Note: This post will contain a tiny TNM spoiler. I’ll label it clearly and keep it after the jump so you can skip it if you want.
The thoughts herein spun off of a discussion I had with my friend Mads a few days ago, which resulted from this blog post of his: Why Writing and Programming are Fundamentally Different. I’ve added his blog to my links out there in the right side.
It made me think about what I’ve learned about branching dialogue during my work on TNM. As I’ve spent 7 years rewiring my brain to a more nonlinear form of creativity than other media call for, my sense of how to pull off good branching conversations has improved significantly, and it seems that the most important thing is to think less like a writer and more like a programmer – in that respect I agree with Mads’s thoughts.
But there’s something else – a trick to preserving the tightness of the dialogue. The more linear a conversation is, the easier it is to make the dialogue snap and spark the way a good screenwriter can. Writing, as anybody who’s taken a writing course will know, is not about producing realistic conversation, but about producing interesting conversation.
Real conversation is often idle, aimless, disconnected, and full of pauses and misunderstandings. In a screenplay, unless you’re deliberately aiming for realism over traditional mainstream writing, everything needs to be in there for a reason. If you include a pause, it must be pregnant with meaning. Misunderstandings must lead to something, they must affect the direction of the dialogue rather than just being a brief diversion before the conversation gets back on track.
Perhaps surprisingly, it’s easier to write realistic branching dialogue than interesting branching dialogue. Dialogue put together by a computer or by a player’s half-informed, tentative clicking more easily comes to resemble the aimless, slightly off-center, disconnected flow of an everyday conversation than the tightly paced and deliberate dialogue of a drama. To stay in control of the conversation and make sure it will eventually lead somewhere, a game writer has to drastically limit the nonlinearity of his dialogue.
Two good examples of different ways to limit dialogue branching without making it completely linear are Deus Ex and Neverwinter Nights. The former has largely linear conversations that only branch at important junctions, whereas the latter never gives the player less than 2 choices per line, but different choices often lead to the same response. The problem with DX’s approach is that the player feels far less in control of the conversation and the character of the avatar. The problem with NWN’s approach is that players often feel cheated upon replaying the game and realizing that by far most of their choices make no difference. Deus Ex’s choices feel far more important because when your choice wouldn’t matter, you’re not given one.
Whichever method you choose, however, the challenge you face as a writer is to make sure every branch of dialogue is equally interesting. Ideally this should go for all aspects of your game – if you give the player a choice, every option should have interesting consequences. This is why people complain about Bioshock’s wrench being overpowered – these are people who feel compelled to pick the optimal route through the game, and if that means they have to stick to the really boring wrench combat, they do so, and they have a miserable time as a result.
Often, as a writer, I feel that I have several ideas for how a conversation could proceed at particular junctions. As a screenwriter or a book author or similar, part of your job is to choose the option that you think is most interesting or will take the dialogue in the direction you want it to go, not to mention the direction that best fits your characters. When writing branching dialogue, you have the option of leaving that choice to the player – you just have to make sure every choice leads to something interesting.
As an example, here’s one of my favourite examples of branching dialogue from TNM. This is not a big spoiler, and depending on how you play and the choices you make in the game, you may not even encounter this conversation on your first couple of times through the story, but if you want to stay completely spoiler free with regards to TNM, best skip this blockquote box. Our player character Trestkon’s lines are bold.
- That was a bad move Trestkon.
- Oh I’m sorry, how inconsiderate of me. I should’ve left you some painkillers for that new headache of yours.
- Heh, nice. So what did you hope to achieve by sparing my life?
- Option 1: I needed a challenge. Killing you would’ve been too easy.
- I’ll give you a challenge. When you least expect it.
- I think you’ll find it quite difficult to catch me off my guard, assassin.
- Good thing I’m up for a challenge too.
- Option 2: A clean conscience. And I thought you’d prefer being alive.
- Morals? That should prove a decent enough weakness to exploit the next time we meet.
- I have weaknesses enough, but I won’t let anyone exploit them.
- You just keep watching your back, “Agent”.
- Option 3: I proved a point. Gave you and Kylie something to think about for a while.
- I’ll mostly be thinking about our next fight. You upstaged me today, but the only result of your arrogance is that I get a rematch.
- I’m looking forward to it.
- Only because you don’t know what I’m planning.
Each of these three short branches is a perfectly fine piece of dialogue that I’d have been happy to write into a linear script. I think they work just fine on their own, with a good amount of friction between the characters – they’re all nice, snappy exchanges between two quick-witted people. If I were writing a linear script, I would have to choose one option based on Trestkon’s character – is he an arrogant, self-confident person? Is he a empathic saint who never takes a life if there’s any way to avoid it? Or is he a psychologically savvy person who manipulates his enemies?
In a film or a book, I would have to make that choice, but in a game, it’s best left to the player. As long as I have good ideas for how to take the dialogue in interesting directions either way, putting the choice in the player’s hands will make the player feel a stronger ownership of the avatar.
I feel that the trick to making this work is to make sure every option maintains the conflict. The above is a conversation between two enemies who quite recently had a fight which one of them lost. If one of the options completely defuses that conflict, leading instead to reconciliation and mutual understanding, the drama is drained out of the situation – not to mention the rest of the game is changed significantly (not that we’d hesitate to carry through with that, but in this particular case the game would be poorer for it).
The player’s choice above doesn’t really change the outcome of the conversation – in all 3 cases, the relationship between the two characters will be even worse than it was before the conversation. But the player has had a chance to pick his own character’s motivations, and I think that’s worth writing and recording the 8 extra lines.
A final thought: The conversation I chose as an example ends after each of the three branches, but the way I see it, there’s nothing wrong with “merging” the branches again after they’ve played out. If more information needed to be imparted during that conversation, each of the branches could easily jump to the same line and the conversation could then carry on from there – as long as the conversation changes noticeably for a few lines, I believe the player will feel that the choice was worthwhile and that it was properly recognized by the game.