This is great. I've been a game dev for about 30 years, much of which I've spent working with narrative design/writing teams. One thing I've learned to watch out for, especially among junior designers, is what the author labels the "Time Cave."
Narrative branching, done well, is fantastic—it gives the player agency and lets them make the story their own (as it were). But when you're creating the story graph, it's easy to get lost in it and lavish care on one path at the exclusion of the others. You can easily end up with one or two long, greatly-detailed paths, and (because dev time is finine, and you need to move on to writing other parts of the game) a pile of other paths that are shorter and less interesting. If the player takes one of the shorter ones, they end up missing out on all your coolest stuff. The tools I would design for the kinds of games I created specifically made it easy to create a main story trunk with side paths (that rejoined the trunk), and more difficult to branch/loop/etc.
Of course, that's not the only (or even the best) way to do narrative design—Disco Elysium is a masterwork because it did the branching, merching, loops, jumps, random checks, and so forth, so well!
I think it's a mistake to try get a story-focused game to have branching paths, akin to the old choose-your-own-adventure books. Until LLMs can proactively create new stories for the player to enjoy dynamically, i think it's always fraught with peril that the player fails to get the full story (or have to repeatedly play it and choose something else to try).
My personal preference would be to have a single, on rails story, where the player don't truly have a choice. It's an interactive movie.
Or, pick a sandbox mechanic, and let the player do what they want directly, and compute the consequence (the most common type being the physics system).
I'm the opposite, apparently. I loved CYOA books as a kid because they could be reread, so I ended up seeking games that boasted multiple endings, including "bad" endings. When playing more linear games, I appreciated them for what they were, but there was a disappointment that I could not try different options along the way.
I think both have a clear place in gaming, since different gamers obviously look for different things.
> My personal preference would be to have a single, on rails story, where the player don't truly have a choice. It's an interactive movie.
I can't bother to play those kind of games. A movie will be able to deliver its stories a lot better than a game.
But with choice and branching you get to appropriate the protagonist(s) and some story events can be a lot more impactful then. Lately I played Cyberpunk for which you have some choices in most missions and the endings hit different. If anyone involved in the DLC story is around: kudos to everyone involved in making the "face in the crowd" ending. You play some almost super heroic character and due to your choices (which involve betraying and killing a lot of people) you get to survive: alone and back to generic human power level.
> A movie will be able to deliver its stories a lot better than a game.
I don't agree. Something like SOMA would just be a generic sci-fi B-movie but it's an awesome game, even though there's no real choice and is in essence just a walking simulator.
If you're not going to have choices matter, why make the story in an interactive medium at all? Branching paths require a lot of compromises, but there are still things you can do much better with handwritten stories than in a sandbox style.
> why make the story in an interactive medium at all?
have you not seen the success of the COD Modern Warfare franchise? Their single player game is essentially an on-rails shooter, with pivotal story points completely scripted (you "press the buttons"). There's no choice, there's no branching (of the story).
But people like to shoot, like to run around, etc. It feels like they have control, and it feels like the heroics in the story is their contribution.
Halo is the same. It is essentially a very long hallway with enemies to take care of before you can move to the next hallway. Also, I recently played through the first Halo again and it was still quite fun.
There a lot of great games with branching narratives, for example The Outer Wilds, Baldur's Gate 3, The Witcher 3, all the FromSoft games have extremely powerful narratives with multiple branches and multiple possible endings based on player choice. Done well there's a lot of replayability that arises from the desire to experience different paths within the overall general experience, as well as the immersion that comes from the player truly feeling they have agency in the world in the way that an on-rails experience just cannot do. For example there was one quest in the Witcher 3[1] where I remember being haunted by the consequences of an extremely difficult choice I had to make and wondering for days whether or not I had done the right thing. I eventually replayed the game and when I got to that point .... decided to make the same choice. sigh. So I still don't know whether I did the right thing. I just couldn't bring myself to try the other branch - it just wasn't right for me. One of the most powerful things a game has ever done to me, and very true to the complex and ambivalent moral tone of the books in my opinion.
Done poorly, you end up with something like Fallout 4. Theoretically there are choices but they don't matter that much and many of the branches get little to no QA love so have bugs and problems (eg when I did my playthrough of fallout 4 I unwittingly managed to avoid the main choice you're supposed to be forced into doing which is to choose which faction to align yourself with) which had numerous buggy consequences because the devs clearly had never expected anyone to play in this particular way (I basically did a breadth-first traversal of all the different quests but put off a couple of key decisions because I couldn't make up my mind. In the end this pushed me past the point where I was supposed to make the choice and I ended up friends with just about everyone. Except the institute. Screw those guys). Yet another reason that game was so disappointing.
[1] The whispering hillock. If you know, you know.
While I enjoyed playing Fallout 4, I was really annoyed with how it forced you into a specific path that lead inexorably towards a very specific ending no matter, as you say, what you do.
At the end, you have just two options: You can join the institute, or make war with them. I actually really didn't care about this faction and wanted to get out of that storyline, but that's what the game offers you.
Of course many games lead you to a specific denouement, which is typically one where you've defeated all the bad guys and win. But those games don't make any pretenses about making choices, whereas Fallout does.
I was really impressed with how many choices you had in Baldur's Gate 3. They definitely recorded an insane amount of dialogue just to cover all the possibilities. But I did find myself annoyed with the ending here, too, which only allows two choices at the end.
If there are two choices to make, it's basically the same as one, because the story just stops there. I'd rather games stop pretending. Don't pretend that there are so many gray areas. Just let the good guys win over the bad guys.
An alternative approach for such a game would be to have the decision tree be oriented around a kind of moral compass throughout, and your deeds decide what you can do. So if you keep killing or betraying people, you become more evil, maybe physically deteriorating into a kind of ghoul, while normal people start to fear you and refuse to barter or make allegiances. But as you get more evil, you gain access to eldritch abilities, make friends with monsters, and so on. There may be a point where you can make amends and return from evil, or vice versa, but at some point there's a point of no return where the final trajectory has been decided.
There are some games that have some something like this, where you have a "reputation" among the good guys that you can lose by doing misdeeds. Not sure if exactly what I describe has been done, however.
That alternative approach is basically what happens in Dishonoured and Dishonoured II where there is a “chaos” meter which determines how much the world descends into disorder based on your choices to kill people or let them live. There are multiple “clean hands” options and there are always ways to complete any particular mission or level with clean hands. There are some elegant ways the world’s chaos state are revealed to the player. For example in Dishonoured, the child Emily’s drawings[1] are more or less happy reflecting the chaos of the world.
yep, and i argue that most game devs wouldn't do it well enough to justify it. The limited time and resources available means they're almost certainly better off not branching, but make one good main story and polish it.
Take a look at the examples you listed - they're all award winning games, from developers with serious experience, grit and determination to make the best game. They spent ages, and lots of resources to do it.
Even big studios, with similar or more resources, fail at making such branching good. It's a folly to think that a smaller game developer with more limited resources could make it better. Disco Elysium is almost an exception that proves the rule (or another example is Pathologic).
imo its not necessary to get the "whole" story, or replay it to see every possibility
just seing "my" story is enough
interactive movie type games are great but they're a different experience from choice style games which are also great, (and ofc the article shows that there's many styles within this too, all with a different experience)
I don't believe LLMs can recreate the same authored experience that has a point of view. I think they'll be okay at genre work soon enough though, for better or worse. But thats not a type of game I'm personally interested in.
I completely disagree, there are plenty of branching games which are extremely good and which would be severely worse if they weren't branching: Disco Elysium and Baldur's Gate 3 come to mind.
We are very very far off from an AI being able to come up with compelling stories that are logically coherent.
> there are plenty of branching games which are extremely good
i think you mean there's barely any good ones. The examples you come up with are the exceptions that prove the rule. Look at a game like Dishonered, where the story have _some_ branches, but it's half-assed imho. There's plenty more games where having gone for a branching story made the game more expensive, less deep, and harder to sell as a result.
I think a good idea is to have a tree-like checkpoint-save system so you can always go back to the state you were in before descending down into a branch and go down another one without replaying everything up to this point. It encourages replays and exploring all content.
Tying together Steam's Web API + Claude 3 Opus/Haiku + ElevenLabs + some Unity magic to create 5-minute YouTube shows to cover hits and surface "hidden gem" games that might not get much video love elsewhere:
(Author) Good point! I picked the top categories by mentions. 24 entries that mentioned "code" or "copilot" (out of all total disclosures), and a third of them actually went out of their way to state that there was NO AI code gen; typically like so:
> it is not used in the game itself in any area: 3D models, code...
I suspect that a more rigorous perusal of the metadata (i.e., more than those quick search terms) would turn up some more, but either way, it seemed like such a tiny fraction of the whole.
Not looking to apply, but having tried this out some years ago, I really love the approach. Using indie games to get kids to handle their emotions is great.
I'm sure he did more than just one game, but the revenue wasn't enough and he had a family to feed. We connected on GarageGames back in the day, which saw some reasonable success. I think it's easier these days to make a living as an indie game developer, but it's still very much a hit-driven business.
> It's impossible for an elephant to be completely legless, as they rely on their legs for basic functions like walking and supporting their weight. Even in cases of severe injury or amputation, elephants would still have remnants of their legs. Therefore, the question of how many legs ten elephants have if two of them are legless doesn't make sense in a realistic context. Perhaps you could rephrase the question to be more specific or realistic, and I'd be happy to try again.
ChatGPT-4:
> Eight elephants with four legs each would have 32 legs in total.
Right now this just plays a canned video on loop—the only thing standing in the way of it auto-generating a new show each day/hour is cost of ElevenLabs TTS; I'd go over quota pretty quickly. I imagine the cost will come down.
Narrative branching, done well, is fantastic—it gives the player agency and lets them make the story their own (as it were). But when you're creating the story graph, it's easy to get lost in it and lavish care on one path at the exclusion of the others. You can easily end up with one or two long, greatly-detailed paths, and (because dev time is finine, and you need to move on to writing other parts of the game) a pile of other paths that are shorter and less interesting. If the player takes one of the shorter ones, they end up missing out on all your coolest stuff. The tools I would design for the kinds of games I created specifically made it easy to create a main story trunk with side paths (that rejoined the trunk), and more difficult to branch/loop/etc.
Of course, that's not the only (or even the best) way to do narrative design—Disco Elysium is a masterwork because it did the branching, merching, loops, jumps, random checks, and so forth, so well!
reply