonsensical and/or contextually useless.

My character had 3 Hearts, Dachary’s had 2. Adversary hits do 1 Heart of damage. Actually we don’t know this. It’s never stated. We’re just assuming, because if they did more than that her character would be taken out with a single hit and mine with two.

This means that if there were a lot of bad guys one of two things would happen. Either, we’d get eliminated very quickly, or the GM would spend all their tokens attacking us, we’d get lucky and dodge, and then half of the NPCs wouldn’t be able to do anything every round afterwards because we’re only generating 2 tokens but there are 4+ opponents. The GM isn’t receiving enough tokens to spend to have them do anything. Eventually, by having each opponent act, the GM spends down to an equilibrium that matches the number of characters or is slightly below it. They’re screwed.

If the GM is managing an ally NPC then they’re screwed even more quickly, because the ally will also cost a token to act. NPCs don’t roll, so having an NPC ally means having a guaranteed successful attack (no roll so it can’t fail), that also takes away an action from the GM.

Missing what makes PbtA great.

One of my favorite things about “Powered By The Apocalypse” (PbtA) is the playbooks. Every character has a description, and some special moves that not only made them special, but also really gave you a feel for the character. ATMA does give you a feel for the character via the backstory on the back of each special ability card, but the abilities themselves don’t feel like they really tell me about the character. The cannon weilding bear can bite things, and shoot its cannon. This doesn’t help me feel the character at all. The teleporting teen can… teleport.

The Titan I played did kind-of achieve this, but it was also contradictory. My backstory said I had been mind controlled and forced to hurt people and I felt bad about that. Meanwhile one of my special abilities was to embrace my violent past, and slam my fists into the ground. Afterwards (or separately? I don’t know.) I could “deny mercy” and “sow terror” or “incite hatred”. Am I a violent raging ass-hole, or am I a repentant person who only fights when forced to?

It would have made much more sense if the special move left me shaken about reverting to my past ways, or something like that.

Despite being mind-controlled for years, one of my abilities was to posses others. No-one who’s been mind-raped for years and forced to do things against their will is going to be comfortable mind-raping other beings unless they’re just a horribly broken being that needs to be eliminated for everyone’s safety. Even if we completely ignore the safety issues around this topic, it just doesn’t make any internal sense. In their defense they did get sensitivity readers for the “Act II” set, so hopefully there’s no more of this in those cards.


This game really frustrates me, which is another way of saying that this game has great potential.

I am fortunate to have a wife who’s amazing at coming up with stories on the fly. This helped a lot, but - without exaggeration - we spent at least half of our time trying to figure out why everything was happening, and trying to figure out what the hell ever card we pulled had to do with anything.

We’d paused before the final scene because dogs needed walking. The next day, we wanted to see how the story ended, but by the time the final climax was reached and we were starting to vanquish the bad guys I did not have the energy to give a shit. I had spent so much time, and energy trying to come up with explanations for everything and constantly having to come up with new house-rules for everything our characters tried to do, that I just didn’t care. It was like “…and the last two guys run away and [quick illegal roll for an NPC] the Magistrate barely wins the election and eventually all is well with AIs and non-AIs. The end. I don’t care. I just want to put away the cards.”

This game could be really good. Every mechanical problem it has is something that was solved years ago by other talented designers.

It doesn’t seem to know if it wants to be a fighting game or not, and the characters all feel completely unbalanced. Either they’re all about fighting or they’re useless in a fight. The ones that’re useless in a fight mostly have no skills that’d be useful in a game about social conflict.

The character’s special abilities all need to be rewritten so that you don’t have to create a house-rule for all of them. Is the hair prehensile? Roughly how long can it become? Can the teleporter teleport places they haven’t been?

The combat needs to match the narrative. It completely breaks me out of the fiction to have a huge titan smashing a tiny teen human do exactly the same damage as a tiny teen human with no fighting abilities punching a massive metal robot.

The description needs to match the images and be internally consistent, and consistent with the other elements of the world. The “Warmech Veteran” is drawn with a human and a robot. The description is about a robot who “suppressed their higher level functions like logic and speech” but it can “befriend another fighter” … isn’t friendship kind-of a higher level function? What’s the deal with the human? Are they a pair? The robot is drawn with some sort of missile launchers on its back, but are they unlimited missiles? does the missile hurt only one character if it explodes beside them? What does it mean that it can it “ride upon missiles”? Why does a big robot that’s obviously designed for war, only have 2 hearts when a squishy human also only has two hearts?

Speaking of hearts, why do some characters have multiple separate lines of hearts and some only have one line that may be long or short. Does that separate line mean something? Does something happen when all the hearts in one line are removed?

I hate that it has forced me to either ask these questions and come up with house-rules for each, or say “fuck it none of these rules make sense, no-one understands what my character or the enemies are actually capable of, and I have to just accept it”?

I don’t think this game is good for new players if any of them are inclined to actually consider the implications of anything written on the cards. If you have players who embrace “don’t think about it.” and are really good at collaborative improve you’ll be fine. Otherwise, I can’t recommend this. It just requires way too much work, because it defines way too little, and doesn’t provide the support you need to make generating the story easy.