Over on the Escapist, Extra Punctuation has an awesome article about a game mechanic of leveling down rather than up. I’ve occasionally kicked around a similar idea, though starting more from “how do we do choice narratives where the choice feels significant?” — and one way to make the player actually care about choices is to tie the results into gameplay.
I’m attracted by the idea of a plot reminiscent of (the movie version of) “Last of the Mohicans,” where our protagonists start out as wealthy, happy, proper young ladies of English extraction and end up as bedraggled, hardened, and — in one case — dead. It’s a story of stripping away all peripherals until each character’s deepest feelings and commitments are revealed. That kind of story could make a compelling tragic game, or a story of triumph at excruciating cost, not far off from the structure of Victor Gijsbers’ Fate.
Leveling down, or gradually giving up your collection of Batman gadgets, or losing one after another of your crew of sidekicks until you stand alone, or burning away one after another of your huge inventory of doodads — that’d be a way to do the gradual-loss plot. Like I said, awesome.
Emily Short's Interactive Storytelling 
I agree that Croshaw’s pitch is intriguing, but I can’t agree with his rationale:
“I was repeating a statement I’d made before concerning RPG elements, stating that I’d played a lot of games with RPG elements where the game gets easier and easier towards the end as you gain more and stronger powers, which is failed design because games are supposed to have escalating difficulty curves.”
As I’ve argued before, most traditional CRPGs have a “long and winding road” structure. That is, the character starts off as an underdog, usually the archetypal “unlikely hero,” and 99% of the game is building that character into a powerhouse capable of taking on the primary antagonist. The feeling of power that a player gets from that process is part of the dramatic arc of that structure. That doesn’t mean that a CRPG should progress to the point where the hero can breeze through any challenge, but it probably would undermine that appeal to insist on an escalating difficulty curve.
I better solution (at least with reference to long-and-winding-road structured CRPGs) might be the “99 lives” solution. If any of you remember the trick from the original Super Mario Bros. where you corner a Koopa Trooper shell on a descending stairwell, bouncing on it to amp point values until it starts giving you 1UPs, maybe you get the reference. Basically, it was an easy way to max out on extra lives, but if you surpassed 99 lives, Mario immediately died and the game started over from the beginning.
To fit that into a narrative context, a CRPG could enforce something like the middle of the difficulty curve by letting power corrupt. If you built your character up much beyond the minimum level needed to beat the antagonist, the game could start to change the alignment of your character, take certain decisions out of your hands, and eventually convert the player character into the antagonist’s ally. When that happens, the player is given the option of demoting the character back down to reasonable levels, or starting out with a new, unleveled, character — who then has to fight not only the antagonist, but also the corrupted hero.
The idea of leveling backwards probably makes more sense in the context of IF, where thematic development can be as important as numerical value. I could see the principle being applied to something like a mix between The Strange Case of Benjamin Button and After Life, where the the process of stripping down is necessarily painful for the character, but also allows them to reach a state of purity qua character. The difficult decision-making process could, in that case, help build a strong identification between the player and the character, since in the course of stripping away the more fuller character they start with, they’ve ultimately shaped the character that emerges at the other end of the game.
Geneforge 4 (maybe other parts too, but I didn’t get to play them) has very similar mechanics. Through the game, you can find “canisters” which give you powerfull skills – free of charge, except that if you take too many, you start loosing conversation choices as you loose control of yourself.
one way to make the player actually care about choices is to tie the results into gameplay.
I’m a little skeptical about this whole idea. When the results of my choice are tied into gameplay, then I tend to care about the gameplay effects of the choices rather than the story effects. If you remember my story about my nethack cat, part of the reason I was so solicitous of the cat was that I was at a point of the game where I could get along without a pet (or could afford to run back up several levels to get the item that would resurrect it). When I was weaker and getting badly beaten up, I let my dog give its life to soften up an enemy at a chokepoint, though I did feel bad about it.
This was one of the problems I had with ‘Mid the Sagebrush and the Cactus — Victor described it as a narrative at war with an RPG, and I found that when I tried to act in a way that made some narrative sense to me, I got absolutely destroyed on easy mode. To make my way through the game I would have to make my choices according to the numerical systems, not according to their narrative valence.
Though I have to say that the stripping-down mechanic does seem promising for narrative — maybe because the loss of power could also tie into a loss of things you care about, so the narrative and gameplay choices could reinforce each other rather than working at cross-purposes.
I’m very much with Matt W here.
If you make my choices game mechanically relevant, I will pick the mechanically optimal choice, and I’ll probably reload until I get it.
Watching your resources dwindle into nothingness the further you get? Sounds like entering the survival drama.
I approve.
One game that did do that was the pen-and-paper RPG Call of Cthulhu, by Sandy Peterson (who went on to work in the video game industry working on some little-known titles like Doom, Quake, and Age of Empires).
Your characters pretty much inevitably slid into eventual insanity from everything they witnessed – assuming they survived (and in the early versions of the game, that was not very likely). So while your other abilities would increase over time, making you more and more heroic, each encounter with the unknown left you weaker, and more likely to become permanently useless as a hero… or worse, a threat to your own companions.
Just as another data point, the recent Mass Effect (PC RPG) games had a fairly typical weak-to-mighty progression, but toward the end you had to start making some very hard choices and sacrificing your carefully cultivated henchmen/friends to overcome the big bad guys. Dragon Age Origins (by the same developer, I believe) had a similar cycle. I had become quite attached to some of the characters, and I found myself replaying from earlier save games, trying to figure out some way to rescue them–not because I couldn’t beat the villan without them, but because I wanted them to live through it.
I’m reminded of ‘Enlightenment’, ‘Zero Sum Game’, and, in a way, ‘Spider and Web’. ‘The Sound of One Hand Clapping’ has a similar stripping-down-to-essentials basic premise as Enlightenment, but I don’t recall if you actually cycled through your inventory in quite the same way.
The difference there (from my point of view) is that the things in Enlightenment and Zero Sum Game that you’re giving up wouldn’t be useful to you if you kept them, and have only fairly limited emotional significance. So when I was playing I didn’t get the sense that I was losing capacity or making a hard choice.
There’s something like this in your Metamorphoses, too, isn’t there? Some of the puzzle solutions can involve sacrificing your belongings, and at least one of them has sentimental value — as it disappears you get the message “You’ll never see that again” or somesuch. (Which I believe actually makes it easier for you to carry out certain puzzle solutions later.)
Though again, I think emotional significance and gameplay significance don’t go together there — the items with the greatest emotional significance aren’t necessarily the ones that are the most powerful in gameplay terms.
Valkyrie Profile: Covenant of the Plume lets you “power up” by sacrificing your companions. (Or rather, they get powered up for a combat, and then they die immediately after, and you get their energy.) These aren’t robots either, they’re characters with motivations and dialogue and scenes you will miss if you do off with them. The effect is exactly that of “losing one after another of your crew of sidekicks until you stand alone”.
I can’t help but think that “leveling down” feels like a cheap gimmick to me.
Yahtzee’s argument seems to be that games should get harder as you go along, not easier. Now I know my gamer cock isn’t as big as Croshaw’s but I’ve never, ever, ever played an RPG in which the endgame was easier than the opening. As you get more powerful, you fight more powerful enemies, you can harp on all you like about how this is wrong or childish or unintuitive but it *works* and “the game should get harder not easier” is only a valid argument if your character gets more powerful and your enemies don’t.
A game where your character gets weaker over time might be interesting as an experiment, but I suspect that in practice it might turn the whole thing into a frustrating resource allocation exercise. What if you level down too quickly at the start of the game? What if you sacrifice the wrong NPC and find yourself unable to solve a future problem.
A game in which your character gradually deteriorates physically, psychologically or emotionally might be interesting. A game where this is reflected by reducing your gameplay options would be badly designed.
In the case of many games that people have mentioned, the effects of “leveling down” can be shown in various ways, not just through the story. Certainly game play mechanics can change a little or a lot. (Witness how Far Cry had your vision get blurry and your motion get staggered if you did too much in the heat due to the disease you had.)
Text adventures couldn’t convey this kind of thing except through the player trying typed commands and having static text come back. There would be a binary feel to this that I think wouldn’t work well in a text adventure format. Going with the Far Cry example, maybe you’d have a text adventure where the player character gets sick and so their total number of possessions that they could carry would drop (because they are weaker). That’s one level of this implementation and probably the simplest. The other is limiting the choices of the player based on actions they have taken. Knights of the Old Republic and Mass Effect did this well also. In some cases, the choices weren’t even presented to you. If you went the path of the Sith, then you couldn’t go the path of the Jedi. What was more interesting were the parts where you were still allowed to take one path or the other, but where you were still limited in how hard or easy it was to take one path, given other decisions you had made.
I’d be curious to hear actual thoughts about how text adventures would do this rather than just theory about the technique of “leveling down.” The player is the one responsible, for example, of thinking through puzzles. So how do, as one example, make the player character “dumber” (i.e., can’t solve puzzles) when it’s clearly the player who is doing so. It seems like you’d have to relate this to opportunities available to the player. But then that gets into how the player of a text adventure knows that an opportunity is now possible, particularly since the only way this can currently be conveyed is through text.
Construct your puzzles with massively multiple solutions. There are a lot of ways to do that, but it might be based on a simulation system as in Magnetic Scrolls games, a physical simulation/magic system as in Metamorphoses or Savoir-Faire, or a tactical combat system like the one in the ATTACK library. Start the player with a large supply of resources (nouns/objects) that can be fed into these puzzles, or else with a lot of action styles (verbs) that can be used to manipulate things. (For instance, in a magic system, that might mean that you start out being able to cast many different spells, but sacrifice access to some of them as you go along.)
Then (and this is admittedly the hard part of the design, but it was always going to be, even in a non-IF context) you have to make sure that all the late-game puzzles can be solved with any of the various resources/skills the player might have remaining at that point. Challenging, but doable.
“Construct your puzzles with massively multiple solutions.”
But is the audience for text adventures receptive to that kind of game play? Even in many of the games (non text adventure) out there, there aren’t “massively multiple” solutions. There are *some* solutions. I guess it depends what “massively multiple” means because then you get into a notion of how truly distinct the end states of many of those multiples truly are. In other words, are you just crafting a series of solutions so that you have a deeper gaming experience? Or are you just doing it to provide more — not deeper — gaming?
“For instance, in a magic system, that might mean that you start out being able to cast many different spells, but sacrifice access to some of them as you go along?”
Okay, but how does that matter to the game play, though? I get what you’re saying: you’d have to make sure that even if someone sacrificed their ability to use certain spells, the end game would still have to work. (Or, rather, one of the presumably many possible end games would have to work.) But why ultimately does that matter?
Originally this post started with “how do we do choice narratives where the choice feels significant?”
I’m not sure that just because I sacrificed a spell, I’ll feel that was significant — unless it’s very clear that I’m doing so at the time I’m doing it and, further, I have some notion of what I’m giving up and why it might matter. Someone earlier said about Mass Effect “I found myself replaying from earlier save games, trying to figure out some way to rescue them–not because I couldn’t beat the villan without them, but because I wanted them to live through it.”
Exactly right. You wanted the person to live because it made the game richer; you had come to like that character and wanted them to survive. It was the very notion of having to make the choice to sacrifice them — without knowing all the consequences, beyond how you felt about it — that mattered.
Maybe what text adventures should do is a little experiment. Take games that have been widely recognizd as having powerful or evocative storylines. Find the scenes in those games that are most commented upon and try to write those as a text adventure. See if it’s possible to convey the same sort of emotional appeal and the same sort of gaming experience.
If text adventures are capable of evoking the same emotions and experiences, then the mere fact of a differing mechanic should not make all the difference. I also think this will take it away from the focus of just “late-game puzzles.” Games beyond text adventures have been successful for many reasons but I think one of them is how they merge the puzzle solving in with all aspects of the environment and the story. Half-Life 2’s gravity gun is an oft-cited example and I’ll even grant it’s not the best of them. But the fact that gamers still routintely talk about it does count for something. It allowed for multiple solutions to “puzzles” but also allowed you to close off one solution while still leaving others open. It’s pretty impossible in Half-Life 2, short of dying, to completely prevent yourself from reaching the end. And Half-Life 2 is a very linear game. Yet, gamers widely hailed the game as a good game and a good story. The same applies to Dragon Age and there, of course, the game mechanics are entirely different.
Another great example is The Force Unleashed. Barring its unfortunate control bugs (a mechanic), the story is considered one of the best — if not *the* best — of Star Wars games to ever be made. Yet your role as Starkiller is really very linear. You start Sith and you sort of just become a Jedi, even if you don’t want to. But then you also get that binary decision at the end as to how you want it to go. And both endings were very emotionally satisfying (one, of course, being emotionally negative and the other being emotionally positive).
Can text adventures give that kind of experience that gamers respond to? If it can, I don’t think “leveling down” is going to be the solution.
But is the audience for text adventures receptive to that kind of game play?
Yes.
Next question?
Okay, I can unpack that a bit if you like.
What I mean by “massively multiple solution” is specifically the kind of solution that results from having a clearly defined simulation system — the text equivalent of the gravity gun in HL2, in fact — and then defining some criteria on which the puzzle is to be judged “solved”.
How the player gets from the starting state to the solved state is then up to him, and there are many ways of achieving it.
Games that implement this kind of content have generally been warmly received by the IF community because of their relative accessibility and the lack of read-the-author’s-mind puzzles, as well as their thematic coherency and the possibility of sneaking up sideways on puzzle solutions. Even if the player doesn’t have an immediate burst of insight about how to reach the solved state, he can play more with the system, increase his understanding of its capacities, and then deploy it better. A tight and focused simulation can give very clear feedback about why a puzzle solution is failing, which both hints towards the correct approach and removes some of the frustration of a player wondering why his solution doesn’t work. Moreover, focusing on a specific simulation type for a game conveys to the player the domain of probable solutions, which makes it easier to guess how to approach new puzzles.
There is absolutely no reason why IF games can’t do this, and there have been a number that do. The subset of things one can simulate is a bit different — it tends to be less about physically stacking crates (hard to express in words, boring to manipulate with a command line) and more about qualitative state changes.
In that context, it’s also not implausible to imagine a system in which solutions get harder and harder as you part with more and more of your manipulation tools/actions/whatever.
(Obviously, I also agree that this mechanic would need to be framed in a way that the player would have reason to care about.)
Maybe what text adventures should do is a little experiment. Take games that have been widely recognizd as having powerful or evocative storylines. Find the scenes in those games that are most commented upon and try to write those as a text adventure. See if it’s possible to convey the same sort of emotional appeal and the same sort of gaming experience.
That strikes me as a pointless experiment. Can you do the end of Citizen Kane as a statue? Can you convey the effect of Michelangelo’s Pieta as a passage of blank verse? Does the result tell you anything qualitative about the positive potential of those media?
Anyway, I think this conversation is rapidly falling into a communication gap that results from disagreements in our starting premises. You’re assuming, first, that the goal is to emulate both the content and the sales of a AAA game, though I’ve said that isn’t my goal; and, second, that IF == text adventure, a genre defined by a profoundly retro sensibility and a devotion to narrow puzzle types. IF is a medium, and it already encompasses a number of genres besides the text adventure, as well as followers and practitioners who are interested in further exploration.
“I was repeating a statement I’d made before concerning RPG elements, stating that I’d played a lot of games with RPG elements where the game gets easier and easier towards the end as you gain more and stronger powers, which is failed design because games are supposed to have escalating difficulty curves.”
This claim is baffling on several levels, and the only way I can make sense of it is by reading “a lot of games with RPG elements” as action-RPGs a la Symphony of the Night and its many descendants. But if that’s the case, the genre intrinsically fails by his metric.
Action RPGs as a rule do not have difficulty curves. A core aspect of their contract with the player is that the player can spend time (overleveling by grinding, or farming useful items) to trade off for skill (overcoming challenges without recourse to these things). When you flatten that out, most modern RPGs do in fact have an escalating difficulty curve. In both absolute numerical terms, and in plot terms, endgame enemies are more powerful than beginning game enemies.
The only place where this complaint could hold is in “persistent world” RPGs, where each area has enemies of a fixed power level, and the player can (theoretically) visit them in any order. (TVTropes refers to this as a “Beef Gate” – until you can deal with the beefy enemies in a region, you don’t get to go there, so it’s a softer way than fixed-plot to enforce an ordering.) This was the basic mechanism in games like the earliest Dragon Quest/Dragon Warrior games, or the Might and Magic games, as well as a few science fiction games like Starflight and Star Control II. It’s not immediately clear that any of these suffer for that.
Not to mention that, as stated, he’s saying that anything that isn’t an increasing difficulty curve is failed design; this is deeply weird because the only place one ever finds that in the first place are games explicitly posed as challenges to the player. (Tetris, Geometry Wars, Chip’s Challenge.) Once you also have “tell a story” in there, dramatic pacing is going to insist on some bumps and dips in the curve to control pacing, and in huge swathes of design space (Madden 2011, the Mega Man platformers, anything with a hub-based map architecture) the concept of “difficulty curve” can’t even be meaningfully posed.
For IF, even in puzzle-based challenge games, it still doesn’t work out too well; I don’t think the idea of a difficulty curve for Varicella is even logically consistent.
You don’t seem interested in hammering out the disagreements in starting premises that you bring up. But I’m confused on this:
“that IF == text adventure, a genre defined by a profoundly retro sensibility and a devotion to narrow puzzle types.”
Didn’t Infocom use the term Interactive Fiction? It’s not like this is a new term, right? Are you saying Infocom games weren’t text adventures? Was Infocom just a devotion to narrow puzzle types? Where was this cut-off when text adventures *became* interactive fiction? Or is this some sort of neo-Interactive Fiction movement? Maybe the reason people outside the community don’t see all these differences is because those inside of it aren’t very good at making it clear.
As far as the “pointless experiment”, if you’ve already determined that without even trying, then I agree there’s no point really having that sort of discussion or even trying that sort of experiment. Talking about rendering scenes between two different game types and talk about doing the “end of Citizen Kane as a statue” seems purposely obtuse to what I was describing.
I agree, though. Further discussion is probably not warranted. I’m finding the text adventure community, or at least representatives of it, are overly defensive. By all means, however, keep at it. I’m sure text adventures (or Interactive Fiction, if you prefer) will become relevant again one of these days.
I don’t know if I’d say that “text adventure” and “interactive fiction” mean different things, but for me “text adventure” certainly does have a connotation of a sort of old-school puzzler — the sort where mazes and inventory limits were potentially a big deal. And a lot of commentary about “text adventures” makes assumptions about the nature of interactive fiction that haven’t been true in a long time. For instance, this comment from a site that has a lot of intelligent discussion of games: “Text adventures are designed to be difficult and puzzling because, gameplay-wise, that’s all they’ve got.” That’s not true, it hasn’t been true for at least over a decade, and that very site reviews a lot of interactive fiction (or text adventures, if you prefer) for which it isn’t true.
So I think people involved in IF do find it wearying when people make assumptions about the medium that seem to be rooted in Zork-style text adventures. For instance, you talked about blurry vision in Far Cry — that seems like exactly the sort of thing that could be done in IF. Why not?
Leveling Down. What a concept. And yet, isn’t that the “Fall from Grace” in a nutshell? Imagine a royal forced to hide as a new recruit or peasant due to a revolution or family crisis. In that case however it’s less about the leveling down than being a fish out of water in a unique situation. i.e. Level 10 Noble becomes dual class w/ Level 0 Warrior/Peasant/Thief…
The trick with that type of setup for a group would be they would have to be very experience roleplayers and be willing to start from scratch. It’s the “Private Benjamin” or “Life Stinks” movie approach.
Definitely a thought provoking post. Thanks and keep up the great work!
Desktop Dungeons has elements of this. While you do increase in power as you level, healing and replenishing magic power occurs only via exploring new dungeon tiles, and the dungeon is VERY small. The game mostly plays as a puzzle game with a strong luck element that focuses on careful usage of this finite element. Near the beginning, killing all the enemies is paramount so that you can level, and you’ll often take on enemies of higher level where you can, in orde to accelerate levelling. Toward the end, that all reverses as remaining unexplored dungeon tiles become ever rarer and more precious.