Flash Duel: *Betrayal* at Raid on Deathstrike Dragon
Wednesday, November 9, 2011 at 12:40PM Flash Duel's new raid mode is cooperative, in that you team up with up to three other players to defeat the dragon. The mortals win as a team or lose as a team. There's a common problem with cooperative games that a dominant player can bark orders at everyone else and basically play the game solitaire. What to do about this?
Deathstrike Dragon tells you what to do!
The most common answer is to do nothing, and "play with different people." Another common answer is an infeasible and sloppy one: rules about how you can't share information with people on your team. Another answer is to have a secret "traitor" on the team, so you can't trust everyone's advice and you have to think for yourself. Finally, there's a very uncommon solution used by Space Alert and Wok Star where there's time pressure (meaning the game takes place in real-time, not turn-based!) and that there simply isn't time for a single person to do everyone's job. In video games, of course there's the solution that your instructions don't replace the skills of other people (hey, just get all head-shots in this FPS!) but we're talking about board games and card games here.
Let's talk about the worst solution first, the one where the game claims that you can't share information. If you're experienced with tournament rules, hopefully you immediately see the problem here. You can "give hints" but you can't say what cards you have? Like "I have a high card" might be ok, but "I have the Jack of Spades" is not? A hint is actually identical to saying the card in high level play. You give enough hints, or you encode information in the hints to make that so. You can also tap your arm or your forehead to pass information, or other such signals. The point is that there is no real way to stop this kind of stuff. In fighting games, it would be like saying "don't use a certain move *too much*" or some such fuzzy, non-discrete, unenforceable thing.
Fuzzy Rules and Battlestar Galactica
Another example of how this type of solution is sloppy and infeasible comes from the game Battlestar Galactica. In that game, each player submits a card face down to a pile that represents a team effort to complete a task, then two extra random cards are added. This allows a traitor to sneak in a card that will hurt everyone, then he or she can claim that card must have been one of the random ones when everything is revealed. Ok, sounds fine at first glance. But what about sharing information? The rulebook says this:
Skill Cards and Skill Checks: Players are prohibited from revealing the exact strength of cards in their hands. They may use vague terms such as “I can help out on this crisis a little bit,” but they may not make more specific statements such as “I am playing 5 piloting.” In addition, after a skill check is resolved, players may not identify which cards they played. The reason for these restrictions is to keep hidden information secret and to protect Cylon players from being discovered too easily.
One player who is not the traitor should announce the following strategy. "I am not the traitor, and it's in my interest to expose the traitor. If you are not the traitor it's in your interest too. If you do not do what I'm going to say in a moment, you must be the traitor. What I'm about to say benefits non-traitors and exposes traitors, so there is no reason to not to go along other than being a traitor. We'll all "hint" at the cards we're going to play, and of course hints and just saying the card are the same in high-level play. Then when the cards are later revealed, we see if every card claimed to be there really is. If anyone lied, they are the traitor. If anyone was intentionally too vague with hints, they are the traitor. (The game pushes us all asymptotically close to the taboo tactic here.) Note that it's possible that a lying traitor could get lucky and his lie matches a random card. That's no matter though because if the cards *don't* match, then we definitely know the traitor. We'll just do this every single time, preventing the traitor from ever doing anything."
Is that a fun way to play? Not really. But that just highlights the problem. Playing well breaks that game because rules trying to limit communication between people who really want to communicate don't really work. Playing that way is also "against the spirit of the game," but with a squishy information sharing rule, playing against the spirit of the game is just playing well, really, and that's a problem too.
A Better Way to Handle Hidden Information?
The problem is that it's infeasible to give players an incentive to share information, then claim that they can't. A better way to handle this is to attack the problem at the incentive level. Make the players not want to share information. Either way, the goal is to make it so not every player knows everything so that players have to think for themselves rather than rely entirely on the advice of the loudest player. If we can give people some reason they don't *want* to share information, we don't have to worry so much about all the annoying stuff above.
We need a traitor who gets his power from information. On the one hand, the more information you share, the better off your team is because you can all plan together. On the other hand, the more information you share, the more powerful you make the traitor, so you should not share everything. The moment you hold back sharing anything, we've already solved the dominant player problem.
In Flash Duel, the hidden information is the cards in everyone's hands. Remember that these cards just have a single big number on them, like a "2" or something, and that you only have hands of five cards. The traitor has a special power where he can voluntarily reveal himself and then attempt to kill off the mortals by naming the cards in their hands. If the traitor can name every card in every other mortal's hand, he kills them all. This would probably never happen in a real game though, because players will know that showing their hand cards can be deadly. What this really does is keep information sharing in check.
By the way, the revealed traitor then fights alongside the dragon, so he's not out of the game when he reveals himself.
The Betrayal Mode
Last time I wrote about the Raid on Deathstrike Dragon mode, the one without the traitor. In that one, the answer to the dominant player problem is "just try to work together and don't play solitaire." But a dominant player certainly could ruin that experience, as with almost any cooperative game. The Betrayal mode is a harder version of that raid where one of the mortals (or zero of them, but you won't know that!) is the traitor.
In playtesting, several players actually preferred the regular raid over the betrayal raid. The regular raid is a bit simpler, and if you are all getting along and cooperating anyway, there is no problem to fix. That said, I think the betrayal mode will really appeal to certain groups in that it's an extra challenge, and really different (and treacherous!) game dynamics. If you're ready to turn your Flash Duel up a notch, then try it. See which mode your group prefers, and feel free to post your experienes, questions, or hype on the boardgamegeek.
Flash Duel 2nd Edition ships in early December.
In closing, have another dragon card image:
Reader Comments (31)
GlenH, I think we have a major misunderstanding here about what I'm talking about.
First, you said the situation I talked about is not feasible BECAUSE It would make the game unenjoyable. Wrong cause and effect. It is feasible and yes it makes it makes it unenjoyable. That it makes it unenjoyable doesn't stop you from doing it, therefore it is unrelated to in feasibility.
Next, you said BSG is not really designed to stand up to this kind of play. Yes I know! And that is the entire point of my post. I agree with you that it isn't. And I'm saying that that's kind of sloppy, and it would be better if rules were more solid rather than less solid. So I have attempted to take a step toward the more solid and pointed out that the entire set of solutions having to do with banning "some kinds of communication" are always going to be less solid.
As I said in my post, if the best way to play something is to make it totally unfun, then that is a rules problem. It's not a problem with the players, it's a problem with the game for not being good enough to stand up to the kind of play the players did.
Finally, the strongest disagreement I have, and this is really the whole basis of everything here, is that the BSG situation I described is NOT analogous to a cheat code. There is no solid rule that prevents what I said from happening. There is a vague rule that is unenforceable and non-discrete, and the result of that vague rule is that it's possible to use "hints" to communicate full information and not break the letter of the law. So the entire problem is that it's NOT cheating, while it should be. So to fix this problem we'd need a way to actually clearly define cheating (we can never do this if you allow "hints" but nebulously don't allow the consequences of giving good hints) or we can attack the problem at the more base level by chaining the player incentives to begin with.
I do agree that players who are not playing optimally could have a lot of fun. But my standard of ok-rules is much higher than that for a game to be acceptable to me. If reading the rules one time leads to such an obvious problem that destroys the fun of the game (while still playing legally!), then some better rules would be in order.
I'm surprised at all the defenses of BSG. The game actually makes it incredibly easy to do what Sirlin describes. The one time I played, I just got the military leader position, declared martial law, threw every other player in the brig, and then let them out one by one using a system of hinting like this to build trust between the humans. Worked perfectly, Cylon was forced to reveal and leave the ship upon entering the second system. (which is really, really early in the game for people who don't play it).
It was incredibly easy to do, and seemed more or less optimal. We never played again because the 2 hours of combat resolution needed to actually end the game is boring.
On the BSG thing, one point of miscommunication might be that some people are looking at the secrecy rules in the rulebook and others are looking at the expanded/clarified secrecy rules in the FAQ.
Under those rules, only binary statements are allowed. A skill card can be identified as high or low, a destination as good or bad, a loyalty card as Cylon or human. The rules are still a bit fuzzy, sure, but identifying the exact cards you play is definitely against them.
In general, yes, it would certainly be preferable for games to avoid the situation where players have access to information, would benefit from sharing that information, but are forbidden from doing so by game rules. Or, worse, discouraged but not actually forbidden. I encountered this recently during a game of Werewolf.
I play Werewolf online, in what would probably be considered high-level competitive play. I was recently at a games day where a face-to-face game of Werewolf was played. While the moderator was explaining the set of role cards he'd included, I realised that his setup included too many special characters, and hence could be broken by the simple of act of everyone claiming their roles.
Now, there's no rule against claiming your role in Werewolf, and in certain situations it's expected. But the strategy of having everyone claim right at the start of the game (what we call an "all-claim") is generally considered unfun. I demonstrated this by badgering people into claims and solving the game before it really began.
Now, in the online community I play in, we balance our rolesets such that the natural consequence of the all-claim is not a solved game, but a game in which the traitors are still hidden and the special characters are likely to be eliminated without using their abilities. In other words, the unfun strategy is discouraged by making it ineffective for the side that would wish to use it.
The face-to-face casual group, however, decided that they wanted to carry on with the same sort of setups (ones that made the all-claim a highly effective strategy) and instead agree not to all-claim. Essentially, continue to give one side a game-breaking advantage but have a "gentlemen's agreement" not to use it.
Fortunately I had to leave to catch my train. I suspect it would have been a frustrating experience to stick around for further rounds.
I'm not sure you're getting what we're saying though. You're saying that identifying the exact card is against the rules. We're saying that identifying the exact card is both against the rules and possible, hence the problem. It's just inherent that allowing some communication is equivalent to allowing all communication amongst highly skilled players, so the entire notion is kind of infeasible from the start. And this leads to me trying to attack the problem from a different angle, by going after the incentive, not band-aids on the part about "you can sort of communicate but not really."
I guess I'm confused by the same action being described as 'against the rules' and 'not cheating'. To me that seems contradictory. If the rules forbid the sharing of certain information, and you deliberately share said information using hints or codes or nonverbal signals, aren't you cheating? Is this a definition-of-terms issue?
Now, that such a rule is required in the first place is certainly problematic, and 'band-aid' is an apt description for it. The non-traitor side has access to certain information and gains an overwhelming benefit from sharing it, which renders the game not-fun. Thus a kludged rule is added forbidding sharing the information.
I don't know if you're familiar with a game called The Resistance, but it essentially revolves around that same core skill-test mechanic from BSG. A group is chosen to attempt a mission, they each secretly add a card to help or hinder, then all the cards are shuffled and revealed. Now, they fixed the BSG problem by taking the information away - there are only two types of card, one to support and one to sabotage. The non-traitor team simply doesn't have the ability to identify their cards like they do in BSG.
Somewhat amusingly, they then managed to then add the same problem back in by giving each of the non-traitor loyalty cards unique artwork (as opposed to the identical ones from BSG) and necessitating a band-aid rule forbidding discussion of the loyalty cards.
There's another communication issue in The Resistance as well. One of the toughest decisions in the game is whether a traitor should sabotage or not when there are multiple traitors on a mission. Because the traitors can't communicate openly, they have to make that choice independently, and if either both sabotage or neither sabotages it can be devastating.
But it's trivially easy to come up with a meta-rule about who has priority over the sabotage decision; say, the first player alphabetically. If you're a traitor it's optimal play to use such a system. And it's completely impossible to rule against this, even with a band-aid, because it doesn't require any in-game communication at all. Once it's been said once, you can't unlearn it and you don't need to repeat it.
Being confused about something being possible by the rules yet against them is the correct state of affairs here. Yeah that is confusing isn't it? And it goes to show that you can't feasibly have a rule saying "you can share some info" and another rule saying "but not too much" because the first rule actually lets you share everything if you try hard enough (if you're playing to win). This is the entire problem, and it's why any cooperative game that is claiming to have this rule is doing it in a sloppy, defeatable way. And it's why I have suggested a different approach.
Interestingly this same topic came up at the NYU game conference called Practice. One person asked the crowd to come up with solutions to the dominant player problem in coop games, and as part of his question, he expressly forbid any rule of the form "you can communicate....but you can't really communitate!" He took it as given that this is stupid and sloppy and he was looking for another answer. An answer that allows general talking to still happen and that doesn't involve the (perfectly fine) time-pressure solutions seen in Space Alert and Wok Star or traitor mechanics. The crowd produced no acceptable answer. Then I gave my answer which was "This sounds like some useless statement, but it's a serious one. There is no known solution to your problem." I later talked to him about the Flash Duel 2 solution, of not just any traitor mechanic, but specifically empowering the traitor to kill when he gathers information, and he thought that actually was an advancement of the field, possibly.
The point is, we shouldn't be stuck on "there isn't a problem to have squishy rules about info sharing that actually don't work when playing to win." Better to move on and try to find solutions. I offered one, but other even more clever ones lie undiscovered, I think.
About The Resistance, yes I am sightly aware of it, but I didn't know the bit about the artwork giving away which card is which. Ha!
Just a few things on BSG:
The rules, while restricting communication, aren't exactly squishy, except in that there isn't really anyone to enforce them. They are mostly strict in a binary sense. High/low, good/bad, yes/no, etc.
And, there is some disincentive to revealing information to a traitor (if each player basically says what they are contributing, a traitor who votes last and/or has an ability to alter the vote could choose the right amount to sabotage. Also, in a game with 5 or more players, there are multiple traitors, who benefit from know who the other traitor is, and knowing who aren't traitors to the cause).
That high/low thing isn't even in the original rulebook, it's like a later patch. And yeah the rules really are squishy because that's the nature of saying "you can talk all you want, but there is certain types of information you can't convey." If you try hard enough, you really can convey it and there isn't really a feasible way to stop it. If you are playing in isolation booths and have just a high / low button to push that sends a signal remotely, yeah that's enforceable and binary. But not when you can have a conversation and say any old gray area thing you want, and make gestures, and who knows what kind of things to circumvent the rules. It all goes to show that stopping the incentive to communicate is a better approach than any rules at all that attempt to squishily allow communication and also stop communication. There's just no good way to do that.
(Edit: removed the part of my post that accidentally responded to a comment on page 1 again)
By the way, the game described with with co-op win and super win, is Defenders of the Realm (eagle games).
http://boardgamegeek.com/boardgame/65532/defenders-of-the-realm
Also with super win in incredibly cool surprise backstabbingly way, Castle of the Devil, in which through concealed hand and card play, a player can betray his team to win on his own.
http://boardgamegeek.com/boardgame/25951/the-castle-of-the-devil
great posts, following trail from Penny-arcade, cheers to all!
GS_topcow, hey that's helpful, thanks for pointing out those other games.
Due check out devils castle, its insanely fun, also The resistance is worth a look:
http://boardgamegeek.com/boardgame/41114/the-resistance
Your blog and designs are an inspiration man, gad to be of assistance.
GS