Win conditions — the rules that tell you what counts as a ‘win’ in a board game — are usually thought of as something defined by the game designer. In this post, I argue that they aren’t. Things are a lot more complicated than that.

There’s a bit of debate among co-op designers about what win ratio they should be designing for. Do you want your players to win 10% of the time? 50%? 80%?

And make no mistake about it, some co-ops are mean. Ghost Stories is famously brutal, even for the most experienced players, and each expansion added a new way to lose. Many patience card games are brutal, too, in their own way. Some 21% of the trillions of random possible Klondike setups are literally unwinnable, in the sense that there’s no move a player could make at any point that could prevent eventual defeat. Most of the rest are heavily stacked against you by the laws of probability, even if you played as well as you could with the information you had.

Competitive games can be deliberately unbalanced, too. Many wargames try to model a particular historical engagement, including the relative strengths of each side’s initial position, putting one side (say, the Gauls) at an up-front disadvantage against their opponents (the might of imperial Rome). In the same vein, Space Hulk – a 1980s two-player match-up between lumbering armoured humans and lightning-fast six-limbed aliens – is deliberately skewed in the aliens’ favour, although the effect is somewhat ameliorated by the designer’s recommendation that you simply play each scenario twice, switching sides halfway through, and tot up your win rate. (Absent any other tiebreak criteria, of course, this makes a draw one of the most common results – one of many similarities between Space Hulk and test-match cricket.)

Like Klondike, some computer games are literally unwinnable. Space Invaders and Tetris, at least in their original arcade iterations, are designed to keep kicking you harder and faster until, basically, you die. Modern tower defence style games have ‘infinite’ modes which work the same way.

But if the aim of every game is to win, why would anyone play a game where that aim was literally unachievable? And, for that matter, why would anyone accept an obviously disadvantageous position as a starting-point? Do these games sacrifice enjoyment at the altar of didactic value? Are they simply designed to teach the defeated player a lesson in history about how miserable it was to be a Gaul under Roman occupation, or a Space Marine dispatched to plant explosives on an alien-infested spacecraft?

In fact, of course, there’s a perfectly coherent reason why people play unbalanced games: to take on a difficult challenge and do as well as they can. As long as you know that the setup is stacked against you, it can be thoroughly enjoyable to take it on anyway. A Gaulish victory over the Romans, when you pull it off, is even sweeter because of the inherent disadvantage. And if you don’t win, well, that’s all right really because the odds were against you.

This explains why people are willing to play even single-player games that are guaranteed to be unwinnable, like Space Invaders and Tetris. The only question with these games is how quickly you lose. But that’s ok, because the challenge is to lose well: last as long as you can, get as far as possible, rack up the most points or the most kills or whatever. The aim is not to beat the game (which is impossible) but to beat the odds, to better your personal best, to go down in an even bigger blaze of glory than last time.

Of course, this way of thinking about “winning” can apply to straightforward competitive games like chess too. A win is a win in chess, according to the rules. But if you’re playing against an opponent who’s better than you, your ambition changes; the game is unbalanced in your opponent’s favour, just like Gauls vs Romans. A win would be great, but it’s unlikely; you’ll still be more pleased with a narrow defeat than a drubbing.

But this is where the rubber hits the road. We’re used to thinking of win conditions as defined by the game designer: to win, you need more VPs than your opponent, or more buildings, or a better village, or a madder castle, or whatever. And in the paradigm case, that’s fine. But the implication of what I’ve just said is that what counts as winning, in the sense of what outcome a player is aiming for, isn’t entirely down to the designer — it’s also partly down to the player.

In one way, this is obvious. My desires and objectives are psychological facts about me, not elements of game rules. Game designers get to determine how the game works, but they can’t determine what I think about how the game works, or what I want the outcome to be. Winning and losing — which are value states in the minds of players, remember — are as much up to the player as the designer.

Imagine a wargame that has rules identical in every way to, say, Warhammer 40,000, except for one difference: the rules say that if I completely trounce my opponent and destroy every enemy unit, then my opponent wins and I lose. Or imagine a version of the game in which the designer specifies that every possible outcome is always a draw. What difference would these changes make to the game? Apart from making the designer look silly, not a lot. No amount of exhortation in a rulebook can make me aspire to something that just doesn’t interest me. A rulebook can say “game end condition X counts as a win for player A”, but it can’t make player B want that.

This simple point has some interesting philosophical implications. Foremost among them is the realisation that the question of player objectives lies outside the formal system of the game.

  • What if I set out to lose a game rather than to win it? Does that “break the game”? Yes, it does, in a strange kind of way. We’ve all played against people from time to time who just don’t care about winning, and that attitude absolutely breaks the game. But these players are still following all the rules. And there’s no rule that the designer could add to the rulebook which would correct for it.
  • More benignly, players can deliberately redefine their win conditions: “I don’t want to win, I just want to finish the game having done X and Y”. This can be useful for kids: they can set the level of challenge that suits them, rather than being set a challenge by the designer which they can’t complete. It can be useful for adults too, either when you’re learning the game or when you realise halfway through that victory is unlikely. But redefining your objectives can also break the game, if the smooth running of the game relies on everyone to play like they want to ‘win’.

Of course, it’s not quite right to claim that game designers are completely powerless to motivate players to win. They have one important tool in their arsenal: the ability to make games annoying or boring to lose, and satisfying to win.

The most obvious way to do this is through moderating the difficulty level. It’s very easy to lose at chess, and very hard to win. That discrepancy is what makes winning satisfying. If I released a new, asymmetrical version of chess where everything was the same as the traditional game except that one player was trying to force checkmate but the other player was trying to get themselves into checkmate, it would be a failure: chess is designed in such a way that deliberately losing is unsatisfying, and winning against an opponent who isn’t trying to win is equally unsatisfying. (Of course there are games where you have to lose all your money to ‘win’: Go for Broke is an old example, and Last Will is a newer one. Fine. But these games are designed so that it is a challenge to lose all your money. Playing that win condition in Monopoly would be boring.)

So the game designer is trying to design a situation where each player is given a goal in the game, and the reward of reaching that goal comes from the challenge of getting there. It’s not doing well per se that’s motivating — it’s overcoming challenges to do well.

The moral of the story is that a game designer can suggest an objective, but it’s up to players to buy into it. They will if the game is well-designed and makes it satisfying to struggle for that objective. But they won’t if that doesn’t work for them.

To be continued

For most board games, the conclusion I’ve just presented is pretty academic. Most games are well enough designed that ‘what players want to do’ and ‘what the designers define as winning’ are the same thing.

But there is a particular kind of game where the stuff I’ve just been discussing really matters — where the attempt to specify win conditions separately from what players naturally want creates some interesting dilemmas. In my next post, due in a couple of weeks, I’m going to dig into how those games work, what makes them problematic, and how designers might try to solve the problems they throw up.