Puzzle Strike and the Quest for a Tournament Quality Deckbuilding Game
Monday, April 23, 2012 at 3:50AM Just a reminder that Puzzle Strike Shadows as well as Puzzle Strike 3rd Edition are up on Kickstarter right now. Let's delve into the design in excruciating detail!
Casual versus(??) Hardcore
"Casual" and "hardcore" can be a false dichotomy. Which one is World of Warcraft, for example? It's pretty casual friendly, and yet it's not at all casual to the hardcore raiders that spend literally more hours than a full time job at the game. Likewise, Puzzle Strike is pretty casual friendly, having kid characters, a pink box, and fairly easy rules. I'd like to talk about the casual side of Puzzle Strike in another post, and the several ways we're really turning up the casual appeal even more in the future.
But for now, I want to tell you purely about the hardcore side--about Puzzle Strike as a serious, competitive, tournament game. Make no mistake, one of the missions of the game is:
"For Puzzle Strike to be the best competitive deckbuilding game there is."
Ok great mission and all, but how do we accomplish that mission? Let me tell you all that's gone into making that happen, and the challenges we've faced along the way. Here's the criteria that have always been at the heart of the project:
- Asymmetric design
- Player interaction
- Getting to the meat of the game quickly
- Strategically interesting dynamics
- Exciting moments built into the system
- Balance of "viable options during gameplay"
- Fairness of the asymmetric choices
Asymmetric design
In Puzzle Strike, you start by choosing a character. Each character has different abilities, and allows for different gameplay, and appeals to different player personalities. I've been involved in competitive scenes for games for a long time and the excitement added by having a cast of characters to choose from is enormous. The 2-player version of the base set alone has 55 different character matchups, while the expansion brings that to 210 different matchups. There are so many nuances to knowing how to play all these matchups differently, that symmetric games feel flat by comparison. Even apart from the big gameplay advantages of asymmetric games, there's a boost to the player community by having so much to debate and explore. Different characters also allow different players to find their personal playstyle in at least one of the many options.
Though your opinion may differ, to me, a symmetric game would be a non-starter here, as in not eligible to even consider as the best deckbuilding game for pure competition.
Player interaction
There's a reason to have games with low player interaction. Maybe you'd rather all play a mostly solitaire game without having the "harshness" of directly competing. Even in games with low direct interaction, there can be indirect forms of interaction. That said, this is not a great recipe for a real competitive game. The more player interaction there is, the more opportunity there is to display the kind of skill that should matter in a competitive game. A game with literally zero player interaction would still require skills of course, and those would probably be the skills of optimization. It's just that a race of several non-interactive players optimizing is a missed opportunity when instead we could have a game of very high interaction, allowing for maneuvers and counter-maneuvers.
I've heard the terms "contested" and "uncontested" skills used, here. Uncontested skills are the kinds your opponent can't do anything about. In a video game like Street Fighter, that would be the part where you perform a difficult combo, for example. Contested skills are the kinds your opponent CAN do something about. In Street Fighter, the example would be getting at just the right range to do your move, because your opponent can move his own character to affect that range. While uncontested skills can certainly exist in a good competitive game, the focus really needs to be on constested skills, at least if long-lasting tournament play is the goal.
In Puzzle Strike, the "crash" mechanic builds player interaction into the core of the game. You are trying to fill up the other guy's gem pile full of gems, and you do that by "crashing" (breaking) gems in your own pile and sending them to your opponent. He can "counter-crash" to stop those incoming gems. He might want to because doing so actually ends up removing gems from the system which slightly lengthens the game. Counter-crashing this way also doesn't cost an "action," so that's another reason to do it. But there's reasons on the other side, too. Simply accepting those incoming gems and crashing on your own turn would require spending an action, but it would also yield you a bit of money to buy some better chips. And it would NOT remove gems from the system, so if you're in a good rushdown position, this might be a better option.
The point is that this kind of direct interaction is at the forefront of the game. Also, the red attack chips all have big effects on the game, and the blue defense chips have pretty relevant effects too. You are very often faced with decisions how about to respond to your opponent, and whether you should try to disrupt them, rush them down, or hang back and build your own economy. All the *indirect* interaction that's common in deckbuilding games is still there too, of course. The part where your choices of which chips to buy depend on which chips you see your opponents buying. Luckily that's not *all* the interaction though.
Getting to the meat of the game quickly
In Puzzle Strike you start with your three character chips in your deck, so you can play those starting on the very first turn. In some games, you start with basically blank cards and it takes more turns to get into the real meat of the game. This might sound like a small point, but in a tournament game, it's actually very important to use every minute of gameplay to its full extent, or to cut that gameplay. If you want to run several games in series, it's kind of boring if the first few turns of all those games take a while to get things going, so it was a very conscious decision to give players character chips they can play right away, even before the buys from the deckbuilding start to kick in.
Strategically interesting dynamics
Of course a game has to be actually interesting to play on a strategic level if it is to be a long-lasting competitive game. I wrote an article about how difficult it was to arrive at interesting dynamics that weren't degenerate.
The short version is that the money system, the purple chips that manipulate the gems in your gem pile, and the delicate balance between rushdown, building econ, delaying the game, and ending the game were tough to get right. It's tough because the game system is interconnected, meaning that just about everything effects everything else. It's easier to balance a game if you have some subsystems that can be adjusted without messing up everything else, but if you do manage to get such a dense system to actually work, it means an even richer strategic playground to play in.
I also was the lead designer of Street Fighter HD Remix, and balancing that game was challenging, too. It was based on a game that had been played heavily in tournaments for 14 years, so changing anything about balance at all is a bit like threading a needle. Also, if you change anything about a character to fix a specific matchup, then it will affect all the other matchups. At first glance, it means the system is so interconnected that it's damn hard to work with. But in Street Fighter, it was actually possible to use a lot of tricks to make that balancing challenge easier. By thinking hard enough, many solutions to balance problems in a matchup could have minimal effects on all the other matchups.
One example is Dhalsim vs. Guile. If you aren't familiar with Street Fighter, Dhalsim has stretchy limbs that reach across the screen, while Guile often likes to stay back and throw his projectile called the Sonic Boom. This was a problem, boring match. One champion tournament player suggested that barely changing the hitbox on one of Dhalsim's stretchy punches would mean the difference between it getting a clean hit against the Sonic Boom and trading hits. And that one change would really improve the gameplay of the match. Changing that hitbox had very, very little effect on any other match because it meant changing something on the backside of the character, in a place where fighting moves don't usually interact anyway.
I'm not sure if you followed that, but the contrast is that in Puzzle Strike, there are usually no such tricks available to us. Every damn thing affects every other damn thing, which means a lot of work on the development end, but also a lot more ability of the player to affect the game with nuanced play than there would otherwise be.
Exciting moments built into the system
That last section might have sounded a bit dry. Although strategy is very important, there has to be excitement in a competitive game. Now that we understand games more than the olden days, I think we know that when making a competitive game, we want to build in exciting moments into the system. I don't mean to force them artificially, but to create a game system that we know is likely to generate exciting moments.
In Puzzle Strike, there's a comeback mechanism that's modelled after the very interesting comeback mechanism in the video game Puzzle Fighter (I was also lead designer of Puzzle Fighter HD Remix, by the way, so it's no surprise I chose this theme for Puzzle Strike!). Anyway in both games, when you have a lot of gems in your gem pile, you are closer to losing in some sense. If your side fills up to the top, you lose. In another sense, you're doing just fine though. One reason is you have more ammunition to fire back at the other player. And on top of that, both games have a "height bonus" that gives you an advantage for having a lot of gems. That means there's a push-your-luck element there, which also helps as a comeback mechanism. In Puzzle Strike, the height bonus allows you to draw more chips per turn the higher your gem pile is. So when you're close to losing, you can do even bigger combos.
Another conscious design decision to increase the drama of the game is WHEN the win condition is checked. The basic idea is that if the various kinds of gems in your gem pile add up to a total of 10 or more, then you lose. But you don't instantly lose; this is checked only at the end of your turn. You often go over that limit, then on your turn manage to save yourself and stay in the game. It gets really exciting when your opponent sends you way, way, way over that limit of 10, and you somehow manage to pull off an amazing turn to throw it all back at him. This isn't an accidentally exciting moment though--it's there on purpose as an example of designing excitement into the game system.
Balance of "viable options during gameplay"
In my series about balancing multiplayer competitive games I talk about the difference between two different usages of the word "balance." Sometimes people mean balancing the set of options available during gameplay. Both symmetric and asymmetric games have to care about that. If there are several kinds of moves you can make, but all of them basically suck except one kind, then that isn't "balanced" in a sense.
Asymmetric games have to deal with that AND then also deal with making sure the different starting options (in our case, all the characters) are fair against each other. Let's talk about that first kind of balance first, though: the viable options during gameplay, regardless of there even being different characters.
The article I linked earlier touched on the challenges of getting this kind of balance to work. After releasing the game, we've had a whole lot experience with it though, and have seen across dozens of tournaments exactly how different strategies are used--or not used. And there has been a threat to balance of viable strategies we've been facing for a long time. The 3rd Edition of Puzzle Strike (and the Shadows expansion) make one change--one seemingly small change--that has a huge effect across the entire game to address. But first, what is the problem?
The problem is "mono-purple." That is, the strategy of ignoring most of the bank and buying only the purple chips that directly affect your gem pile. Playing in this way is kind of short circuiting the game, avoiding big swaths of it. That could be fine depending on how powerful such a strategy is. So is it powerful? Well, yes and no. Some characters in the 2nd Edition tended dangerously close to mono-purple power, while others used more diverse strats. Then we released the Puzzle Strike Upgrade Pack to address that. The situation was much improved, as more diverse strategies were viable than ever.
In developing the expansion though, we were very often faced with too small of a design space. We make an interesting character, but then the game system's reward for playing in the boring mono-purple way is a bit too much unless we take specific steps to fight that with various extra clauses on lots of chips that punish such a strategy. It also left us very little design space in which to create new puzzle chips (those are the ones in the bank that change every game). If a puzzle chip is too weak, people will ignore it and just buy purples. But purples are so strong that when we turn up the power of puzzle chips to compete, they often have to be so strong as to be game-breaking if they are tuned just a hair wrong. What we need is more breathing room here, more space to create chips that are of a reasonable power level compared to purple chips.
In another article, I talked about how I looked toward Starcraft for an answer to something, and their model of late-game units like Carriers that could smash early game defense sparked me to create uncounter-crashable 4-gems in Puzzle Strike. So again, I looked to Starcraft to answer our troubles here. Our trouble is that a player who buys only purples is trying to end the game as soon as possible--he is doing a 6 pool zergling rush, or something. But if the opponent holds off this rush, he is no better position.
In Starcraft, the rushing player would have a big economic disadvantage, so there is more of a tradeoff in whether to rush. What makes matters worse is that in Puzzle Strike, it's not really even analogous to the rushing player having zerglings. Those are early game units that fade in effectiveness later (yeah, yeah they can be upgraded in Starcraft, but that's beside the point). Anyway, all those purple chips in Puzzle Strike are just as good late as early. So it's like rushing for no economic disadvantage with hydralisks or mutalisks or something that you can win the game with later anyway.
This maneuver needs an economic disadvantage for the system to make strategic sense. And now it does.
The Combine chip (the basic purple chip that combines two smaller gems into one bigger one) now costs $1 of in-game money each time you play it. If you buy and play only this one chip over and over, you are rushing to end the game, but if your opponent buys just one or two to hold you off, he will be able to survive and extend the game. At that point, you will have spent several turns buying low cost chips, while his economy was not really affected, so he will have better tech going into the mid-game.
Along with this change, we also adjusted several other chips to allow for rushdown to still be possible, just in a way that requires actually using your character chips and puzzle chips from the bank. Overall, in high level tournament play, there's a more diverse set of viable strategies now. Rushing, econ, disruption, and engine building strats all coexist.
Fairness of the asymmetric choices
Once the game system works, we need to have a set of fair characters. That is, no character can be too good or too weak. Too good is a much worse problem because that invalidates all other characters. Too weak is just minorly unfortunate because no one will play that character. After years of iteration based on tournament results, I think we're in good spot now. Twenty different characters(!) that all seem to have their uses in high level play, without any particular one of them dominating too much.
I could go on forever about the balancing process of these characters, but instead I'd rather talk about the goal of even balancing them in the first place. It seems that most games are interested in releasing more and more and more content. Like expansions every three months. New, new, new. I'm not interested in that at all, and it actually runs counter to the goal of creating a highly polished competitive game. Instead of adding more and more, we are zeroing in on a better and better game. Each iteration has been more polished than the last, better gameplay dynamics, and better balance. If we simply add more and more, yeah that appeals to some players, but it doesn't actually produce something legitimately great. It means instead of fixing whatever issues older chips / cards have, we would be waiting for them to rotate out of tournament play. We'd be forsaking those earlier sets and letting them lie with whatever issues tournament play had uncovered.
I'd rather give you all the very best versions of my games that we're able to produce, at that given moment. And with years of development effort now spent on making Puzzle Strike 3rd Edition (plus the Shadows expansion!) the best competitive game it can be, I can truthfully say that this is the best version we've produced so far, by a big margin. I look forward to seeing the competitive scene grow, and for years of Puzzle Striking to come.
You can also play Puzzle Strike at www.fantasystrike.com for free, by the way. Some players have logged THOUSANDS of games of Puzzle Strike there, and there are tournaments all the time, in addition to casual play. Thanks to the entire community of players who have all contributed to refining the game into its current awesome state.
If any of this sounds good to you, get in on the Kickstarter project for the Puzzle Strike Shadows Expansion as well as the 3rd Edition base set.
Thanks!
Reader Comments (22)
awesome article, Sirlin! A lot of awesome perspective into how you actually built the game. :)
Just out of curiosity, what is the current state of Yomi? Will there be a similar set of balance updates and so forth, or is it considered in a balanced state now? Fwiw, I would have backed a project similar to your 135 dollar option with the Yomi game added in (with the additional cost obv), as I already have Flash Duel and didn't want to more than double my price just to get Yomi.
Anyway nice work; I'm looking forward to getting Puzzle Strike later this year!
David: Two paragraphs before "Fairness of the asymmetric choices", you suddenly use "tech" in a sentence in a way that wasn't alluded to anywhere in the article. Using "chips" instead would probably make it clearer to people who haven't read your previous articles.
Nathan: Other kickstarters (Steve Jackson's Ogre, to name one David's already referred to) have created "add-ons" before, where you pledge more than your chosen reward and pick something else that you want to add; maybe David can accommodate you that way.
Good stuff, as always. Can't wait to read the strategy guide, now that I've skimmed the drafts.
So, does that mean Puzzle Strike is basically done? I mean, is the third edition going to be the last? Or is there a chance you'd make a fourth edition at some (much) later date?
Nathan, Yomi is considered very balanced now. No idea on if we want to make it even better or just do nothing and release an expansion. In any case, pretty long before *anything* will be released. I shudder at the thought of how much work to create 10 new card backs, more than 10 new boxes, and art directing 100+ pieces of art. Oh and also balance the new characters! ;)
Albey: PS done? Maybe? I mean who knows, it seems great as of this moment. We've worked on it so much, it would be nice to let it be and work on other things though. There's Yomi plus at least one other game you've heard mention that I should be getting to, though I haven't had much time for either one lately!
It seems like it would be possible to make a tournament-worthy game without very heavy player interaction, though.
Take the deckbuilding game that I'm enjoying at the moment -- Dominion. It's possible, through random setup mechanisms, to have a setup where there aren't any attack cards (witch, militia, etc.), and thus there's practically no player interaction during the course of that game; yet, the game is still quite enjoyable as it becomes a rush to see who can build his/her economy the fastest in order to get the provinces/colonies, and it doesn't become stale because there are a number of ways to go about doing it.
Granted I don't know how serious the Dominion tournament scene is, just playing it casually.
Just my two cents.
I think you'd find that Dominion is much easier to "solve" than Puzzle Strike, making any tournament situation less serious.
Peacewood: Imagine a game that had no player interaction at all and where all players had the same starting conditions. If all players started with the same resources and had no disruption from anyone else, then the competition would really just be one of optimization ala "Who can figure out this puzzle in the fewest number of moves?"
In this particular imaginary game, the optimal, "correct" solution would be the same every single time, even if it might be difficult to figure out what that optimal solution is. Many games without much player interaction mitigate this problem by adding randomness, so that the particular puzzle being solved is different with every playthrough.
But if a game instead chooses to include heavy player interaction, it allows the game to be different with every playthrough without adding randomness. In Yomi, for instance, the 10 character decks never change at all, and yet each match can still be interesting because your optimal play is heavily based on the actions of your opponent.
So in a more theoretical sense, a game with player interaction is much harder to optimize for than a game without player interaction, meaning that the gameplay will be more likely to still be interesting even if people's brains one day evolve into supercomputers.
But on a personal level, I simply prefer games with heavy player interaction because it allows the player's personalities to shine through so much more. That kind of game design insists that you care about what the other players are doing, and you get to see how the other human beings around you react when you poke them with a crash gem or whatnot. And I think that's kind of a big deal. :D
I also think more importantly, in a game with player interaction, it matters who you're playing. In a game without player interaction, I could be playing an expert, a computer bot, an AI, nobody or even 500 people at the same time and it wouldn't change my actions. That feels to me a whole lot less like competing with someone and a whole lot more like competing against myself with them happening to be there at the same time. I used to be a competitive runner, so I'm not saying there isn't merit in that, just that it isn't strategically or tactically very interesting.
Awesome article, David. As an avid gamer (both video games and board games) I like that you can draw clear parallels between highly competitive video games and boardgames. Do you find that a lot of your design inspiration comes from video games? Are there any non-competitive games that have influenced specific mechanics in your games?
Keep up the good work. I can't wait to get my hands on the 3rd edition + expansion.
Thanks Gordo. I've worked on the design of many things over the years, and many were outside the realm of competitive video games, but many were video game projects that never ended up shipping, for reasons unrelated to design. That was one reason to make board/card games, so there's less uncertainty about being able to ship them. And I've only shipped 3 so far, all of which are competitive (though Flash Duel is also cooperative). While there are a lot of mechanics I like in various games, including ones not about competition, the games I've made so far are pretty directly influenced by competitive games.
The one I'm planning after Yomi is the most intensely competitive of any of them, actually. Maybe I should make a pure co-op game to balance it out!
that would be very interesting.
One thing I haven't enjoyed about some co-op games I've played (Shadows Over Camelot, Pandemic, Flashpoint - Fire Rescue) is that they sort of devolve into one person calling all the shots. Space Hulk : Death Angel has some mechanics that allow you to get around this problem, but they're easy to play around and that's also not a very good game unless you really like Warhammer 40k. Do you find this to be a problem with co-op games? Any ideas or examples of games that solve this pretty well?
The dominant-player-problem is a common one in co-op games. Space Alert and Wok Star solve it by adding time pressure (measured in real time, not turns). I think that's a great solution, but not for ever game, because a) you might not want that kind of pacing and b) you might want a design that requires more thinking and rules questions about how various interactions work. Several games solve the problem with a traitor, though I think often in very iffy ways. You can read my post somewhere about how Flash Duel specifically attacks that problem at the "incentive not to share info" level rather than the more common "you have an incentive to share info, but there are band-aid-ish rules that say you sort of can't (but if you try hard, you still can)".
There aren't really any other known ways of solving the dominant player problem. That said, one player told me he was enjoying the co-op game Ghost Stories, so I asked him what steps the design takes to minimize the problem. He said it takes no steps at all, and that his particular play group is made up of all smart people who each contribute to decisions, with none of them being dominant. So even if the problem exists in the design, it didn't exist for his particular play group. I kind of wonder if we are stuck with that kind of solution for games where you don't want a traitor and don't want real-time elements.
I should point out the common video game solution to this: that's it takes lots and lots of skill and knowledge to play something right. For example, me just telling you what to do in Counter-strike or even in high level World of Warcraft is no substitute for all the nuances of knowledge and execution that you might have as a player in those games. So I wonder if giving people different roles that build up complexity over time could work. Even if all the game state ino is freely available to your teammates, maybe you have developed the best intuition for how to deal with playing all the nuances of Role X.
Such a game may be too complicated to be popular though, as most co-op games are pretty simple. I can see the point of that in that they are often social experiences primarily. Forbidden Island, for example, is so easy you can play it with anyone. Pandemic is not much harder, and you'd need a vastly more complicated game to have the effect I talked about in the previous paragraph.
Great article. It is nice to see all those aspects that were integral to the design of Puzzle Strike in one place.
I never actively thought about the benefits that “Getting to the meat of the game quickly” has for tournament play. Obviously, the potential for player interaction from the very first turn onwards is a plus: It gives you more opportunities to outplay your opponent by making optimal decisions, while creating more situations where he/she could make bad ones. However, that it is annoying to “go through the motions” by having very similar first few turns again and again in a tournament that consists of many rounds/plays is a pretty good point, too. In retrospect, this is clearly one of the reasons why Puzzle Strike quick matches never get boring, even if I play against the same opponent again and again: There is no sameish set-up to get to the “real turns”. Your first turn is already very important and what you do and what you buy can differ from game to game, even when playing the same character match-up.
Regarding the problems of cooperative games: My favorite solution is to have a semi-cooperative system. The players have to work together to beat the game and they will all lose together if the game wins. However, once the game ends and the players have successfully beaten the game, a winner is determined among them by comparing how well they did individually (maybe through victory points or their general advancement). So, while the players have to cooperate, they can't just follow the tactical advice of a single player, because this will most likely lead to him becoming the overall winner. Each player is inclined to keep some information hidden or to simply not do exactly what others want them to do in order to win. I concede that this is not a perfect solution, because it makes the game play differently than a purely cooperative game, but it is still my favorite one. Maybe I'm just not a fan of purely cooperative games though. ;)
@ Star Slayer
Isn't that just a competitive game then? I mean, the potential outcomes are still:
1) Everyone Loses - I Lose.
2) Everyone "wins" but I'm not the overall winner - I Lose
3) Everyone "wins" and I'm the overall winner - I win
I don't see the appeal of option number two where I am basically enabling somebody else to win the game. In fact, I could see option 2 being more of a way for players who aren't in the running to sabotage the game so that nobody wins. Maybe everyone I play games with is just a jerk, though.
Gordo, it is different from a purely competitive game, because you have to work together in order to not lose. So, while the end result may fall into one of your three categories, the actual game play is different from a game where you solely compete against the other players. The key to victory is to find the right balance between working together and advancing your own agenda. Come to think of it, there are nuances beyond your three options: If the players don't work together, *everyone* loses. You were beaten by the game... how shameful! You should cooperate better next time. If the players beat the game however, you may not make the first place, but you may finish second or third and you didn't outright lose. You achieved something, even if you don't take the first place - you survived / beat the system. You and your buddies did it. Yay!
Also, I think there is something like a scale, with cooperative on the one end and competitive on the other end. Some semi-cooperative games fall closer to the competitive end, like Chaos in the Old World, where you mainly fight against the other players, but where all players lose if they squabble for too long. Others fall closer to the cooperative end (I can't think of a good example from the top of my head, but I've seen a dungeon crawler once that worked this way), where you mainly try not to lose against the game mechanics and have to work with the other players so that you aren't overwhelmed. The latter type of game has a strong cooperative feel, while still maintaining a certain rivalry among the players.
Agreed with Gordo; losing is losing. I actually like the idea of an "everyone loses" condition you have to watch out for, but that's just a competitive game where you're forced by circumstance to work together. That's a tense and fun situation. But that's hard to do well, and I wouldn't put it in the co-op genre.
Some games (maybe the dungeon crawler you were thinking of?) start with a cooperative phase, and later transition to direct competition. That's also pretty tricky, because you'll have some players who don't want to cooperate at all, and some players who seal their fates by cooperating too much, and sometimes the whole thing falls to table politics. "True" co-op with a competitive element is basically impossible unless everyone plays suboptimally.
"There aren't really any other known ways of solving the dominant player problem."
When making broad proclamations about an entire genre of games, stick to the medium you actually know ... video games. There are numerous co-op games that deal with the dominant-player problem in ways other than those you describe, including Hanabi, by Spiel de Jahres winner, Antoine Bauza.
Hi Ben, that's really condescending of you and an inappropriate way to talk to another person. I suggest straightforwardly contributing information, if that is your goal.
This question was actually discussed quite a bit at the NYU Practice conference about games (which included board game players and board game designers). The result of the discussion was that even though many things have been tried in co-op games, there haven't really been any good answers to this yet. It's not that just I personally cannot think of or am not aware of any. Traitors and real-time gameplay and scores amongst the winners are all answers, though they don't quite get to what we want. (Well the real-time one does, but has other problems.) Games with "fake hidden info" aren't a solution to the problem.
Hanabi actually contributes nothing to solving the problem. To restate the problem, the issue is that if all players have access to all the same info, one player might dominate the decision making process, and further that it's not reasonable to have hidden information in such a game.
For example, we could attempt to solve this by giving everyone a hand of cards, then saying these cards are secret and can't be shown to other players. An excellent point brought up at the NYU Practice conference is that we also assume we want a game where socializing is possible and encouraged. That means expect everyone to talk to each other. And if we say "communication is allowed, but not too much communication" then we're using sloppy rules. If you're just playing for fun, yeah the sloppy rules can be enough. And if they are enough, then simply having hidden information would be fine.
I think (and several others in this discussion also thought) that a less sloppy solution would be better. Something that doesn't pretend to hide info from players, while simultaneously giving them incentive to share info, and band-aid rules that supposedly stop that communication...yet allow communication in general. I think it would also be counter-productive to say the sloppy solution isn't sloppy, as it would lead to not even looking for a more solid answer. Anyway Hanabi is in the same category here as any other game that would have supposedly hidden information. If we played it isolation chambers with no social communication, then yeah it would be a great solution (as would any game with hidden hand cards). But the actual goal is solving the dominant player problem with solid rules while still allowing communication.
Yeah, David, a less sloppy solution is what I was getting at originally. The game I mentioned earlier, Space Hulk: Death Angel, tries to solve the problem with a hand full of cards that are secret actions you choose each turn. The problem there is that each player only has three cards and you can't play the same one twice in a row, so between that and sloppy rules about communicating the problem doesn't really get solved.
I think traitors is an alright solution for some games, especially a game like The Resistance. I guess that's kind of a competitive team game though.
@StarSlayer, I think it's a bit of a stretch to call Chaos in the Old World cooperative. The "everyone loses" scenario in that game has been pretty rare for me in my games, and it doesn't seem to me like it was put into the game to encourage cooperation, but rather as just a fun alternate ending for the game. That's just my feeling about it though.