I imagine one could have a game where multiple players are competing to be the traitor. I'm thinking something with a hand of cards or something, where the stars could align for you to have a shot of ditching the team objective and winning on your own. Everyone is counting the cards and trying to figure out if the traitor pieces are in play and who might have them. Maybe there would be some kind of audit action which would impose a cost to let you go through the cards of another and force them to turn over their cards and you can trade your own cards for them or discard a number of them to get the traitor cards back into circulation and/or ditch them but obviously if you kept cards people would want to audit you...
Or maybe a game where everyone was collaboratively building something on the board that would be hijacked by one of the players, who at that point becomes the traitor against which the other players are aligned, and everyone wants to be the traitor so they can be the sole victor. There would be a whole element of trying to figure out when you can pull off betraying the team and still win; maybe the traitor has more limited resources and so they need to make sure they are in a position of strength to betray the team, and the strength is contextual (there is more than one working part) so like, players are all sabotaging the building of the machine in specific ways and stockpiling resources so that when they betray the team the machine is complete but for a critical part they can alone can make...
That second game also begs for an excellent mad scientist theme.
"Play sub-optimally to avoid detection" is possibly the single worst game mechanic that exists in modern tabletop. It's unsatisfying in the extreme.
I completely agree. Sandbagging sucks to the extreme.
I disagree. If playing "optimally" would lead to your being detected and losing, it's not optimal.
What you're talking about is an arbitrarily defined subset of the more general mechanic of "play in a way that doesn't give away too much information", which is a very important consideration in games like, say, poker and Pandante. Those two are perfectly good games, and the mechanic works totally fine in the case of many other games as well.
I'm entirely happy to defer to your judgement in the context of "traitor games", because I haven't played many of them and I have no problem accepting that that mechanic is universally terrible across those games. However, assuming that is true, I suspect it's much more likely to be a case of bad design across all those traitor games rather than a fundamental problem with the mechanic itself.
I suspect it's impossible to fix in turn-based games where the traitor has the same inputs as a regular players and where detecting the traitor is a boon to the rest of the players.
Simply put, every set of rules I can come up with to make the traitor mechanic actually work make the rest of the game incredibly unfun or eliminate all player communication.
In my personal experience (I play with a group of 5 - 6 very experienced in bluffing/traitor games). There is no one person who tops every single time we play Avalon or Coup. However when there is a new player they tend to get destroyed as sub-optimal plays or bluffs are recognised by the regulars from the group.
When there are are only two or three players the games are broken though.
Dimitri and I found this to be true at Pax Aus when we tried to play 2 and 3 player Coup. A round would last all of 30 seconds as both of us were playing almost perfectly.
These games are also pretty fun as opening games at parties before a more strategic long ass game. Sometimes also played at University between classes (one of the few good things about going back to school).
Most memorable moment being a traitor - playing Baratheons in Game of Thrones.
Manipulating Tyrells to kill Martells and pressure Lannisters while I killed the Starks then swept over and won the game by backstabbing the Martell's unprotected castles.
"Play sub-optimally to avoid detection" is possibly the single worst game mechanic that exists in modern tabletop. It's unsatisfying in the extreme.
I completely agree. Sandbagging sucks to the extreme.
I disagree. If playing "optimally" would lead to your being detected and losing, it's not optimal.
What you're talking about is an arbitrarily defined subset of the more general mechanic of "play in a way that doesn't give away too much information", which is a very important consideration in games like, say, poker and Pandante. Those two are perfectly good games, and the mechanic works totally fine in the case of many other games as well.
I'm entirely happy to defer to your judgement in the context of "traitor games", because I haven't played many of them and I have no problem accepting that that mechanic is universally terrible across those games. However, assuming that is true, I suspect it's much more likely to be a case of bad design across all those traitor games rather than a fundamental problem with the mechanic itself.
Those are good points. I guess in the end, whether a mechanic is suitable or not depends on the core of the game. Still, I don't like the feeling of having to keep a low profile by playing in some wierd way that just "doesn't feel right". I suppose it's generally a problem in games where leader bashing is significant, and so, good players feel inclined to sandbag to avoid having to fight too many fronts. Nonetheless, there are actually direct conflict games which minimize kingmaking and leader bashing, most notably Titan, WITHOUT using pseudo-hidden "victory points". In fact, Titan doesn't even have victory points (although you could argue that in some cases, your point total, which is sort of like the amount of EXP you have, do function as VP) since Titan is a game about player elimination.
I'm sure some are starting to groan about my gospel of Titan, but it really solves many common problems with wargames while retaining a high level of player interaction, rewarding skill over luck, and maintaining near-limitless replayability and a huge level of strategic depth. I've played over 2000 games and it still continues to fascinate me; I can also recall times when I've spent several hundred games just figuring out when, where, what, and how to attack using my kill stacks.
Of course, it still has its flaws, and it's a pain in the ass to learn if everyone who's playing the game is a beginner, but once players start being truly aggressive, the game ascends to a transcendent level of gaming bliss.
What could work for a traitor game would be where everyone is playing cooperatively but at certain points a player could openly decide to go traitor and pull off a single player win.
Example: everyone is playing together against the game. They move their pawns around the board and complete objectives to beat the game. However, every time a certain something happens the player that triggered that something draws a card from the traitor deck. On their own these cards do nothing and some of them literally do nothing (blank). However the cards that are not blank have text that only activates when the player chooses to go traitor.
As a player accumulates these cards the other players will begin to wonder if that player will turn on them. If several players accumulate these cards at the same time they could choose to go traitor as a group. Once someone goes traitor the game changes into a two team versus game. At this point the game could offer an elimination or redemption/conversion mechanic and then resume as a co-op game after resolution.
If not by choice the traitor cards could have a value on them and once the value exceeds X that player has to go traitor.
What could work for a traitor game would be where everyone is playing cooperatively but at certain points a player could openly decide to go traitor and pull off a single player win.
Example: everyone is playing together against the game. They move their pawns around the board and complete objectives to beat the game. However, every time a certain something happens the player that triggered that something draws a card from the traitor deck. On their own these cards do nothing and some of them literally do nothing (blank). However the cards that are not blank have text that only activates when the player chooses to go traitor.
As a player accumulates these cards the other players will begin to wonder if that player will turn on them. If several players accumulate these cards at the same time they could choose to go traitor as a group. Once someone goes traitor the game changes into a two team versus game. At this point the game could offer an elimination or redemption/conversion mechanic and then resume as a co-op game after resolution.
If not by choice the traitor cards could have a value on them and once the value exceeds X that player has to go traitor.
We hated it enough to not let it beat us. We played it until we beat it.
It has a lot of the problems of traitor and co-op games in general, but it did the traitor thing well enough to where you could feel the tension when the person who had the ring had to choose whether or not to throw it in the lava.
I suspect it's impossible to fix in turn-based games where the traitor has the same inputs as a regular players and where detecting the traitor is a boon to the rest of the players.
Simply put, every set of rules I can come up with to make the traitor mechanic actually work make the rest of the game incredibly unfun or eliminate all player communication.
Okay, let's break this down a bit more.
As I mentioned before, there is nothing wrong with the basic idea of "play in a way that doesn't give away too much information". This is a mechanic that is necessarily present in any imperfect information game, but it's still present to varying degrees in different games. It seems to me that the key issue with "traitor games" is that there is often a very sharp cutoff in terms of the information you leak---there is usually a very sharp line between "traitor" and "not a traitor", which leaves the traitor with little to do except bide their time.
Granted, I think biding your time and waiting for the right moment to strike has some appeal in and of itself, but it lacks the desired subterfuge for something you would call a "traitor" game. Thus there is a desire to blur the line between "traitor" and "not a traitor" so that play doesn't just work as a binary switch from one to the other. (1) Players must be prevented from keeping everything in the open (2) The traitor's actions need some kind of cover
Now, (1) is pretty important, but I don't think Rym is correct in saying that you need to eliminate all player communication to achieve it. You just need to make it so that information shared by the players is at a high risk of being exploited by the traitor, a la Mafia or The Resistance.
As for (2), there are several ways to achieve this.
First of all, it can be done with the game mechanics---players' moves can be partially hidden. The most basic mechanic here is playing of face down cards, in conjunction with at least one of the following: - a prolonged build-up period before the cards are revealed, which can give the traitor greater leeway before their actions are revealed. - mixing of cards played by different players and/or deck cards to hide information
Secondly, the traitor's actions can be covered by giving incentive to normal players to make traitor-like moves. Once again, there are a couple of ways of achieving this: - along the lines set out by open_sketchbook, have the traitor not be fully determined until the end of the game. This could include mechanics by which the traitor is randomly determined, mechanics by which players can choose to become the traitor, or mechanics by which the traitor can pass on their role to another player and stop being a traitor. Any of these rules would incentivise players to spend some effort setting up for the possibility that they will be the traitor. - on a similar note, give players some individualised incentives. For example, the non-traitor "team" could get utilons (i.e. real/fake money, or points in a multi-round game a la Hearts) for collective victory, but also bonuses for accomplishing various individual objectives that could also serve to give the traitor cover. - have a multi-round context where winning as a traitor scores very highly. This way, players are incentivised to play non-traitor rounds somewhat traitorously to give cover to future rounds in which they could be a traitor.
On the whole, there are some design challenges but based on my analysis of the core mechanics I don't see any fundamental reason why "traitor games" cannot work.
I'm trying to work out a mechanic that lets one person be dealt a different hand of cards than everyone else. For example, for 5 players, shuffle different decks of traitor and non-traitor cards, then arrange a 50 card deck so that every fifth card (counting from the first card) is a traitor-themed card. Then never shuffle this new deck! Instead someone cuts it, puts the top half on the bottom, and then the cards are dealt to the players.
Every round, all the players play one card face down, and then those cards are shuffled before being revealed.
Something something something. Probably the team's victory points can be traded for the ability, in the last stage of the game, to reveal a single player's card, to see if they are the traitor or not.
After a certain point, people have to draw cards to replenish their hands, these come from a mostly non-traitor but partially traitor randomly shuffled deck. Now the traitor will start drawing non-traitor cards, and some others will draw traitor cards. This is like the growing influence of the traitor!
Later the team can start drawing single cards from a player's hand, up to the number of victory point they won earlier, which of course is in the traitor's interest to minimize. If it comes up traitor, the team wins. If the game ends without anyone revealed as traitor, the traitor wins.
I don't think you need to go through all the trouble of managing a deck like that. You can just have beneficial cards and detrimental cards all mixed up. The non traitors will only want to play beneficial cards. The traitors will want to play traitor cards. But sometimes the traitor will have to, or want to, play beneficial cards if that all they have. Non traitors may have to play traitor cards if that's all they have.
That kind of sequential card stuff would be crazy fiddly anyway. Setup would take forever and the game would probably break more often then it worked.
I might be neat if every card had both a beneficial team effect, and an effect the traitor could eventually exploit. You are creating something on the board, filling in a pattern or something with your various cards. The traitor is trying to put down cards that will eventually benefit their traitor and fitting it into the scrabble pattern of other cards that is, say, building pots of victory points. Make it so that the "stopping the traitor" mechanic is detached from the "team building whatsit" part, so both the traitor and the other players are both trying to build these pots of victory points while also maneuvering to claim them; that way during the sneaky phase, its less obvious the traitor is the traitor because the cards that help the team could also potentially be benefiting the traitor. Maybe the players have a super, super limited number of "fuck the traitor" cards they can put down in the pattern to block the traitor from claiming them, and perhaps they are temporary so putting them down too early, or if you aren't sure who the traitor is, is a waste.
If you have a hand of cards, the capabilities of the traitor are obfuscated, so the traitor can play closer to optimally as a traitor without giving away their role, and play well as a player while still benefiting themselves as a traitor. There's also the potential to make the traitor traitor-ing a reveal on the traitor's part; they suddenly laugh evilly and collapse the house of cards they have been secretly building, and all the players scramble into damage control mode.
My point is that the weight of traitor cards in the hand of a player determines if they are a traitor or not. At the start, everyone has (for example) 7 cards, and one player has 7 traitor cards, and everyone else has 7 non-traitor cards.
As the game progresses, the traitor will lose some traitor cards and gain non-traitor cards. They will never become a non-traitor, because one of the cards won't be able to be played at all, and that will always have enough weight to make whoever has the card a traitor.
Yet at the same time, others may pick up traitor cards, and the weight of those cards will beat out their non-traitor cards.
I'm not sure the start would be more fiddly than other games. For 5 players:
1. Select the 7 traitor starter cards. 2. Shuffle the 28 non-traitor starter cards. 3. Place a traitor card on the table. 4. Place 4 non-traitor cards on it. 5. Repeat steps 3 and 4 six more times. 6. Cut the deck, put the the top half under the bottom. 7. Deal the 35 cards to the 5 players. 8. Shuffle the rest of the non-starter cards (both traitor and non-traitor) and use that to replenish the hand after each play.
It's really not so hard. It will be immediately obvious if it didn't work, as one player will have all traitor cards and everyone else will have none.
The problem this is solving: How do we give one player all-one-kind-of-card and everyone else random other cards? And how do we do that in a way that nobody knows who got all the traitor cards? One way would be to sort all the card into 4 non-traitor piles, and one traitor pile, but then how do you shuffle those sets of cards without mixing them?
Some kind of bigger card sleeve that could hold easily a handworth of cards, alternatively small envelopes. Not the most elegant solutions, but I'd think it better than some "every fifth card" -arrangement.
I think it would be fun if every round each player could play 0, 1 or 2 cards, and each player has their own deck of cards (from the envelope) to replenish their hand for the first phase of the game. Then until that point, there could be a logic puzzle to see who the traitor could be by how many cards are played in the round and how many cards are revealed to be traitor cards.
Comments
Or maybe a game where everyone was collaboratively building something on the board that would be hijacked by one of the players, who at that point becomes the traitor against which the other players are aligned, and everyone wants to be the traitor so they can be the sole victor. There would be a whole element of trying to figure out when you can pull off betraying the team and still win; maybe the traitor has more limited resources and so they need to make sure they are in a position of strength to betray the team, and the strength is contextual (there is more than one working part) so like, players are all sabotaging the building of the machine in specific ways and stockpiling resources so that when they betray the team the machine is complete but for a critical part they can alone can make...
That second game also begs for an excellent mad scientist theme.
What you're talking about is an arbitrarily defined subset of the more general mechanic of "play in a way that doesn't give away too much information", which is a very important consideration in games like, say, poker and Pandante. Those two are perfectly good games, and the mechanic works totally fine in the case of many other games as well.
I'm entirely happy to defer to your judgement in the context of "traitor games", because I haven't played many of them and I have no problem accepting that that mechanic is universally terrible across those games. However, assuming that is true, I suspect it's much more likely to be a case of bad design across all those traitor games rather than a fundamental problem with the mechanic itself.
Simply put, every set of rules I can come up with to make the traitor mechanic actually work make the rest of the game incredibly unfun or eliminate all player communication.
When there are are only two or three players the games are broken though.
Dimitri and I found this to be true at Pax Aus when we tried to play 2 and 3 player Coup. A round would last all of 30 seconds as both of us were playing almost perfectly.
These games are also pretty fun as opening games at parties before a more strategic long ass game. Sometimes also played at University between classes (one of the few good things about going back to school).
Manipulating Tyrells to kill Martells and pressure Lannisters while I killed the Starks then swept over and won the game by backstabbing the Martell's unprotected castles.
http://boardgamegeek.com/boardgame/72/verrater
I'm sure some are starting to groan about my gospel of Titan, but it really solves many common problems with wargames while retaining a high level of player interaction, rewarding skill over luck, and maintaining near-limitless replayability and a huge level of strategic depth. I've played over 2000 games and it still continues to fascinate me; I can also recall times when I've spent several hundred games just figuring out when, where, what, and how to attack using my kill stacks.
Of course, it still has its flaws, and it's a pain in the ass to learn if everyone who's playing the game is a beginner, but once players start being truly aggressive, the game ascends to a transcendent level of gaming bliss.
Example: everyone is playing together against the game. They move their pawns around the board and complete objectives to beat the game. However, every time a certain something happens the player that triggered that something draws a card from the traitor deck. On their own these cards do nothing and some of them literally do nothing (blank). However the cards that are not blank have text that only activates when the player chooses to go traitor.
As a player accumulates these cards the other players will begin to wonder if that player will turn on them. If several players accumulate these cards at the same time they could choose to go traitor as a group. Once someone goes traitor the game changes into a two team versus game. At this point the game could offer an elimination or redemption/conversion mechanic and then resume as a co-op game after resolution.
If not by choice the traitor cards could have a value on them and once the value exceeds X that player has to go traitor.
http://boardgamegeek.com/boardgame/823/lord-rings
We hated it enough to not let it beat us. We played it until we beat it.
It has a lot of the problems of traitor and co-op games in general, but it did the traitor thing well enough to where you could feel the tension when the person who had the ring had to choose whether or not to throw it in the lava.
As I mentioned before, there is nothing wrong with the basic idea of "play in a way that doesn't give away too much information". This is a mechanic that is necessarily present in any imperfect information game, but it's still present to varying degrees in different games. It seems to me that the key issue with "traitor games" is that there is often a very sharp cutoff in terms of the information you leak---there is usually a very sharp line between "traitor" and "not a traitor", which leaves the traitor with little to do except bide their time.
Granted, I think biding your time and waiting for the right moment to strike has some appeal in and of itself, but it lacks the desired subterfuge for something you would call a "traitor" game. Thus there is a desire to blur the line between "traitor" and "not a traitor" so that play doesn't just work as a binary switch from one to the other.
(1) Players must be prevented from keeping everything in the open
(2) The traitor's actions need some kind of cover
Now, (1) is pretty important, but I don't think Rym is correct in saying that you need to eliminate all player communication to achieve it. You just need to make it so that information shared by the players is at a high risk of being exploited by the traitor, a la Mafia or The Resistance.
As for (2), there are several ways to achieve this.
First of all, it can be done with the game mechanics---players' moves can be partially hidden. The most basic mechanic here is playing of face down cards, in conjunction with at least one of the following:
- a prolonged build-up period before the cards are revealed, which can give the traitor greater leeway before their actions are revealed.
- mixing of cards played by different players and/or deck cards to hide information
Secondly, the traitor's actions can be covered by giving incentive to normal players to make traitor-like moves. Once again, there are a couple of ways of achieving this:
- along the lines set out by open_sketchbook, have the traitor not be fully determined until the end of the game. This could include mechanics by which the traitor is randomly determined, mechanics by which players can choose to become the traitor, or mechanics by which the traitor can pass on their role to another player and stop being a traitor. Any of these rules would incentivise players to spend some effort setting up for the possibility that they will be the traitor.
- on a similar note, give players some individualised incentives. For example, the non-traitor "team" could get utilons (i.e. real/fake money, or points in a multi-round game a la Hearts) for collective victory, but also bonuses for accomplishing various individual objectives that could also serve to give the traitor cover.
- have a multi-round context where winning as a traitor scores very highly. This way, players are incentivised to play non-traitor rounds somewhat traitorously to give cover to future rounds in which they could be a traitor.
On the whole, there are some design challenges but based on my analysis of the core mechanics I don't see any fundamental reason why "traitor games" cannot work.
Every round, all the players play one card face down, and then those cards are shuffled before being revealed.
Something something something. Probably the team's victory points can be traded for the ability, in the last stage of the game, to reveal a single player's card, to see if they are the traitor or not.
After a certain point, people have to draw cards to replenish their hands, these come from a mostly non-traitor but partially traitor randomly shuffled deck. Now the traitor will start drawing non-traitor cards, and some others will draw traitor cards. This is like the growing influence of the traitor!
Later the team can start drawing single cards from a player's hand, up to the number of victory point they won earlier, which of course is in the traitor's interest to minimize. If it comes up traitor, the team wins. If the game ends without anyone revealed as traitor, the traitor wins.
Something like that.
I might be neat if every card had both a beneficial team effect, and an effect the traitor could eventually exploit. You are creating something on the board, filling in a pattern or something with your various cards. The traitor is trying to put down cards that will eventually benefit their traitor and fitting it into the scrabble pattern of other cards that is, say, building pots of victory points. Make it so that the "stopping the traitor" mechanic is detached from the "team building whatsit" part, so both the traitor and the other players are both trying to build these pots of victory points while also maneuvering to claim them; that way during the sneaky phase, its less obvious the traitor is the traitor because the cards that help the team could also potentially be benefiting the traitor. Maybe the players have a super, super limited number of "fuck the traitor" cards they can put down in the pattern to block the traitor from claiming them, and perhaps they are temporary so putting them down too early, or if you aren't sure who the traitor is, is a waste.
If you have a hand of cards, the capabilities of the traitor are obfuscated, so the traitor can play closer to optimally as a traitor without giving away their role, and play well as a player while still benefiting themselves as a traitor. There's also the potential to make the traitor traitor-ing a reveal on the traitor's part; they suddenly laugh evilly and collapse the house of cards they have been secretly building, and all the players scramble into damage control mode.
As the game progresses, the traitor will lose some traitor cards and gain non-traitor cards. They will never become a non-traitor, because one of the cards won't be able to be played at all, and that will always have enough weight to make whoever has the card a traitor.
Yet at the same time, others may pick up traitor cards, and the weight of those cards will beat out their non-traitor cards.
I'm not sure the start would be more fiddly than other games. For 5 players:
1. Select the 7 traitor starter cards.
2. Shuffle the 28 non-traitor starter cards.
3. Place a traitor card on the table.
4. Place 4 non-traitor cards on it.
5. Repeat steps 3 and 4 six more times.
6. Cut the deck, put the the top half under the bottom.
7. Deal the 35 cards to the 5 players.
8. Shuffle the rest of the non-starter cards (both traitor and non-traitor) and use that to replenish the hand after each play.
It's really not so hard. It will be immediately obvious if it didn't work, as one player will have all traitor cards and everyone else will have none.
The problem this is solving: How do we give one player all-one-kind-of-card and everyone else random other cards? And how do we do that in a way that nobody knows who got all the traitor cards? One way would be to sort all the card into 4 non-traitor piles, and one traitor pile, but then how do you shuffle those sets of cards without mixing them?