First off, do you ever ask yourself if the world really needs one more game and, along those lines, why do you keep making games yourself?

James Ernest: Constantly. It''s a running joke around here: "Game Publishers everywhere realize that there are officially enough games, and decide to stop making new ones." But really, this is an entertainment business, so it is reasonable to argue that the public will always want something fresh, even when "everything" has already been done. There are also new techniques in game design that make today's board games mechanically better than those of ten or twenty years ago, and I think that makes new games genuinely better. Obviously as a designer and a publisher I'm driven to create new games and new products, so it's easy for me to rationalize these things.

What is the most important thing you keep in mind when designing a new game?

James Ernest: Two things. Why to pick it up, and why to keep it. In other words, why will someone play this game the first time (theme, format, concept, etc), and why will he play it again (balance, ease of play, fun).

In the "pick it up" department, I usually start with a theme or a story, not a game mechanic, because I think that's what drives a player to learn a new game. If I told you I had a great new game about rolling dice and counting, would you want to play it? What if I told you I had a game about using mad cows to discover unexploded bombs? Yeah, okay. They're the same game. I prefer to come up with the hook first, rather than searching for the perfect hook for an already-written game, because if that's your plan you can spend months working on a great mechanic and then find yourself with no way to sell it.

I also find that truly inventive game mechanics grow out of the story. In Kill Doctor Lucky, you believe you're the only person who wants to kill him (otherwise the game would be a lot easier). Because of this, you not only have to be alone in a room with Doctor Lucky, but you must also be in a place where no one can see you. This line-of-sight mechanic makes the board and the strategies more interesting, and it would never have come about without the story.

In the "keep it" department, I think the time spent playing the game should vastly outweigh the time spent leaning, setting up, explaining, and taking apart the game. Game rules should be simple and obvious; if it's an economic game, you should have money. If it's a fighting game, it should be fast. I try to be inventive but not to the extent that no one can understand what's going on. Falling is a good example of a game that was too different: lots of people never figured it out. Those who did think it's a blast.

I also, of course, try to build the game so that it's fair and balanced. That's a tall order, and it's easier to give the appearance of balance than to actually create it. For example, a game might be "fair" at the start, but badly out of whack by the end of the first turn. For example, in a purely random race where every player rolls fifty 6-sided dice, a player who rolls a 6 on turn 1 will beat a player who rolls a 1 on the first turn almost 65% of the time. If the gap is higher a few turns in, the lead is almost insurmountable. That means that players who got off to a slow start have no reasonable chance of catching up, so the game quickly becomes less fun. Designers compensate for this problem by installing "catchup" mechanics, which are often obvious and forced. For example, they might force the player in the lead to step back one space at the end of every round, and the player in the back to step forward. To me, that's embarrassing. The game ought to have more texture, and allow better strategy to make up for poor luck. More on that below.

Let's say you've got a game that's working well, but there is one clear problem that's still present. Either...

• the game is too dry.
• there is not enough player interaction.
• it's too long.
• or it's not making sense within the current framework of the theme.

Which one or two of these problems would you consider to be most crucial and how might you go about fixing it?

James Ernest: These problems have varying priority depending on the type of game and the audience. For example, if a game is designed for the European market (or American connoisseurs of those games) "too dry" is not really an issue. But for these players, "interaction" is critical. For the American youth market (specifically, selling games to publishers with distribution in that market) the opposite is true: the game must come with a cartoon show, a million dollar ad campaign, and a line of chewable vitamins, but it's okay if mechanically it amounts to two players drawing for high card.

I'll discuss solutions to each of these problems in turn.

The game is too dry...

"Dry" has lots of shades. A game can simply take too long, require to much math, or be about a subject that is not interesting to the players. Usually a game is too "dry" when the mechanics don't suit the theme, or when players spend more time optimizing their moves than taking them. This is in part a function of the group playing the game. However, there are certainly faults with any game that make it more dry than it might be.

I personally like pretty "wet" games: lots of flavor, luck, and fast action. I'm more entertained if I don't have to think too hard, and if there's something funny about the choices I make and the way the game proceeds. A good theme will go a long way towards lightening the load: in part, the theme gives people a reason to learn the rules, and in part (if the rules make sense) it makes them easier to remember. In the Kill Doctor Lucky example above, "I can see you" makes perfect sense and it's not hard to recall. It might be harder if the same mechanic was used to represent something else, such as orthogonally-firing laser-shutdown beams.

If a game is too dry I will usually try to fix the rules, because I have already decided that I like the theme. My guess is that something about the rules isn't engaging enough, or something about the theme isn't coming through For example: I don't take enough chances, get enough little victories, or have enough reason to choose one option over another. Things can also dry out if there seems to be only one clear path to victory. Once everyone decides that Orcs are the most powerful troops, everyone always buys Orcs.

There is not enough player interaction.

The concept of "player interaction" is one of my pet peeves. Not because games don't need it, but because game designers don't seem to know what it is.

Basically, player interaction is what makes multiplayer games different from several simultaneous games of solitaire. In a game with no interaction, such as the 50-die race described above, players are just moving forward as fast as they can and not affecting each other's position. Many complex games can feel just like this. For example, a MMORPG might have its users spending most of their time alone, killing monsters and leveling up. The game is only interactive when players make raids together, or when they buy and sell their characters on eBay.

Board game designers frequently use "player interaction" as a tool with one purpose: to give players in the rear a chance to pick on the players in the lead. One common solution goes something like this: Each player has one rock, and can throw it at any other player to slow him down. The net result: everyone but the leader throws his rock at the leader, and the leader throws his rock at the player in second place.

The worst version of this rule is one in which players must choose whether to help themselves or hurt the leader. In other words, each must decide whether to move forward, or throw a rock. In other games, the rocks are free, so everyone throws them all the time, and the leader always gets hit. In a two-player game, rock-throwing is not so bad: everything you do is targeted at one opponent anyway. But as a leveler for multiplayer games, it's a weak solution at best. At worst, it actually creates a game where no one wants to be in the lead.

What is player interaction, really? In multiplayer games, it's a game structure in which a player's every move can affect the position of other players. It's not a set of superfluous choices that compare to the rock-throwing rule above, but a set of primary choices that often, or always, affect the whole group. For example, in a stock trading game, since everyone might own shares of Company X, when I affect the value of those shares I understand that this will benefit (or hurt) all players who own part of that company. In Kill Doctor Lucky, if I stand where no one can see me, I also stand where I can't see anyone else. Thus, when I am alone, I make everyone else more alone as well.

Player interaction of this sort is not something that can be created with an extra rule or a few extra cards. It has to be woven into the system. (Now, before I get email, I have learned this principle through designing plenty of games with rock rules, so thanks, but I know.)

Another form of interaction is player communication, in the forms of concealment, bargaining, and bluffing. Bluffing is the main form of interaction in Poker. Players must behave in a way that conceals the quality of their hand, good or bad, but it is understood that this is their goal. Lying to liars who think you are lying is a tricky sort of interaction, to say the least. There is no mechanic in Poker that allows you to "pick on the leader" and in fact, in a tournament setting, that's the last player you want to tangle with.

The game is too long...

"Too long" is another of those terms that means different things to different people. I like extremely short games. You can play Brawl in 35 seconds. A hand of poker takes about 2 minutes. I will play poker for hours, but the game is always being reset, so a mistake I make now is effectively erased at the start of the next hand. Yes, I have less money, but that's not the point. Contrast this situation with any hour-long board game in which I can spend the entire game trying to make up for one foolish move.

I find games often go too long because the midgame stalls. A decent game of any length has three phases: a setup, a midgame, and an endgame. Even Brawl has this to some degree. In the setup, players are getting their resources into play, configuring their pieces with opening moves, or otherwise getting ready for battle. In the midgame, the game is in a more or less steady state in which all players are vying to advance. For example, every minute of Scrabble. In the endgame, the game is nearing conclusion, and players are trading in everything they have collected for whatever wins them the game, doing things that might seem foolish in the midgame but are perfectly appropriate at the end.

As a designer, it's important to give players as little midgame as possible. In my opinion, as soon as the game reaches a steady state, it's time for the endgame. Better yet, if the game never quite levels out, people won't be exactly sure when the endgame begins. So people can take chances on exactly when they start "cashing out."

Games can also bog down if individual turns take too long. This can happen when, to borrow a phrase from my friend Dave Howell, the solution space is medium-sized. If the solution space is small, as in Tic-Tac-Toe, a player can analyze all the possible moves and choose the best one quickly. In Chess, or any game with a suitably complex solution space (this can also be created by sufficient randomness), analyzing all possible moves is practically impossible. In these games, players make choices knowing they're not perfect, and they don't take forever to do it. But when the solution space is medium-sized, a player could trace out all the possible solutions if he really wanted to, and some players feel compelled to do this because otherwise they will not be playing their best game. Again, this will mean different things to different groups, but if you're finding that you are taking too long analyzing your strategy, you may be able to solve it by growing or shrinking the number of possible outcomes.

It's not making sense within the current framework of the theme...

When you start with a theme, and your mechanics don't match, it's usually correct to go back to the theme for a solution to your problem. I will typically change a game mechanic to fit a theme, not change a theme to fit a mechanic. A far more respectable game designer than myself, Reiner Kinizia, takes the opposite tack. "Theme is very important in my games," he says, "so I will frequently change the theme several times until I find the right one."

For me, the game is the theme, and the mechanics are just a way of telling the story. So I will redo mechanics from scratch until I find the ones that match the theme. In designing U.S. Patent Number One, I did a total of four completely different game mechanics before I settled on one, and frankly I'm still not happy with how it came out. But the story of that game is the same as when I began.

When the mechanics aren't meshing with the theme, I usually ask myself: what system am I trying to model, and what elements of that system make sense in the game? If it's a real system, it must work in real life (to some extent), so if it's not making sense in the game there's usually a disconnect between the game and real life. For example, in a game about building casinos, I was having trouble coming up with a way to help players who got off to a shaky start. I was rather embarrassed when I finally hit on the solution: gambling! If you're down on your luck, you can enter any player's casino and bet with what money you have. It's a little like a "rock," since the richest players are typically the targets, but unlike the rock it has slightly less than a 50% chance of working. After all, in the long run, the house always wins.

Sometimes efforts toward balancing a game lead to uninteresting, nearly equal choices. Do you have any advice on keeping things fair, yet still offering the players interesting decisions?

James Ernest: As far as I'm concerned, a nearly equivalent choice doesn't count as a choice. "Balance" is another one of those buzzwords that game designers throw around, like "interaction," without really knowing what it means. "Balanced" doesn't mean that no matter what road I take I always get to the same place.

For me, that choice is interesting if, for one player, choice A is clearly better, and for another player choice B is clearly better. It's not interesting if A and B are the same for all players. If Player A can have a strategy that he likes, for whatever reason, and Player B can have an equally effective strategy that leads him to make different choices, that's balance.

In CCG design, creators try to envision as many different combinations as possible, and make sure that none of the prevailing strategies is markedly better than the others. (It's okay if there are also plenty of strategies that are terrible, but no single strategy should be the best). This design parameter embodies the concept of "multiple ways to win" and, when executed correctly, is a brilliant exercise in game balance. No particular card is of equal utility to all strategies: if it were, it would be too strong. When building a deck, one player might desperately need Card A, while another player has no use for it at all. As long as every card has a place in some deck, and no single card is too strong, the game is fair and balanced.

Board game designers can take a lesson from this: no choice has to be equivalent for all players, as long as different strategies make all choices interesting to someone.

How do you decide upon and achieve a proper balance between randomness and pure strategy for a game?

James Ernest: This is the most subjective factor in game design. Some players like games with no luck, some players like games that are entirely luck. (Note that the amount of skill in a game, as Dave Howell would point out, is independent of the amount of luck. More about that below.)

Luck serves several purposes in games. It allows players who are not skilled in the game to have a chance of winning, and it gives them something to blame when they lose. A Chess player never says "the pieces were against me" or "my opponent lucked out," and for some players this is why Chess is a great game: players take full credit for their wins and losses. By contrast, few people claim to have "skill" at playing slot machines, yet they are a hugely profitable business, because sometimes people don't want to be responsible for whether they win or lose. Personally, when I am tired or drunk, and in the mood to gamble, I will seek out a game that requires no strategy at all!

One German game publisher wrote that luck in games represents the influence of death in real life. I'm not sure where he got that, but I take it to mean a game with luck is not perfectly predictable, and is therefore not worth over-thinking. Luck provides the "cognitive fog" required to grow the solution space I mentioned above, making it impossible for players to look too far ahead, and therefore limiting their decisions to a manageable few. In a game like Button Men, where dice are rolled every turn, the number of possible outcomes becomes overwhelming pretty quickly: enough so that one can play the game reasonably competitively by only looking one or two moves ahead. Analyzing all possible outcomes becomes geometrically more difficult as you look further ahead.

On luck versus skill: Many players like to rate games on a single axis between luck and skill, but this is a limiting metric. Luck and skill actually fluctuate independently. The linear metric is descriptive of "what percentage of the game's outcome is reliant on luck or skill" and in that context the two attributes can be seen as opposites. But games can have very little skill and absolutely no luck (like Tic-Tac-Toe) or a whole lot of both (like Poker). In a game with both, short-term fluctuations mean anyone can win, but in the very long run skilled players always come out ahead. I like games with this pattern because I like to give everyone a chance to win, while still rewarding good play over the long term.

The number or length of rules may be a factor when creating a game. How do you make the decisions of adding or removing rules during development?

James Ernest: A rule should carry its own weight. The hassle that a player must endure to memorize the rules must be justified by the benefit that it has to the game. For example, "You need the most points to win" is probably worth keeping. "On a turn number divisible by 2, if you have fewer points than the leader, you can call a vote among the players with fewer points than you to take one point away from the leader, but only if a majority of those voting can decide which other single player to give it to" is probably not. I don't care if this rule perfectly recapitulates the semi-annual General Council meeting in Burma, and if you feel that without these rituals Burma would be no different from any other place. No one is going to remember this rule, and even if they do, they're rarely going to invoke it. (This one also requires remembering which turn number you're on...)

It's always tempting to add more rules than you need. For example, when it's discovered that a player who collects more than three coconuts can win the game by standing on a certain space on the island, the designer might add a rule that says no player can own more than three coconuts. This isn't a rule you would expect from the context (what island nation prohibits people from owning more than three coconuts?) so players have to remember this rule in a relative vacuum, and make decisions based on the fact that while they can and should accumulate coconuts, they can never accumulate more than three. It's even more perplexing if the stand-on-that-one-space trick isn't obvious, since even once they know the rules, people are still mystified as to why they can't own more than three coconuts.

Rules like this are called "patch" rules, and they are annoying. The real solution to the coconut problem is obviously not the coconuts, it's the space, or the rules that let you stand there and win the game, or something else entirely. Chances are, the designer has got it in his head that the space and the rules are critical, and the max-three-coconut rule, irritating as it is, is the best solution. I think that's ridiculous, of course, and I'm always looking for a more invisible solution to a problem like this. The challenge is that it's always harder to change or remove existing rules, than to add just one more.

At what point in the game design process, if any, do you typically become bored or disinterested in a design and what methods do you use to help you push through to completion?

James Ernest: Actually, boredom is a sure sign that a game is going wrong. I won't "push a game through" if I am getting bored with it; I will analyze why it's getting boring. Obviously, playing the same game over and over can get tedious, but I have to hope that fans of the game are going to do the same. So the worst thing to do is ascribe the boredom to repeated play and assume it won't affect anyone else.

Now, if a game is boring me and I don't see an immediate solution, I will usually shelve it. There are plenty of half-baked games filling the bookshelves of this house, waiting for the day when a second look or a brilliant new idea makes them more fun. Deadwood sat around for a year like this, and in fact, when I found it again I had the cards and board but no copy of the rules. When we made up new rules to play with the same pieces, it worked much better!

If they haven't been addressed already, what types of design issues did you find yourself having the most difficulty with when you first started designing games? How did you overcome these design difficulties and what did you learn from them?

James Ernest: Designing games is pretty complicated, and I'm seriously considering writing a book. (I feel like I have today). I think the right approach is to design with the player in mind, and remember that no one cares about your game nearly as much as you do, so they'd better have a really good reason to want to play it. A lot of designers kid themselves and think that other people really need to see their new version of Chess...

We're seeing a lot of new games these days, but not many are hailed as innovative. How new or different from other available games do you think a design should be to be considered worthy?

James Ernest: Innovation is a double-edged sword. You can be so imaginative that no one understands you. Nobody bought Magic: the Gathering at first, and for awhile before it shipped, it was indistinguishable from a flop. In fact, most innovative new ideas are just that: flops. Once Trading Card Games became a known quantity, plenty of other companies tried the same thing. They usually made a little money, or lost a little money, but they could never do as well in the same market, because Magic was the first. So, if your creative idea is a success, it's got the potential to be huge. When you copy someone else, you have a more guaranteed result: you won't have a giant success, but you're less likely to have a flop.

I don't know what makes a game worthy to be hailed. But I do know that there are periods in any industry when the manufacturers are not as risk-oriented, and are therefore less likely to come up with anything truly innovative. Puerto Rico was hailed as innovative, but from where I sit it's just another Euro-style game with most of the same ingredients, and this is why it sells well, but not better than Settlers of Catan. Settlers wasn't the first game to work like that either, but it was the first to take hold in the American market.

At the moment it's really hard to make money in gaming, so most established companies are hunkering down and sticking to what they know. It will take another upstart like Wizards of the Coast (as they used to be) to get lucky with the next big thing. Big things come from little publishers, because big publishers don't like risks. Little publishers take risks because they have to, and that's why most of them fail. But that's also why, once in a while, you hear about the one that succeeds.

It's often hard for new designers to find good playtesters. Do you recruit your best playtesters or in some way train them? Whichever the case may be, how do you go about it?

James Ernest: I think anybody can be a good playtester, in the right circumstances. When I test a game I am always looking for people's reasons for making all the choices they make, and their immediate reactions to the rules and the story. My group is happy to tear a game apart and make frequent re-starts until the core mechanics work. I'm open to comments about mechanics that make no sense or turns that take too long. And by playing open-handed (sometimes) we can collaborate on strategy that would be harder to see if we played normally. Basically, they don't need to know a lot about game design, as long as they know when they are having fun.

If you have any other comments for aspiring game designers, we'd love to hear them.

James Ernest: There's a sign on my wall that says "write the game you want to play." I think designers write games for other people, especially when they perceive their fans as somehow different from themselves. As a result, some pretty unplayable games come out under the banner of "being for someone else." For me, if I'm not having fun playing my own game, I can't really expect anyone else to like it. So I always have to decide at the end of a session whether this game is something I'm willing to play. If the game isn't for me, I need to change it.

Now, that said, it's also important to realize that other people don't like exactly what you like. So be aware that while you're writing the game you want to play, and making sure it's fun for you, you should have some indication that it will also be fun for someone else.

I like a lot of different kinds of games. Not all kinds, but plenty of them. I like the formats I've pioneered, and I like the other formats I've worked in, like board games, card games, CCGs, and dice games. I won't usually do a historical wargame or an RPG just because I don't play those games and I don't feel like I bring much to the table. But once in a while I get the chance to do a miniatures game, and it turns out pretty well. :)