,
59 tweets,
6 min read
Read on Twitter
Alright, I'd better get started :D It will of course be based on the particular kind of games I worked on (Warning: probable Frenglish)
1. Narrative design = creating an inhabitable fictional space through game mechanics and conventions. Story comes second, always.
2. A videogame is a world to inhabit for a few hours, without ellipses or editing. The simple experience of “being there” is everything.
3. Your job as game narrative designer is to define what “being there” means for each particular game, and how to engage with the world.
4. This means defining where/how to add body awareness for 1st person games, the types of interactions with the setting, how you talk, etc
5. There’s a running joke about flushable toilets in games :) Trivial things can make a fictional world feel real, anchor the player in it.
6. Imagine the trivial narrative things that can anchor the player. You don’t have to stick to existing solutions like notes and audiologs.
7. Narrative mechanics can be systemic (the Heart’s secrets in Dishonored) or non-systemic (the vignettes in Edith Finch). Both have value.
8. The best narrative mechanics fit into the gameplay flow (instead of interrupting it like when you open a fullscreen UI to read a note)...
9. ...But if done right, a note or audiolog can feel like a reward after making your way to it through danger or navigation puzzles
10.Ex of narrative mech. intertwined with gameplay: clockwork soldiers with recorded messages from your target as barks. I want more of this
11. Most of the writing in a game is background noise: overheard conversations, barks… But it’s what makes the world feel alive and real.
12. Have several different voices be heard. A world never has only 1 voice or 1 message. Use different “channels”, with diff. points of view
13. Ex in Dishonored 2: notes, books, diaries, overheard convos, player one-liners, cutscenes, player's travel log, 3 different newspapers
13b. barks for half a dozen factions and 3 social classes, posters in the streets, streetspeakers, audiographs, Heart's secrets, 2D videos
14. The game that you have in mind when writing and planning art is nothing like what the actual game will end up being. Get over it.
15. Know when to interpret and reinvent a request from the level designers or system designers, and when to stick to the exact demand
16. Corollary: know when to give creative freedom to the writers you work with and when you need something specific. Be clear.
17. In other words: often you don’t do what you want or like, because a game is the result of dozens of people’s ideas and constraints.
18. Find your own little thing within the big project and fight for it: a secondary character, a cheap little feature that will feel special
19. For your own sanity, find an enjoyable way to make what you don’t enjoy. Be creative with how you tackle it. Don’t be a killjoy.
20. Of course always make sure that the gameplay tells the same story as you do. “Ludo-narrative dissonance” is such a useful concept :)
21. Many colleagues are biased against narrative and will consider compromises should always be made by us. It’s not true.
22. A good game can be made truly memorable by the power of fiction. Portal is a great game, but would it be as successful without GLaDOS?
23. Storytelling is incredibly difficult to playtest. A narrative with placeholder texts and art will never work...
23b. You need to take feedback but also be able to project mentally how the game narrative will work out once everything’s in place.
24. Make sure the player *wants* to complete the mission you’re giving them as much as the protagonist does.
25. Don’t judge the player. If they want to go with a violent playthrough, don’t punish them for it. But show them the consequences.
26. Don’t "over-simulate". Plan for most consequences to players’ actions, but don’t bother with edge cases – unless they’re fun.
27. Avoid memes, real world jokes, pop culture references. It’s never worth the immersion breaking, it’s lazy and cheap. Own your fiction.
28. Don’t work in a bubble. Go speak with the other teams, see what they’re up to. It can give you new ideas or help spot misunderstandings.
29. Participate in the VO recording sessions. There are a million details that only you can know and it will spare you a lot of problems.
30. Know how your world works: society, history, fashion, religion...But don't make a lore dump out of it. It belongs in game documentation.
31. Try to be consistent when referring to characters, places, items and always call them in the same way, or players will be confused.
32. A hint should never be told only once. Players will forget, miss it, interrupt, skip it, break it. Repeat it in various places and forms
33. If possible, find a way to present critical information visually, in the world. Most players don’t pay attention to texts and VOs.
34. Even when you think you wrote something painfully short, it’s probably too long.
35. Writing barks: most of them should be super vague and generic - but not bland! If too precise, they will feel wrong half of the time.
36. ...But do work on some super precise ones, and make sure that they trigger correctly, so that the AI look clever and feel more human
36b. examples from Dis2: guards bark when the player is out of ammo, when you shoot their helmets off, when you're close in shadow form...
(all these neat bark ideas came from @FloppyBrown :))
37. Barks are gameplay feedback, they don’t need to feel super realistic. Players actually do like some goofiness (whisky and cigars anyone)
38. The final game story is a sum of solutions to problems brought up by level design, technical constraints, gameplay limitations, etc.
39. So, since as a narrative designer you work with all the different teams, the overall consistency of the game relies heavily on you
40. the good side of this is you often get to tailor your job to your affinities: work more with gameplay/AI, or more on environment/art...
Okay, I need to go to bed, I'll stop at 40 at least for now :D Thanks everyone who read and commented!
41. Always write overheard convos from start to end, so that they can be triggered from a distance: the player will naturally miss the start
41b. If you write them as if NPCs were already talking, like in a film, it will sound weird when triggered in game, if the player notices it
42. Put two ideas in each text/convo: one to give gameplay info, and another one for character building. Random NPCs deserve depth too!
42b. It will also make the player feel clever for guessing which part is relevant to their mission.
43. If possible put the gameplay info at start of text/convo, for players who don’t want to read or in case the scene gets interrupted.
44. This shouldn’t apply to “cinematic” scenes which should have their proper structure and pacing, while players just sit back and watch.
45. Find a way to offer a recap of what happened in the previous missions for players who pick up a game after some time
(We had the Travel Log in Dishonored 2, which also added more personal insights from the protagonist)
46. Texts are the cheapest way to show consequences and variations (like various newspaper articles acknowledging the player’s actions)
47. … so be sure not to deprive yourself of this option by adding to them expensive illustrations that can’t be as flexible.
48. Everyone tell you that the “briefing” cutscenes are too long, until after the playtests when everyone tell you they don’t explain enough
49. Force your colleagues to record placeholder VO, 3 people crammed in a tiny recording booth because it's hilarious
50. most of the time, use TTS to generate placeholder VOs. But be aware that the timing will never be right, so animations will need retakes
