Losing and wanting one more

There's a very simple test to know whether a game works, and it has nothing to do with winning.

You lose a match. What do you do?

If what comes out on its own is typing again, the game is well made. If what comes out is closing it and doing something else, something went wrong. And the interesting thing is that those two reactions almost never depend on how difficult the match was.

The oddity of paying to fail

The researcher Jesper Juul wrote an entire book about this contradiction. We spend our lives dodging failure — we avoid it at work, in relationships, anywhere we're being measured — and then, in our free time, we voluntarily seek out activities whose most frequent outcome is losing. It's strange. And yet it's the foundation of almost all interactive leisure.

Juul's explanation is that defeat in a game brings two things that defeat outside it almost never brings. The first is an understandable cause: you lost because of something, and that something can be named. The second is that you can try again right now. Together, they turn failure into information rather than a judgement on who you are.

Take either of the two away from defeat and it stops working. If you don't understand why you lost, there's no information. If trying again costs too much, the information is of no use.

Legible defeats and opaque defeats

Almost everything people call "bad difficulty" is really a lack of legibility.

A legible defeat is one you can reconstruct. You know you got overconfident opening the chest, that you crossed without looking, that you spent what you had left too soon. It hurts, but it leaves something behind: a specific sentence that begins with next time.

An opaque defeat leaves nothing. Someone appeared and eliminated you in a few seconds with something you never saw coming. Or you were behind before you'd even started, because the other player had better gear and there was no way of knowing. Or something random happened with no room at all to react. In those cases there's no lesson to draw; all that remains is the vague feeling of having wasted your time.

The curious thing — and it's what makes the word "difficulty" so confusing — is that a brutally hard game can hook you for hours if its deaths are legible, and a fairly gentle one can wear you out in twenty minutes if they aren't. Hardness isn't the problem. Opacity is.

Optimising for skill has a price

There's a design decision almost nobody states out loud but which defines the whole experience: what you optimise for.

When a competitive mode is optimised for the skill ceiling, something perfectly logical happens. The best players have an enormously deep space in which to improve for years, and that's a real achievement. But the typical experience of someone new becomes this: you join, you run into someone with thousands of hours and you vanish in ten seconds.

That defeat is one of the opaque ones, even if technically it's entirely fair. There's no decision of your own to review. There's only a gap. And repeated three or four times in a row, it stops reading as a challenge and starts reading as a sign that this place isn't for you.

The alternative isn't to dumb the game down. It's to optimise for something else: for memorable moments to happen, for there to be several ways a match can get interesting besides aiming better, for occasional chaos to hand out chances to those who haven't mastered the mechanics. An improbable comeback, an absurd fluke, a thirty-second alliance with a stranger. None of that takes depth away from those who seek it, and it gives everyone else a reason to type again.

What makes trying again cheap

There's a less elegant but equally decisive factor: how much it costs to play again.

A short match forgives a lot. If losing costs you three minutes, defeat is almost free and curiosity always wins. A long match hardens every death, because what you lose isn't the match but the afternoon.

And then there's the punishment that accumulates. When losing doesn't just end the match but also takes away something you carry afterwards — position, resources, a streak you'd been nurturing — defeat stops being an episode and becomes a loss. At that moment the game changes its nature without warning: you no longer play for the sake of playing, you play so as not to lose what you've built up. It's a very effective way to keep people in the short term and a very effective way to burn them out in the medium term.

The only metric that really matters

You can measure many things about a game mode: matches played, average time, who wins most. None of them says much on its own. Someone can play a huge number of matches and be having a so-so time, trapped in something they no longer enjoy.

The good signal is smaller and harder to fake. It's that word that appears in the chat after a defeat, typed almost without thinking: again.

It means the person has lost, has understood why, has been left with an idea of what to do differently and wants to test it. It means the defeat hasn't pushed them away: it has invited them in.

Anyone can design a game where winning feels good. Winning always feels good. The rare thing, and what sustains a community for years, is designing one where losing also leaves you wanting to stay.