Cooperative board games

Cooperative board games are difficult to design. One major failure mode of a cooperative game is one player directing all the others what to do. At that point, it becomes a one-player game, only there are several additional people sitting around being bored.

One of the best-known cooperative games, Pandemic, has this problem really bad. Pandemic requires players to carefully coordinate in order to contain several diseases as they spread throughout the world. Unfortunately, the “careful coordination” often amounts to the most experienced player telling everybody else what to do. This is often a winning strategy, because the more experienced player can usually unambiguously identify the correct move. I refuse to play Pandemic for this reason.

But I do like other cooperative games. Here I’ll go through some examples and describe how I think they address the issues with Pandemic.

Asymmetrical information

In Battlestar Galactica: The Board Game, players overcome obstacles by playing cards face down, and then summing up the points on those cards. The rule is, you can’t say what cards you have available, or which ones you’re choosing to play. Otherwise, it would be too easy for people to coordinate and play just the right number. You are, however, allowed to give ambiguous information like, “I’m helping a lot for this one”.

This solution works, although personally I find it annoying how ambiguous the rule is. It’s obviously against the rules if we use the phrase “just a tad” as code for “I’m playing a 1”, but at what point does the code become illegal? I prefer games that have more formalized rules about what kind of information can be exchanged. Here are several examples:

In Hanabi each player holds their cards backwards, so that they are visible to everyone except themselves. In order to tell people about the cards in their hand, you have to expend clue tokens, and you’re constrained to giving only specific types of clues.

In Mysterium players cooperate to solve a murder mystery by communicating with a ghost of the murdered person. The ghost (a player) is not allowed to say a single word the whole game, but can only communicate through mysterious visions (represented by cards with surrealist artwork).

Codenames isn’t exactly a cooperative game, but there are two teams, and people within each team cooperate. It’s a bit like Taboo or Pictionary in that one player tries to get their team to guess the right words. But the player’s communication is restricted to a single word per turn.


Many cooperative games, such as Shadows over Camelot and Battlestar Galactica: The Board Game, feature a hidden traitor. Everyone pretends to cooperate, but there’s an element of distrust because some players are secretly working against the team. You wouldn’t want one player controlling the whole team, because what if that person is a cylon?

Unfortunately, there are a few issues with the traitor mechanic. What if the traitor doesn’t know how to play very well? And once you identify the traitor, the game transforms into just another competitive game, possibly not a very interesting one. One easy solution is to make the game short. Once players have identified the traitor, the game is over, or basically over. And then you just play again. There’s a whole genre of hidden identity games along these lines, including Mafia, Werewolf, Avalon, Saboteur, Shadowhunters, and Spyfall.

Another issue I have with the traitor mechanic, is that the traitor’s options are often too constrained. It’s in the cooperative team’s interest to be as transparent as possible, and the traitor has to behave like they also want transparency, even though they don’t. This is particularly an issue in Battlestar Galactica: The Board Game, where the rules are unclear about how transparent you are allowed to be. I consider Spyfall to be a model solution. The cooperative team knows their location, while the spy (the traitor) does not, and is trying to figure it out. The spy tries to be cagey to not give away their identity, while everyone else tries to be cagey to not give away their location.

Aside from having traitors, I think it would be interesting to give players secondary goals, meaning that some players can win “more” than other players. Probably some games already do this, but I’ve never played one.

Dexterity and time pressure

Escape: The Curse of the Temple is a board game that comes with a soundtrack. Players explore a temple, Indiana Jones style, and try to escape before the end of the soundtrack. Players don’t take turns, they’re just limited by how quickly they reroll dice. “Time pressure” refers to the fact that there’s a time limit, and “dexterity” refers to the physical feat of rolling dice quickly. Even if you have some expert player on your team, they can’t roll dice for you!

This is a post about board games, but I should mention that this is the solution used by nearly every cooperative video game. Your teammates aren’t going to take your controller from your hands and play for you.

Lightness and randomness

I recently played a game called Star Trek: Five Year Mission. Players take turns rolling dice, and if you roll the right numbers you can use them to overcome an onslaught of obstacles. There are several mechanisms to prevents one player from controlling everyone else. But for me, the main thing was, this game is so light.

I don’t mean that pejoratively, sometimes we just want a light game. What I mean is that the game is so random that it didn’t seem terribly important to optimize our decisions by asking the experienced players for advice. It also moves quickly enough that asking for advice would slow it down way too much.

Overwhelming complexity

On the other end of the spectrum, we have Arkham Horror. This game has an impressive number of moving components.

Photo of Arkham Horror. There are about 20 piles of cards and 5 piles of tokens, and each player has a bunch of different resources.

While an experienced player can direct a team of Pandemic players to victory, it would be a bit overwhelming to direct a team of Arkham Horror players to victory. There’s just too much to keep track of. Also, the most experienced player is likely the one managing things like the rules of monster behavior, so they’re too distracted to micromanage other players.

The problem with using complexity, is that complexity is relative.  Sometimes it’s too complex for new players and simple for experience players, which can lead them to dominate the team even more. Of course, if one of the players ever becomes too controlling, you can just let them play as two characters at once, maybe that’ll shut them up.

Conclusion and questions

When I researched this problem, I found that it’s often called the “alpha player” problem. Apparently people disagree whether it’s a problem with the game, or a problem with the players. It turns out that I have a strong opinion about this. It’s caused by both. But that’s besides the point. It doesn’t matter what causes it, it matters what the easiest fix is. I don’t know about you, but I find it a lot easier to switch games than to switch player groups.

So there you have it, multiple ways to prevent one player from controlling their whole team. Are there any solutions that you particularly like, or don’t like? Have you played any cooperative games that you felt had elegant solutions?  Or perhaps you’ve found solutions by changing player behavior?


  1. oualawouzou says

    My (almost) teenage daugther just LOVES Mysterium, though she makes a poor ghost as she routinely sighs and rolls her eyes (as teens are wont to do) when we get her “visions” wrong. That said, I think it’s more than asymmetrical information that makes this game work: it’s that the surrealist visions can be interpreted in so many ways, even the most experienced player is bound to make numerous mistakes over the course of a game. Leaving lots of room for error is one way to circumvent the “alpha player” problem. I haven’t come across another game using this mechanic though.

    There’s a Marvel board game (I forget the name, we just call it “Marvel”) that has you recruiting heroes to attack or to boost your ability to recruit more heroes. However, many of these cards have interactions with other cards. That means as the game progresses, each player’s hand becomes more unique. That makes keeping track of all interactions available to any given player quite a feat. Our own alpha player (who’s pretty good-natured about it and can be asked to dial it down without ripping off anyone’s head) always offers advice, but is often overruled because he has overlooked some combo or other. This game also has a built-in point system to pick an “individual” winner once the board is beaten, so it may be in a player’s best interest to keep his strats to himself, but we never tally points that way (why play cooperative when you want to play competitive?).

    We play “The Big Book of Madness”. In this game, we are supposed to keep our cards to ourselves and only offer vague advice, but like you said, after playing with the same people for a while, it becomes trivial to guess what other players are or aren’t holding. However, sometimes, the board throws us a monster that forces all players to stop talking until it is defeated. Coordinating using only gestures and annoyed grunts makes it much more challenging. I could see a game where players are made to draw challenges that screw up communication before each turn (or whatever) and to keep them secret. Challenges such as “say the opposite of what you’re thinking”, “approve everything the player to your left proposes” or “neither agree nor disagree with anything” could make coordinating hell.

  2. bryanfeir says

    Inspired by PZ’s post about Potlatch?

    Yes, I’ve had the same problem with Pandemic. A friend of mine had it worse earlier; our current group tends to be relatively evenly matched, so there ends up being a lot more discussion rather than one person directing it all.

    I think a lot of the problem there is that the amount of randomness in Pandemic is limited to just the card decks, and some of the mechanics of the game actively reduce that randomness even further by reshuffling and placing the old cards back on top of the deck. Every player has a constant number of actions they can take per turn, which makes it possible to make plans multiple turns in advance to some extent.

    Arkham Horror, on the other hand, is not only more complex, it’s a lot more random in terms of what happens, making any plan more than a turn ahead a lot more difficult. (Granted, I’m mostly familiar with the 1987 Chaosium version; does the newer Fantasy Flight version still use dice to determine how many spaces you can move?)

    For traitor mechanics, there’s also Betrayal at House on the Hill, in which the traitor doesn’t actually know they’re the traitor until the endgame starts. Of course, it still has the issue that it becomes a competitive game once that endgame starts.

  3. says

    Yeah I wrote this when I saw Potlatch was a cooperative game, but I don’t know anything about Potlatch and this is a post I had in me for quite a while.

    I’ve only played the Fantasy Flight version of Arkham Horror, which I think is very different, and more complex. You don’t roll dice to move, but there’s still a lot of randomness in the encounters, the Mythos cards, monster selection, and dice rolling for skill checks.

  4. Siobhan says

    Aside from having traitors, I think it would be interesting to give players secondary goals, meaning that some players can win “more” than other players. Probably some games already do this, but I’ve never played one.

    Dead of Winter. The players are a survivor colony of a zombie apocalypse and they all have a shared objective (survive), but “win more” if they complete their secret secondary objective. The tension is that the secondary objective often involves antagonizing the other players (e.g. you have a secondary objective to hoard all the medicine, and one of your teammates is about to die). The tension diffuses the alpha player problem, because they can’t be too good at completing their second objective since the colony has to survive to claim a win in the first place. If all the players are competent, they might pursue too heavily their secondary objective at the expense of the colony’s survival, thus nullifying the win.

    I think it’s an elegant solution.

  5. bryanfeir says

    It sounds quite different. The Chaosium version did require die rolls to move (I think to get into one of the locations to investigate like the library you had to have an exact count, but you could count the two dice separately and use either or the total; it’s been long enough that don’t quote me on that) and when you got into one of the places to investigate you rolled again to look up on a table what happened, and then skill checks as well.

    I don’t think our group ever actually won a game of it, which is kind of appropriate given the source material.

    Good co-operative games do seem to be hard to design, all the more so I suspect because so few people are actively trying.

  6. anothersara says

    There’s Dead of Winter, in which players (other than the secret traitor) win only if the main mission succeeds AND they meet their secret objective. The secret objectives may be complementary (i.e. one player meeting their secret objective does not interfere with other players’ meeting their secret objective) or they may be conflicting, but even if the secret objectives are conflicting they still need to cooperate on the main mission (except the traitor).

    As it happens, I only played Dead of Winter once, and I didn’t fully understand the rules, and the other players said ‘it’s a coop game, we’ll help you’ but I just happened to be the traitor, and I didn’t even fully understand that I was the traitor and all that implied BECAUSE I didn’t really understand the rules. I ended up quitting after the second round because I was bored watching what the other players were doing when I didn’t even understand what was going on. Yeah, that left a bad taste in my mouth. It may be a great game, but it’s so complicated that I think it would be better not to have traitors when there are newbie players.

  7. suttkus says

    oualawouzou, the game you’re referring to is Marvel Legendary. It’s fun, but I find it a bit swingy for my tastes. The right villain/scheme combination and the villain wins before you’ve had a chance to actually use any of the cards you add to your deck. Another villain/scheme combination and you can play cards at random and not worry about losing.

    anothersara, the rulebook for Dead of Winter agrees with you, bluntly stating that a new player should get a game without traitors to learn the rules before having to delve into a game with a traitor. I love Dead of Winter, though. I just wish my playgroup loved it as much as I do. : – )

    A co-op game that’s follows the OP’s suggestion of “some players can win more” is Castle Panic. It’s a tower defense game where everyone works together to defend a castle at the center of the board from all the onrushing monsters. If the castle falls, everyone loses. But players collect the monsters they kill as trophies and the one with the highest scoring trophies at the end wins. The problem with this mode of play is that if someone’s not into the spirit of the game, this becomes, “Well, if I’m not going to win, I might as well make sure everyone else loses, too!” It only works as long as everyone considers losing by loss of castle worse than losing by points. You can play it without points, but then it’s a seriously alpha-tastic game. (My brother’s ability to glance at the board and everyone’s resources and announce exactly what everyone needs to do for optimal play is simply staggering.)

    One of my favorite cooperative games is Sentinels of the Multiverse. Everyone takes on the role of a superhero, and faces off against a supervillain in a comic-book-style location. The game avoids alphaness by the simple virtue of having everyone play with a concealed hand of cards. Sure, you could tell everyone what you’ve got and ask for advice, but then that’s on you. Don’t want alpha’s advice, don’t tell anyone what you’ve got. Some villains may be prone to more alphaness than others (Spite, I’m looking at you, and I don’t like looking at you!).

    Another of my personal favorites is Elder Sign, another game from Fantasy Flight built on the ever-reliable Cthulhu Mythos. Basically, you roll dice and attempt to match them up against certain patterns to dispose of monsters and plots. Doing so earns you resources which can count toward winning the game or manipulating yours or your allies’ dice so that the game is a bit less luck dependent than it might appear at first glance. Not to say that the game isn’t luck dependent, because it really is, but it’s also loads of fun. I don’t see a lot of alpha-ing in this game because nobody’s advice is ever really certain, thanks to the high-luck factor.

  8. says

    Battlestar Galactica added Personal Goals in the Exodus expansion, which are secret goals that you need to achieve or else Galactica will lose resources at the end of the game, which could lose it for you.

  9. says

    I’m getting ready to experiment with Arkham Horror with some friends. So reading about all the complexity doesn’t make me super happy, but I guess I’ll see.

    The obvious way to handle complexity in cooperative games is to let a computer do it. That is, after all, what computers are best at. Some of my friends and I used to get together in person and eat pizza, drink beer, and play something like Diablo. It’s less anti-social than wading around in lots of table-top book-keeping and most of that friend-group constituted the remains of my college-era napoleonic miniatures group: we felt that having the computer take all the “no no noooo keep the cat off the table!” and such was a huge step in the right direction.

    I do think there is a huge niche for social table-top computer games. Games that would be played by passing an iPad around a group, or something like that. There are some really interesting potentials in that sort of system, which I don’t think are adequately explored (and since a play-and-pass game wouldn’t typically need blazing state-of-the-art graphics, it could be implemented easily enough in something like unify) The trick is always really to come up with a great idea and good artwork and balanced and interesting game-play. The people who can do that – I don’t understand how they do it. But the rest is implementation details.

  10. says

    @Marcus Ranum,
    Arkham Horror is complex, but if you’ve had experience playing hobbyist board games it’s nothing to be afraid of. If your play group likes the idea but finds it too complex, I think Eldritch Horror is supposed to be a simplified version of it.

  11. suttkus says

    I had a great deal of fun playing Arkham Horror over the years. It’s not as complex as it looks at first glance. Yes, there are a massive number of piles of cards and more than a few tokens, but a lot of the piles are just different types of the same thing (go here, this happens). I’m not saying it’s simple, because it’s no Ticket to Ride, but it isn’t as complex as it first looks.

    These days, Arkham Horror has been replaced in my heart by Mansions of Madness, second edition, a game I cannot believe I didn’t mention in my last post. Mansions of Madness takes the same summary as Arkham Horror: players take the role of investigators investigating strange happenings in Lovecraft-esque terrain, and end up fighting monsters and struggling to stay sane and alive. However, where Arkham Horror only has the semblance of a plot and character, Mansions of Madness has an actual plot! (For actual character, you probably still want to go with an RPG.)

    Here’s the thing, the first edition of MoM required one player to take on the job of managing the game-state. Essentially, he picked the scenario and red the text boxes and set up the tiles to make the board and moved all the tokens around… and didn’t really do anything that could be called “playing a game”. It was his job to facilitate the other players playing the game by being privy to the secrets of the scenario and telling everyone what happens when they do things. It’s all the work part of being a Dungeon Master in an RPG and none of the creative or fun parts. Worse, as the game is more than a little complex, if he made a mistake in his setup or read the wrong passage of text or moved a token to the wrong place, he could reduce the scenario to an unwinnable quagmire and ruin it for everyone. Mansions of Madness first edition garnered a reputation of being a great game that nobody actually wanted to get out and play.

    So, how does second edition fix those problems? By taking Marcus Ranum’s advice and putting a computer to the job of doing all the hard, boring, automated tasks. When you buy the physical game, you also download an app for your laptop or smartphone or whatever other computer you happen to have available. The players pick a scenario and the computer, inerrantly, informs them where to place tiles and tokens and reads the right text boxes. Second edition is much, much, much more fun than first edition! This is a great cooperative game, and I just dropped a hundred bucks on another expansion box for it! Yeah, it’s not a cheap game, but you get a lot of miniatures, tiles, and, of course, something has to pay for that gorgeous app.

    If you like cooperative games at all, I strongly suggest you find a nerd friend who has this and check it out.

    Bonus: Alpha-players aren’t much of a problem because the theme is investigation. Simply put, the right answer isn’t immediately obvious. There’s some alphaness about (especially with the puzzles, but depending on how much you like puzzles, you might just be glad for it), but it’s less than in a lot of other games.

    Unlike some similar games, Mansions of Madness doesn’t have a traitor, per se. However, if your character goes insane, you might find yourself with an insanity card that obliges you to act either against the group’s interests, or just make you less effective in pursuing the group’s interests. For instance, if you become a pyromaniac, you will lose the game if you don’t have at least so many tiles of the board set on fire when the game ends, so you have to spend a lot of actions setting fires, which doesn’t HELP your allies in the slightest, but at least you still want them to win in general. Still, at least you didn’t join the evil cult in town…

  12. kremer says

    I hadn’t quite considered this issue in board games before.

    My wife and occasionally play some games of Shadowrun: Crossfire, a cooperative game that calls for keeping everyone’s hands secret and merely asking for help as needed, but which we decided long ago was simply too difficult without thinking up tactics together to win. However, the characters in the game can be leveled up between games which makes me wonder whether, now that we’re leveled up, if we couldn’t try out one of the easier missions following the rules as written.

  13. anothersara says

    I also don’t consider Shadowhunters to be a cooperative game, and it’s definitely not over once the affiliations of the players have been revealed. Even if one has figured out every player’s affiliation, someone still has to meet their win condition in order for the game to end. Knowing the other player’s affiliations usually helps a lot, though there are a few neutrals who generally do not care about other players’ identities (the last time I played Shadowhunters I was Bryan, a neutral who only cares about other players identities to a limited extent – I won by getting other players to suspect I was one of the neutrals who wins by dying so that they were reluctant to attack me, and I focused on attacking players who I had a chance of killing, not really caring whether they were Shadow/Hunter/Neutral).

  14. says

    I was thinking of Shadowhunters more as a hidden identity game than a cooperative game. I’m not sure how cooperative the other hidden identity games are either, really, since it seems like the appeal is not in the cooperation, but in pretending to cooperate and then backstabbing people. So that’s one of the problems with using traitor mechanics–for some players it defeats the point of why they want to try a cooperative game in the first place.

  15. Callinectes says

    I actually like Pandemic, but my group doesn’t really get the dictatorial leader dynamic, and we have more or less equivalent experience levels in every game. We rapidly alternative between total confusion, perfect cooperation, and abject chaos. There was one memorable game in which I predicted disaster unless we took specific and drastic action, but I was decried as a melodramatic doomsayer and outvoted. The vindication I felt when the world ended in precisely the manner I predicted was more consolation in the face of the apocalypse than I’d like to admit.

    One game I’d very much like to try is Pandemic Legacy: Season One. This is a game with spoilers. I don’t know that a sequel in implied, the name is supposed to suggest the likes of a major HBO series, and the reviews I hear say that it gets this spot-on. The game can only be played a few times, either 12 or 24 depending on how well you do, because each session is an episode in an arc, and your mistakes and successes in one game impact your place in subsequent games. This includes permanently altering the board, and sealed black boxes and pouches containing cards and pieces that you cannot open until events in the game instruct you to, as the fight against the outbreaks and social collapse go form bad to worse. Death, civil unrest, attacks on your own people, political failure, loss of funding, injuries and psychological trauma on your player characters. It came out in 2015, but for most reviewers still holds the title of “best board game ever made.” Hopefully the curveballs, plot-twists, and non-linear play will prevent single players from dominating play.

    I hear that Risk has also had the Legacy treatment.

    Another favourite of my gang is Betrayal at House on the Hill. You play as an eclectic Scooby Gang of randos (with differing levels of ability, which grow or diminish during play) investigating a haunted house. You build the house as you go, drawing from a deck of tiles to reveal each room and corridor of the ground floor, first floor, and basement as you go. Thus, each game has a different house, usually evidencing an utterly deranged architect. You discover both useful and creepy items, including “omens”. Each omen causes a haunt roll, and each new omen increases the likelihood of the haunt roll starting The Haunt. The original box has 50 different haunts, determined by which omen triggered the haunt and which room is was found in. Then one player is revealed to be the traitor. Usually it’s the one who triggered the haunt, sometimes it’s someone else. The book tells you why: possession, a spell, sudden madness, premeditated malice. Each haunt is a very different scenario, with different pieces and different rules. The win conditions for the traitor and the team players are always different, and either group usually doesn’t know what the other side needs to do. In all my games I’ve never had the same haunt twice. On one occasion the house was transported to an alien dimension with a poisonous atmosphere, on another cursed music would erode our sanity until we couldn’t help but dance eternally, Another saw someone inadvertently release a blob of flesh from a glass orb, which began to expand and fill the house, fully occupying each room, forcing the rest of us to discover and assemble a counter-measure while the traitor tried to force us into the expanding flesh. We all ended up one with the blob in that scenario. One memorable game saw us shrunken to miniature size by a witch, and had to undo the spell before the witch’s cat ate us, though we managed to evade it and navigate the house on board a toy airplane.

  16. suttkus says

    Callinectes, a sequel to Pandemic Legacy is not only implied, it’s on sale at your favorite local game store right now!

  17. Jenora Feuer says

    Just adding another thought on here:

    While I’m not familiar with Marvel Legendary as mentioned above, there’s another superhero card game called Sentinels of the Multiverse which is co-operative, as you play a team of superheroes out to take down a master villain somewhere. The ‘villain’ and ‘environment’ decks are always purely mechanical in terms of their actions (and so can be run automatically), while the heroes each get a hand of cards they can pick and choose from. Villains get attacks which specify ‘attacks target with lowest/highest HP’ so there’s no decisions to be made on targets. Several heroes get cards that don’t do much for themselves but which actively assist others. Some environment cards can affect both heroes and villains.

    It’s pretty good, the creators spent a fair bit of time building world backstory as well as card decks, and there are computer versions of it as well.

Leave a Reply

Your email address will not be published. Required fields are marked *