red-headed android

you know squid game?

deathmatch island is squid game: the tabletop rpg and it’s sick as fuck.

the first thing to know about deathmatch island is that it does one of my favourite things a tabletop game can do: it has mechanics only for the stuff it cares about, and spends all its time pushing those mechanics. on competition show media, like squid game or battle royale or survivor or like, beast games or whatever, we learn stuff about the competitors over time. so the game has mechanics for deepening our understanding of characters over time, which is to be expected.

but you know what else it does? it forbids you from revealing anything about the characters it doesn’t explicitly tell you to!

this is amazing game design! it creates a framework for not only what the cool stuff about the game is, but also what it looks like; it controls play by cutting off things that would be tempting in freeform roleplay, but clash heavily with the themes. no! you may not determine your backstory now! that’s not how this genre works!

and it does this everywhere.

at the start of the game you’re launched into a contest–every PC striving for success, represented by a roll-off between their assembled dice pools–and that contest gives you tangible rewards you mark down on your sheet. so you’re always thinking about the next contest. your competitors win a contest and are provided with weapons? those weapons are not going to be used like ‘weapons’; you do not wave them around to intimidate other contestants, or if you do, that only has narrative or mechanical weight if that scene is part of a contest. they are a bonus for a future contest, plain and simple. so you’re always thinking about the next contest. you decide the next location you’re going to go to–which is inevitably the site of the next contest–as soon as you finish a contest. so you’re always thinking about the next contest.

the pace is breakneck. the game has no time for anything but getting to the next contest, seeing how it turns out & its effect on the competitors, and driving further and further towards the violent climactic showdown of each act of its story (called ‘islands’, because they’re each set on a different island). the game makes sure you don’t have time for anything but that either, by being entirely about that. this is bold. it’s intense. it makes the game extremely fast and smooth to play and run. it’s some of the best game design i’ve seen ever.

but of course, this is a roleplaying game, and people overthink their characters and get bogged down in glacially slow roleplaying scenes about nothing. how does it solve this? well, besides making background information about the characters verboten, it does something else genius.

it randomizes your characters. yeah, buddy, screw you–you can choose these elements if you really want but you’re still choosing from the devs’ list, which is pretty barebones. a couple minor details of look. a name. a job, which provides mechanical benefits and decides what fundamental kinds of contests they’re good at. it takes literally five minutes at maximum. because we know nothing about these characters (as nobody crafted them, they were just assigned these traits five minutes ago), but you’re inhabiting one right off the rip, you spend time stewing in them as a player. who is ‘Jet Briar, Scoutmaster’? you get roleplaying prompts that slowly tease more and more out of the character: a secret one-word motivation at the start, flashbacks between acts, icebreaker questions at the start of the game, prompts to describe their personality and actions during each contest. eventually, you become the person most curious about the character–which makes you the person most eager to get to the places where the game will let you establish things about them.

this is where i get to my first point of criticism. the players need to be locked in for this, and the GM has to communicate this expectation well. if the players start wallowing in angst and doing random freeform roleplay scenes between the important scenes–the contests–the game loses focus rapidly. the GM has to shut this down and keep the pace, and the players have to want to accept that pace. they also need to all agree on the tone; some characters are goofier than others when everyone is random, and some players will want to lean into the absurdity of the game’s corporate dystopian aesthetic and the incongruity of the disparate, random people thrown into its Squid Game-style meat grinder.

so you need good players. but, like–that’s good, right? good players deserve a good game. they deserve a game that matches their energy. if you have a group of people who are mature, trust each other, and are willing to trust the game as it tokyo drifts through its whiplash-paced televised nightmare, you will be rewarded massively for this. most games do not reward you for having good players; the players are the reward. here, the game multiplies qualities like improvisational skill, social tact and the willingness to immerse into a strange world, making them more powerful and creating a stronger experience for everyone. just don’t bring the guy who’s always scrolling his phone during your D&D game. leave that guy at home he can’t handle it. this is deathmatch island, we engage with sincerity and commitment in this motherfucker.

the second thing to know about deathmatch island is that its actual numbers & systems are unbelievably slick. this is the system from John Harper’s Agon, which is already a good game. PCs are working together, but the winner of a contest gets the greatest rewards, which often help them with the next contest–but there’s always enough variance that even the currently-strongest PC can lose, and the currently-weakest can win. injuries, which you can take by pushing yourself too hard, are a catchup mechanic that helps people with bad rolls get stronger over time, while also rewarding risk-takers. those are the themes: you will be meaningfully rewarded for taking the biggest risk you can, but you are also not going to fall so far behind your fellow PCs that the game stops being an exciting competition. the mechanics nail this effortlessly.

and the game is a competition. even though you’ll be a team until the very end, it’s a real fight with real stakes. you earn lasting advantage through doing well early on. the other competitors–NPCs–can and will kill you. you are always incentivized to maximize your odds in each individual contest for a shoot at that sweet loot–shotguns and hand grenades and rations and climbing gear and stuff–that will hopefully ensure your survival in the later contests that really matter. you’re fighting the NPC competitors for that survival, but you’re fighting your teammates to survive best, to move forward best. the tension of knowing you’ll eventually have to kill them too is an incredible driver for this style of play.

since your dice pool is built from the actual advantages your character has narratively–their job being relevant to the contest, the contest relying on a personal ability that they have a good rating in–each setup to a contest feels like a long narrative justification of why you have the best odds, and then a tense roll-off to see if those odds translate into success. the players are given additional buy-in because the winner of a contest gets to choose the context of the next contest. the GM is not shepherding them around from set piece to set piece and choosing their approach for them. the players decide. in my experience, they drive the flow of the game and the narrative of each act far more than the GM. this process is so subtle and elegant i didn’t even notice it until now, but rpg designers should take notes on this one: if you want players to care about a scene or a roll, put them in charge of it and show them the stakes up front.

i did not GM deathmatch island. i played it–and had a great time, partially because my GM was awesome–but i didn’t get a chance to see under the hood.

so i asked my GM. and he says that while the player-facing mechanics are indeed baller, some of the GM-facing ones can be a little hard to parse.

see, Deathmatch Island has no real official setting and, critically, no baked-in explanation for why the contestants are even fighting to the death. they don’t initially remember how they got there, and the game doesn’t tell the GM what’s going on, just how to run the current situtation.

instead, the players theorize what’s going on, and this is translated into a currency that the GM spends on introducing clues that back up the players’ best theories. theories that more players favour get more signs that they might be true. when i played, my character was the only person insisting that the entire deathmatch was the Denghuo Diyu, the first and most violent layer of Chinese Buddhist Hell–this meant that there were relatively few signs that this might be the case, though he kept rationalizing every new circumstance as proof anyway.

more concerning than the lack of hell evidence is the fact that the provided evidence for the GM is actually pretty vague and spotty. a lot of it comes from random tables or lists of suggested elements to introduce to back up one theory or another. everyone might like a version of one theory–it’s aliens, for example–but the game might then tell the GM that if everyone says it’s aliens there should be psychic shit going on, even if the players think it’s more of a biological experiment by aliens. that sort of thing. this takes finesse for the GM to work through and the vibe i got from my GM is that some of these suggested elements to include–as well as some elements of the finale, like ominous (and often laughably un-fitting) titles given to the player characters based on how the other PCs perceive them–mostly just serve as chaff thrown at the GM, slowing them down in what would otherwise be a pretty smooth process of riffing off the players’ theories.

so the idea of this GM guidance is a good one. i could definitely see it helping my GM determine what to do with the NPCs as time went on. one of them was almost assassinated by a corporate SWAT team that airdropped in by helicopter, because he turned out to be a paid actor who was cracking under the stress, and that was because most of us agreed it was a game show, and that was awesome. but there were other times where following the advice of the book on how to develop the behind-the-scenes was clearly putting undue stress on the GM, and that was less awesome. the section could probably do with a rewrite and less focus on prompt lists and random tables. the same is true of the ‘redacted’ section for determining what the behind-the-scenes artifacts and materials look like when the players break the script and go somewhere weird.

so what’s Deathmatch Island? it’s an engine for making your own weird fucked up Squid Game or Battle Royale in fifteen hours and change. it’s a one-shot or short campaign you’ll remember forever. it’s an invitation to create and play fascinating characters you never in a million years would have played in any other game. it’s a strong statement about how mechanics create emergent gameplay experiences, and about how you cannot substitute GM or player skill for real rules–real rules must amplify that skill, not ignore it or lean on it. it’s a reminder to designers that they should honour and support their players.

it’s watching a snarky fuckup kid who lies like he breathes spend days slowly turning from a bunch of descriptive phrases into a person. it’s seeing him fall into despair, convinced he’s entered the endless hell his distant father always raves about, and soldier on anyway. it’s finding out he’s alone, and that’s why he’s here, and that’s why he lashes out, why he’s so smug when he gets one over on the world. it’s learning to hate his lies and then learning to love him through them. it’s watching him get shot through the heart, right when he finally admits who he really is, because the other PCs, they’re playing to win too.

i miss you Tamlin, you little shit.

most campaigns i’ve been in for months and months with my closest friends can’t do what Deathmatch Island did for me, emotionally, in five three-hour sessions with a group of (previously) perfect strangers.

go play it. and play to win.