This post about third acts continues a discussion of dramatic structure in games. The introductory post in this series is called I, II, III, the post about first acts is First Act, Ask Questions Later, and the post about second acts is Too.

There’s a thing people say about how a third act should be: unexpected but inevitable. As much as that’s a “sound of one hand clapping” thing to say, it’s true. A story should end with a twist that catches you by surprise, but when you think about it later, it has to make sense to the point that you can’t imagine things having turned out any other way.

In RPGs, the GM has the usual bag of narrative tricks. He can reveal that the focus of the second act didn’t tell the whole story, or that the premises of the second act were flawed or based on filthy lies, or engage in all kinds of similar what-have-you. The third-act reversal is tricky but no less do-able than for a novel or script.

What about the unexpected and competitive games, though?

In a board game, card game, or the like, if a player lays good pipe and plays well for the majority of the game, and everybody knows it, doesn’t that player deserve to win? There’s an argument for that. On the other hand, slavish devotion to outcomes based entirely on what’s gone before can create a wide swath of drudgery for the players who aren’t in contention to win.

Introducing (or at least allowing) the unexpected is a good way to apply enough emotional throttle to keep players seriously interested in the unfolding of the endgame, to stop them from thinking about what’s going to happen after the game is over, and instead focus them on what’s happening now. The fact is this: as soon as the end starts to seem foregone, “gone” is the mental attitude the players will adopt. (Or, perhaps or perhaps-not worse, they’ll conspire with one another and get all kingmaker.)

But in games that rely on player opposition over a relatively even field — that is, where the “narrative” elements arise from the actions of the players themselves — it’s difficult to script the unexpected.

It’s not impossible, though, to include elements that allow for the unexpected. Game-changing effects (cards that alter victory conditions, for example) and secret effects (cards or tiles kept in the hand, or face-down) are the relatively obvious ways to accomplish this.

The trick is that if major shifts become so common that it’s not possible for anyone to actually enact a strategy to win, the game becomes less like a game and more like a group of humans carrying out randomly generated computer code. Fluxx, for example. Andrew Looney is a fantastic human, but God, I hate that game.

That’s probably where we have to go back to “inevitable” as a key point.

An unsatisfying endgame based on “unexpected” might involve a seismic change stemming from one player revealing a card that changed the victory condition from the accumulation of victory points to the demolition of enemy bases, especially in a game where the two were not procedurally similar. But a satisfying variation might involve an opponent who slyly but right-out-there-in-the-open followed some unorthodox path to a relatively standard victory that no one saw until thwarting it (or abetting it?) became the defining thrust of the end-game.

None of These Acts are Quite Like the Others

Although it may go without saying by this point, an additional and critical feature of a third act is that it is not part of the second act. That is, the main concern of the third act is/should/must be distinct from that or those of the second act. There has to be a change of gears between them.

In traditional dramatic structure, the end of the second act answers the chief dramatic question and the third act asks and answers, “What then?”

It’s been proposed that in a serial dramatic structure (TV, or an RPG campaign), the third act can or should instead or also be either the first or second act in a new story’s dramatic structure. And I’ll buy that.

In a good board or card game, there should be a point at which the players begin to realize that the game is about to end. Whereas earlier on, the field of activity seemed open — even if one or more players dominated or seemed to — and although there may have been a feeling almost of perpetual motion, at some point, everyone realizes that one or more of the players has gotten themselves within striking distance of victory, and the focus of the game becomes that precise situation.

But if a key point of the third act is that it must push toward an ending, an additional critical factor is that it must arrive at an ending. There’s a limit to the number of times a thing can seem to be pushing toward an ending and then not arrive at one. Both instances of The Return of the King are heinous, heinous violators.

But as worthless a waste of English words as the Scouring of the Shire may be, my card game, Spammers, is worse: it can go nearly forever without ending, and it’s an object lesson in a particular game design problem, where a winner is required to maintain a winning position while all of the other players take their last licks, and said licks can easily be more than sufficient to drag the winner back out of his comfortable circle.

In the case of Spammers, the ease of dragging down the winner was a failure of playtest adjustment. That is, playtesting suggested more cards that allowed players to affect one another would improve the game, but too many were added.

But the thing that bothers people in Spammers is less the fact that a winner can have his parade ruined, but more that the game grossly violates their sense of when it will end. When one or more players start to close in on the victory point goal, everyone starts to think about what they’re going to to next — after the game is over, that is. When the game then violates those expectations and keeps the players stuck in their chairs, they get annoyed. Especially the players who see themselves as unlikely to win, or who only agreed to play because that’s what the rest of the group was doing. (That last happens more often then you think, I’ll wager. There’s a critical analysis of that player type in RPGs in Robin’s Laws of Good Game Mastering, which you should read immediately if you haven’t.)

Timing, Then

Whether or not dressing games up in a traditional narrative dramatic structure is putting a two-headed guy in a one-headed suit, I think there’s something to be said of a common expectation — even in games — about the respective temporal footprints of the early game, mid-game, and endgame being 25%-50%-25%. Vary any given component of the ratio by more than 10-15 percentage points, or invert the weight of the middle versus either of the others, and you start to break player expectations, even if they’re unconscious. That is to say, these will work:

10%-65%-25%; 35%-50%-15%; 30%-40%-30%

These, on the other hand, will provoke annoyance and confusion:

40%-35%-25%; 25%-25%-50%; 60%-30%-10%

I can definitely think of games I don’t like because they’re structured with weightings much more like the second set than the first, whether their designers thought about it or not.

More To Come…?

Although this is the last post I’ve mapped on the subject of dividing games into acts, it won’t be the last. I’m really intrigued by Will’s suggestion that there may be an entirely unprecedented structure to be derived for games, totally unlike the one that’s commonly accepted as the suit worn by drama. So, if you’re a forward-looking kind of person, look forward to that.