Cave Story is a classic of the indie games movement. It single-handedly showed many people that a single developer could make a game with dated graphics that was as good as AAA commercial games. This was already clear to some, but Cave Story‘s prominence means that it has heavily inspired much of the work done by the modern indie games culture. There are a lot of things that Cave Story does well; its handling of mood and narrative structure are great, as well as its balancing of humor and pathos. One thing it does badly at, however, is providing the player with effective choice and agency.
There is a point midway through the game where the protagonist, Quote, is exiled to a labyrinth. While trying to find an exit, Quote enters a room with a large gap between him and the other side. Immediately, a character falls from the top of the screen to the bottom and lands with a thud. This is Professor Booster, a friendly face in a hostile environment. This scene encourages the player to do one thing: drop down and talk to the Professor to see if he’s okay. The gap looks too big to cross, and indeed is very difficult to surpass (especially if the player doesn’t have the jump-extending machine gun). The Professor isn’t moving, and Quote has just encountered a different friend in the Labyrinth which he needed to speak to to help.
However, speaking to the Professor is the wrong thing to do. Doing so causes him to die, and provides Quote with the Booster 0.8, a double-jump item. This also makes it impossible to get the best ending to the game. There is no announcement of this fact. Even if the player suspects something, and restores her save, it is not at all clear that she has any option but to speak to the Professor. However, if she makes the (very difficult) jump and abandons Quote’s ally to an unknown fate, he later emerges from a teleporter, quite alive, and gives Quote the Booster 2.0. The Booster 2.0 is a superior jump enhancer, and unlocks the best ending.
The sin committed here is the removal of the player’s agency over the game’s narrative. As Emily Short has pointed out, agency requires that the player be able to make a reasonable guess as to the result of her actions. There can certainly be unforeseen consequences, but they should make sense to the player in retrospect, and should serve a narrative purpose. In Cave Story, choosing to help the Professor kills him and dooms the player to the suboptimal endings. Choosing to ignore Quote’s friend saves the professor and unlocks the best ending. The game might as well flip a coin, or base its ending on whether the player pulled one of two unmarked levers.
A choice made without the most basic of information does not provide interactivity. I won’t say that this should never be done, but it should certainly be done for a reason. Any justification I can think of for this design decision in Cave Story — an existentialist comment, say, or a promotion of selfishness — doesn’t match with the rest of the game. If you are presenting the player with a choice, make it clear that there is a decision present, and give some clue as to the result. This doesn’t mean you need a big flashing dialog box, or that you can’t surprise the player later. But to have a major decision point be this counterintuitive and unexpected is just a jerk move.
How, then, do you reconcile a desire to give the player meaningful choice with a desire to surprise them? There are a few ways, but one is to imply the general results while withholding specific information. For example, tell a player that destroying the ship’s infrastructure could be dangerous, then give them the option of blowing open a door or going around through a dangerous air duct. Later, reveal that blowing open the door depressurized a habitation wing. In the case of Cave Story, make it clear that approaching the Professor will kill him (an unstable walkway?) or maybe let the player choose between taking the Booster 0.8 or leaving it to help the Professor escape. At least then it’s clear that you’re putting the Professor (and maybe more!) at risk by checking on him.
And if you’re going to break this rule, have a damn good justification for it.
I’m sure there’s discussion to be had on this topic, so comment if you have something to add!
I wasn’t sure I agreed with you when you mentioned this on twitter but reading your post in full I think I mostly do.
One problem I have with decision making in a lot of games is how clear the consequences are. Especially when you’re making a narrative decision I hate being told here’s the two options and here’s a full run down of the results of your choice. Because rather than the player thinking what they’d do in that situation and how they want to deal with the issue, they instead think about what’s the best for them to complete the game. Which gives them the best items or affects their karma gauge how they’d like.
My favourite example of what I prefer is the Tenpenny Tower dilemma in Fallout 3. The ghouls want to live in the tower, the humans don’t want them there. The game leads you to three choices: help the ghouls break in, kill the ghoul leader or negotiate to get them to live together. What I love is that you can’t force a happy ending to the situation. Even if you choose the middle ground option it still ends tragically. I like that the game gives you a choice but you don’t really know how it will pan out until much later. The game doesn’t really inform you what the results will be.
However I don’t think this is the same situation as the Cave Story example because the decision isn’t hidden or misleading as such. You know you’re making a choice you just can’t forsee how it plays out. I don’t think I like the Cave Story choice either since it seems like an attempt to trick the player into ruining their own play through.
I agree that the design of the Tenpenny Tower dilemma sounds better (although I’ve read criticisms of how it assigns Karma to that situation). The key there is that the player can imagine the possible consequences. What will the ghouls do to the residents? Will they be able to live together, or will the situation regenerate? What will happen to the ghouls if their leader is killed? The player doesn’t know what will happen, but she can weigh the possibilities.
In Cave Story, though, it is not really foreseeable that speaking to the Professor rather than ignoring him will cause him to die, and it’s totally unclear that the choice is even available. And it’s made worse by the fact you mention, that you can’t easily undo the choice if you save afterwards.
My wife and I both almost stopped playing Fallout 3 because of the Tenpenny Tower story. To us it felt like utter cheating: the designers imposing their cynical ideology on the player no matter what the player did.
That’s funny because I often give up playing RPG games exactly because I feel the creators are ruining the believability of the game world by trying to make the player feel like he/she can be the master of every situation and convincing the player that a good outcome is always an option.
Tenpenny Tower sounds like the kind of dilemma/situation I wish RPGs were brave enough to employ more often.
I haven’t done this quest, but Shamus Young has a great condemnation of the quest on his blog.
The problem is that the way in which it always turns into a disaster is not believable. If you give the player the ability to go back and do things differently (i.e. save/load), there must be logical consequences to these actions. Tragedy is a valid scenario, but for tragedy to be *inevitable* the circustances must *create* and *reinforce* that inevitability. In Tenpenny Tower, they don’t; the tragic ending to getting everyone to cooperate is based only on the game’s constant attempts to shove its nihilistic “people are bad” ideology in your face.
Ever played alter ego? It pretty much does nothing but offer you choices (in this case in how to live your life). This game very elegantly gives you hints about the results of your choices, both before and after you’ve made them. The second part (much like your depressured wing example) solves the additional issue where you have to play a game x amount of times equal to the permutations of your choices. However, a small afterward message can give enough of a hint that makes this unnecessary to all but te most die hard of fans.
Incidentally this is something starcraft 2 does somewhat badly by having the universe somehow bend so that each choice you’ve made was automatically in hindsight the right one.
Alter Ego sounds interesting.
I’m divided on the idea of after-the-fact assignment of consequences; some days, I think the world should be established ahead of time, but on other days I’m okay with giving players the illusion of agency while still making sure they don’t get left out for making the wrong choice.
It’s freely playable at playalterego.com (I just discovered!)
Saves you from having to dosbox it.
I’ve just read that it was designed by a psychologist who conducted over a 100 interviews about the most interesting life experiences of people and kept the most universal stories. Those stories are the game.
—–
Good thinking. I believe the illusion of agency is one of the most powerful storytelling tools in a game designers arsenal. On the other hand… it’s less replayable. On the third hand… a good story is infinitely replayable.
Do you think players resent illusion of agency on a second playthrough?
I think that players are quite likely to get resentful if they realize their agency was an illusion. It depends, though, on how the game is presented; in a lighthearted game or one that’s light on lore, I think it’d go over fine.
I think that the design decision probably has to do more with difficulty curves than with trying to give the player agency. If you talk to Professor Booster you miss out on the Booster 2.0 and lock off the best ending; but you get the Booster 0.8 much in time to use it in the Core fight. I’m mediocre at platformers, and I think I would’ve had the devil’s own time getting past the Core without the booster. (As it was, I had the devil’s own time getting through the waterway, which I think is kind of supposed to be a relaxing area. Could never avoid the last set of spikes, and the changed control scheme for Ironhead threw me for a loop.)
So the thought may be: If you can’t jump across that gap, you need the booster, now. And you won’t be able to get anywhere in the Sacred Grounds, anyway. (Also, I didn’t find the jump that hard — I think I could make it about one of three times, without the Machine Gun.)
This isn’t the only case where you can trade in a resource for a boost early in the game, locking yourself out from an improvement later in the game. If you trade the Polar Star for the Machine Gun or Snake, you won’t be able to get the Spur later. Here the player does have more agency — you stole that gun and shouldn’t be giving it away, and the layout of the Mimiga village gives you a hint that you may be returning to wherever you fell from. But it’s the same dynamic as far as in-game difficulty goes.
I think Pixel probably envisions the experience like this: On your first playthrough, you won’t jump that gap. You’ll get the booster 0.8, which you’ll need for the core fight, but a lot of bad things will happen to your friends and allies. The message in the prefab hut will clue you in that you need to do something else with Professor Booster. (Though I’m not sure how you’re supposed to get what.) Then you’ll play it again, and you’ll be better at the game, and you’ll be able to make that jump and beat the Core without the booster, and do more to get the best ending.
One way of thinking of this is that the plot branches are more like Easter eggs than attempts to give the player agency. If you explore more, you can find different byways in the game like the Littles’ room and the door in the Waterway (I’ve been to the first but not the second, and it turns out it’s pretty much impossible to get back from it with the 0.8.) And you may start to wonder what happens if you make that jump — the layout of the area very much suggested to me that I should try to jump across. If you do, you’ll discover a whole new plot branch; but I don’t think you’re meant to feel guilty for not having found it the first time.
There’s a different problem here, though, which is that blocking off the Booster 2.0 makes the very end of the game harder for worse players. (Though I understand that players with the 2.0 have to face a more difficult Last Cave.) I still haven’t beaten the final bosses with the 0.8, though it’s been a while and I’ve been meaning to give it another shot.
I totally agree with you on the difficulty tradeoff issue. And that’d be a great justification for this situation… if it played like a challenge gate, where you were given the option of bypassing this challenge or making a sacrifice to get help.
As is, however, a single failed jump means you have to accept the Booster 0.8 or reset. And it’s not even a fair challenge; there’s a block placed in an unexpected place that makes the natural jump (jump at maximum speed at the last moment) fail. You actually need to jump early to make that attempt. And it’s entirely unclear that it’s a challenge, especially if you get the Booster soon afterward.
The way the save system works is what really ruins that method. If you could save off a backup save from in-game at that point, or even better, if there was an autosave at that decision point, it would make it less arduous to come back on a second play-through. Cave Story is hard. It’s tricky even on easy mode on the Wii. Playing through a second time is a significant time and effort investment, especially when it doesn’t seem very strongly-clued that you can get a different ending or how to do it.
Even better (from a difficulty-management perspective) would be to make the Booster 0.8 pickup revokable somehow. Say you could take the Booster 0.8 and leave the Professor stranded, then come back to the Labyrinth later and give him the Booster back, but you have to make that jump and maybe pass another test before you get the 2.0. That way, if you gain expertise, you can trade in for the better tool and the better ending.
I agree completely about the save system, which is a pain anyway — I was just thinking that I’d like another shot at the Waterway, but I can’t do that without nullifying my save. I probably will try playing again sometime after I finish it (though man, the last boss rush is kicking my butt), but it’d be nicer if I could practice more in some of the less vicious parts — which is basically what your suggestion for giving the booster back to the Professor amounts to.
I will defend the jump this much, that there’s a mark that shows you where to jump from. Also I think there’s a save point just before this jump, so resetting isn’t very punishing.
There is such a mark, which is useful if the player realizes that the jump is possible, decides to try and make it, and recognizes the significance of the mark. Better than nothing, I suppose.
Yeah, it’s still pretty Easter eggy.
To be honest, despite being an avid indy gamer, I didn’t dare play cavestory for the single reason that I’m afraid to have to replay large stretches for missing this moment (and hypocritically not wanting the moment spoiled)
I can enjoy masochistic experiences like that only if they are intentional (I’m looking at you ‘I wanna be the guy’)
In general I don’t feel like I’m handled gently as a player by most games. Are your experiences similar?
I don’t know about “gently,” but I’m a big supporter of Ernest Adam’s Player’s Bill of Rights, and I don’t believe that this set of rights is often upheld by developers.
That list is very helpful. Thank you.
I enjoyed daggerfall which had a manual with a note from the designer to the players: “You know, sometimes bad stuff happens. That’s okay. That’s part of this game. Sometimes good stuff happens later because something bad happened before. So refrain yourself from saving and reloading and take the whole experience in stride”, which helped a lot in the enjoyment of the game.
(in this case getting caught stealing is necessary to attract attention and become part of the thieves guild, for example)
I would contrast Cave Story with a game like Silent Hill 2, which analyzes your gameplay style to guess what sort of ending your character deserves. Instead of making conscious choices like which route to take or which character to believe, it made use of the kind of stats you’d see at the “Mission Complete” screen after beating a game. It was a really interesting concept, but there were issues, at least for me, where it was taking my choices a little too literally. I can get more specific if you don’t mind spoilers.
I’m fine with spoilers.
There are three endings to the game (plus a joke one). The main two depend on how you treated a character who accompanies you through most of the game — as you eventually learn, she isn’t real, but instead a creation of the town itself. She’s an idealized and sexualized version of your wife who died recently. There are no dialogue trees, but if you stay around her while you play and otherwise seem fond of her, you get an ending where you accept this creation as real. Likewise if you don’t, you get an ending where your character accepts actual reality. This has the side effect of that you can induce yourself towards getting a “happier” ending by hitting your companion in the face with a two-by-four.
Where I stumbled a little bit was the third ending. There is a character you meet who is suicidal; she gives you an item at one point that turns out to be totally useless. If you spend a lot of time looking at it or trying to use it — *and* you leave yourself injured a lot during the game — you get an ending where your character commits suicide because he can’t deal with his own situation. My problem was, I really wanted to help out this other character, so I fiddled with her item a lot. And because this was survival horror, I tried to use healing items as infrequently as possible. So it felt kind of unfair that I ended up with that ending.
To bring things back to my original point, I really liked the idea of tracking choices that players make unconsciously. It is perhaps tricking the player — I certainly felt annoyed with it — but I think it leads to some more interesting possibilities than the more common plot branching techniques.
Tracking choices! That’s awesome. But as you said, in your example their triggers completely misinterpreted your motives and even actions.
And if you consider their triggers, it isn’t hard to imagine a thousand other misinterpretations.
The elusive goal would be to craft an experience where you can interpret the players actions and motives very precisely.
I haven’t seen a better execution of this than planescape torment. A simple question of “Do you believe that everything we do matters?” allows such as answers as: “Truth: yes, Truth: no, Lie: yes, Lie: no”.
And will possibly change your alignment, your relation with the character and because planescape is a world where belief manifests physically, it might change the world.
I also agree that Pixel didn’t seem interested in providing any particular player choice here. As far as I can tell, aside preventing you from accessing the hard route on a first playthrough, it’s simply intended as a homage to obscure secrets in NES games that you might have had to consult a magazine or your friends about.
This is one area where the homage hurts the quality of the design, though (in other areas the homages are extremely well implemented, and indeed many aspects of the game are fairly progressive for its genre), and provides a missed opportunity for a more integrated and meaningful scenario.
By the way, note that you can backup your save file called Profile.dat (or something similar) in Cave Story’s folder on your hard drive, effectively giving you infinite save files if you want to experiment and save time.
I agree with your perception of Pixel’s intent.
Unfortunately, I can’t backup saves that easily, since I’m playing on the Wii, but I did copy my save to an SD card before a particularly iffy choice.