Player death and the suspension of disbelief

[Michaël Samyn]

I don't like dying. At least not in a game. And definitely not when the game doesn't end there.
"Game over" is a lie. The game is seldom over when it says those words. Usually it means "Start again, you clumsy fool".
And I hate that! If I'm clumsy in a game, it's because the designer made it so. It's not my fault. I would not have made the same mistake in real life as my character did in the game. The only reason why I made the mistake is because the designer did not give me a proper interface (or even worse: proper information).

But of course, the designer did this on purpose! Because it's a game! And it wouldn't be a gmae, if you couldn't lose. And it has to be a game. For some reason.

Player death/Game over is most painful in games that do make an effort to immerse you in a narrative atmosphere. There you are, thoroughly believing the world you're interacting with, really enjoying it and then you make one false move and you're zapped out of the world, and you have to start again. This completely ruins the immersion. And what for? The story doesn't end here (otherwise I wouldn't be forced to try again). I'm not supposed to die in this game. Yet I am. Because it is a game.

I'm all for abolishing the death penalty! If the story is not going to end when the player character dies, then the player character should not die. If you really need to punish the player (make sure it makes sense in the narrative, though), find some other way. Something that does not break the illusion, or better even something that enhances it.

[Thomas]

This is a really interesting issue! How did you like the system in braid where you could easily rewind? I think Braid is an interesting example, since although death is gone, you still need to get have obstacles (that sometimes are very tricky). This means, that having rewind does not take away the problem. At least in my opinion.

I have written a post a about this a while ago:
http://frictionalgames.blogspot.com/2009/07/nothing-will-save-you.html

In our upcoming game Amnesia we will also be a addressing this a bit. While there is still death, the game always changes after death, so the player never know what to expect. There is also no Game Over screen or similar, so the player stays in-game.

[Michaël Samyn]

Braid just exposed player death as the mechanic in gameplay that it had become (or always was?) and then built puzzles around that. Since Braid is a highly stylized game, I didn't mind too much that I died or didn't. It didn't make much of a difference whether my failure was depicted as death or not. In essence it was the same structure. Not that i played much of Braid. Like so many other games, it simply forced me out by putting obstacles in my way for no other reason than it was a game.

I guess I have a problem with failure in interactive works. I don't think there should be such a thing. They shouldn't be tests, they should be experiences. And everything I do as a player, should be dealt with by the system in an interesting way. I don't like the idea that the system knows what I'm supposed to do and then I just have to do it (over and over again until I get it right). If the system knows so well, why doesn't it do it? I want to play with the computer, not against it.

I'm quite curious about how you will deal with this in Amnesia, because I must admit that this is the reason why I couldn't play Penumbra. I want to, I really want to. I'm very attracted to the story, to the visuals and the audio, and to the interaction mechanics. But making me die just because I did something wrong is something I can only handle once or twice before I close a game.

[Michaël Samyn]

I've read your post, Thomas (or reread it? -it sounded familiar).

I think the reason why you would even consider including death/failure in your games is because you feel that the emotions coming from interacting with the game system, somehow relate to the emotions that the story is trying to evoke. As a designer, I dislike or even disagree with this concept. Fahrenheit did this to a large extent: making you move the controller rapidly so you would supposedly feel as exhausted as the character in the game. I think this is mistaking the reality of the player's situation with the fiction of the game and in facts breaks the illusion more than it reinforces it.

Knowing that you can lose a game does not make your connection with its story stronger. Instead it forces you to focus on the game as a system, and effectively takes you away from the game as experience, as story. Yes, it might make you feel anxious but not because you're concerned with ghosts or murderers but because you know you didn't save the game and you'll have to restart the whole level if you fail. And there's not much meaning in that.

Movies don't need to kill you, or shock you or threaten to rewind to make you feel fear. Books don't have teeth to bite your hand off when you're not scared enough. This would be silly and undesirable (and even ineffective as per above). In interactive entertainment too, we are dealing with fictions, and the emotions our audience feels need to come from the fiction and their involvement in this fiction through their imagination. Otherwise we force them to focus on the system and then they are interacting with a controller and a processor, and not engaging with a story.

[Thomas]

Regarding Amnesia:
I do not want to talk too much about Amnesia yet, even though is a pretty small forum. Wanna keep some secrets for release One of the thoughts I have had in my when designing stuff is that my father should be able to play it, yet we also wanna cater to a more general audience, so trying to find some kind of middle ground. What we are essentially trying to do with death is that the game never stops or repeats, the player should never find herself doing the same thing over and over. Instead what the player needs to do changes so that if you just push forward you will eventually move forward in the game. And while doing this, the game never repeats the challenges exactly the same. Do not except a revolution, but I think we are moving a little step in a new direction

Regarding blog post:
It was a while since I wrote that and I actually lean a lil bit more towards your views now. I think that the entire failure system that is built into games is by far its greatest baggage. I also think (like you say) that high tension can be made even though there is no risk of failure. As you said, failure is a bit like exposing the hidden mechanics and unveiling the illusion that the player is in. "See, you thought you where in a fantasy world, but as you can see it is just a simulation which will always repeat itself over and over again.". Failure is also a large reason why so many people do not play games, they simply can not move on.

This brings me to a another point: How would actions scenes look like in a game where there is no failure? I have a hard time imagining this, and I think there is a strong underlying bias in me, caused by 25 years of video game playing, that is holding me back. Of course, I would not like for notgames to produce the same kind of shooters that normal games do, but I think it would be interesting to discuss the possibility. Most (all?) games where there is no death, have cutscenes as soon as there is some action element, but that is no way to solve it. I am curious to hear ideas about this!

[Michaël Samyn]

I think action games might be the one exception. Which is why they are, more or less, the only kind of games that are being made. I don't mean to say that all games should abandon the failure "trick". It's a very suitable mechanism if you're trying to tell as story about failure (which is only the flipside of winning).

Action games appeal to a certain group of people who enjoy fiero. But since most people in the world are not fans of fiero, they, well, they don't play games. Also, most stories in the world are not about winning and losing. So they cannot be told adequately through the "failure format". For those stories, we need to find new formats.

If you want to make an action game, by all means make the player fail over and over again. They love that. But if you want to make something else, do something else, find something new. Figure out how the people who would be interested in your content work, what they enjoy, what challenges them. And create a design around that.
(I know this is easier said than done, but it's the general idea and for now, we'll probably have to work on instinct most of the time.)

[Thomas]

Still you might wanna have shorter bursts of action in games with no failure too?

For example, take the novel The Road, which is far from an action packed novel, but the "action sequences" that are in it are quite important. Should this just be be skipped for games that have no failure, or is there perhaps some other way to do it? This is kind of dilemma I am in myself, since horror games are mostly about non-action elements, but often requires small bursts of action.

I think you mentioned in some post that the problem you had with failure is that he protagonist did not do what you wanted him to or what you would have done in real life (due to controls). Hence the problem with normal failure based action is that the mechanics take training (meaning failing a lot) in order to get the protagonist more closely to what the player would really like to do. Perhaps having "simplified" controls and instead make it as a choice is a way to go? I think that there must be some kind of fear of loss and stress involved and the trick would be to do this without making the player focus on underlying mechanics.

If the actions is constrained to short bursts it should be easy to control it and hence make the world react to the player's choice, adding a sense of the possibility of loss, but without ever making the player fail. An example would be if the player while walking encounters some bad-guys and now has to deal with this. The player could just hide in the bushes and wait for them to pass, but if moving close enough the protagonist is captured and then requires the player to escape from another place. In the end, the player will end up on the same story path, but since there where unknown consequences depending on the actions at that point, it should instill enough stress and fear of loss. I think simply having an element of the unknown could be what it takes to create a dramatic action sequence.

This is of course only viable for short instances. The permutations for many of these kinds of situations would become too much in the end.

[Thomas]

Another thing I thought: Would it be OK to kill the player if a really bad choice is made?

I would like to compare this to adventure games. Early adventure games contained situations where the player could get stuck because of something done earlier. For example, if you forgot to pick up a key in a room, it was impossible to escape the cell you later found yourself locked up in. This is of course really annoying, so it was gotten rid of quite quickly. However, it still exist to a certain extent in IF games because you have much more freedom there than in normal adventure games. For example, the player can always drop an item, and doing this at a really bad time, can make the player stuck later on. We also have the same problem with the physics in Penumbra, in some scenes it is possible to do some really "stupid" choices that will make you stuck. The reason why this exist is because the game would be very restrictive otherwise and needed to be very linear, which we do not want. So the rule: "If the player is really that stupid he earns getting stuck" is applied at times. Generally, the player will need to want to crash the game in order for it to happen.

Lets now go back to action sequences and player death. Lets use the player encounters bad guys while walking example again. Would there be actions where it would be okay to allow death? One could perhaps even say that death would not be immersion breaking in these cases because the player reacts to irrationally that she cannot possible be immersed in game. For example, if the player just runs up towards the bad guys with no weapons or anything then perhaps it would be okay to kill her off and restart the game from some previous position? I think this might be a possible design, but there is always the problem that what a designer thinks is obvious for everybody might be far from the case. Some play testing should solve that though.

[Michaël Samyn]

Smart. The idea of being captured and having to escape turns the problem of failure into a puzzle rather than a twitch-based interaction. And puzzles are often a lot easier to blend with stories. But of course, puzzles come with their own issues: if it takes too long for a player to figure out how to solve the puzzle, they will be forced to "leave the story" and start focussing on the game system instead.

For example, take the novel The Road, which is far from an action packed novel, but the "action sequences" that are in it are quite important. Should this just be be skipped for games that have no failure, or is there perhaps some other way to do it?

Having never made (or even been inclined to make) an action game, I'm not the world's expert on this issue!
But I think you should only have failure as a game mechanic if failure makes sense in the narrative. The problem is that if your narrative does not involve resurrection or immortality, failure represented as death needs to be permanent.

Of course, in most stories, you don't want the protagonist to die before reaching the end of the story. So the solution is to represent failure in another way, not as death. If it's suitable for the story, perhaps you can hurt the character instead of the player. I remember feeling terrible about how my autonomous sidekick got hurt in Silent Hill 4, even if that didn't have any impact on the gameplay. Just seeing her bleeding was very painful and thus something I wanted to avoid.

This does not solve the problem of repetition, though. Repetition is not suitable for most stories. And in most games, repetition only confronts you with the systematic nature of the game. So either you figure out a way to make repetition an integral part of the experience (won't work with many stories) or you find a way to not have repetition (at the expense, possibly, of shortening the experience).

I think that there must be some kind of fear of loss and stress involved and the trick would be to do this without making the player focus on underlying mechanics.

This is easy: you just make sure that the "fear of loss and stress" come from the narrative and not from the game system. That what you lose is something you loose in the story, and not in the game-as-system. The advantage of this approach is that in most cases you won't even need character death.

[Thomas]

Some quick notes:
The capture does not needs to be a puzzle, for example the player can find the place deserted or what not. Or just have some puzzle-less escape. I feel that a notgame should try and avoid puzzles unless they are optional and does not hinder the player from progressing. Of course that is up for debate, but I would like games to be interactive in away that they give freedom instead of being in the way (which is what puzzles do). I am sure you agree more on this, but just wanted to state me position

Of course death should be avoided unless it fits with the story, but I think one could get away with a "The story should not end like this, should it?" thingie if the player does something truly stupid I know, best avoided, but sometimes it can give you more freedom as a desinger if you can allow for some immersion breaking but very unlikely moments. At least, that is my feeling after dealing with annoying physics for a couple of years

As for tension from the story, I think what one needs to think about is that bad things are out of the player's control. For example, if you walk through a dark corridor which ends with the protagonist dying, the death should be caused by outside forces (a killer jumping out from closet) and not from protagonist induced things (the character trips and the killer catches up). I think the latter makes the player feel cheated. Not sure though since I cannot really compare any experiences. All games are failure based so I guess we need to try this stuff out

[Erik Svedäng]

Very interesting discussion!

One thing I've been thinking about is that the majority of board games and competitive games actually handle death in a much better way than the games that try to tell stories.

Death in games like chess or Go is very finite (both for the pawns/stones and for the players). Same thing with Starcraft or Counter Strike. That also makes them much better since it creates a really huge tension. Losing there is also more fun in a way, at least to me. I guess it's because 1) I can play again without any repetition, knowing more than last time and 2) it's more of a ritual between humans than "doing a test"

The death that story-based computer games have introduced is so washed out compared to that, it's no wonder it only frustrates people. And even when I win such a game (by beating it?) the feeling is not very strong since it lacked both tension (because of the save functions) and someone to defeat.

I guess I'm basically repeating what you're saying. My point is that death in games can work beautifully, actually that used to be the whole point of them. The computer messed it all up.

[Michaël Samyn]

I guess I'm basically repeating what you're saying. My point is that death in games can work beautifully, actually that used to be the whole point of them. The computer messed it all up.

Actually, it's humans who messed it up. By falling into McLuhan's trap again. Every medium apparently needs to have as its content the previous medium. At least in the beginning. Early novels were like songs and poetry. Early cinema was like stage plays. Early television like newspapers and magazines. Computer entertainment does the same by being games. It's a phase. But a phase that has been warped by its commercial success.
 
I completely agree: games are things that you play with another person. And it's almost as much fun to lose them as to win. But videogames can't really be lost. Most single player games do not have a losing condition that is an acceptable outcome of the game. "Game over" is a lie. Because it doesn't mean the game is over at all. It just means you have to start over until you don't make the mistake you made. A videogame is only really over when it ends in the way that the designer has foreseen.

As games, videogames suck.

But, as "something else" they are wonderful! This is why we need to concentrate on that "something else" and not waste time making things you could make better with pen and paper, dice and cardboard.

[Erik Svedäng]

Actually, it's humans who messed it up.

Yep, that's what I really meant. Finding a local maximum and not searching further, instead settling for "good enough". Sounds kinda human to me for sure

As games, videogames suck.

Haha, someone should write a book with quotes of you

[Jeroen D. Stout]

I completely agree: games are things that you play with another person. And it's almost as much fun to lose them as to win. But videogames can't really be lost. Most single player games do not have a losing condition that is an acceptable outcome of the game. "Game over" is a lie. Because it doesn't mean the game is over at all. It just means you have to start over until you don't make the mistake you made. A videogame is only really over when it ends in the way that the designer has foreseen.

As games, videogames suck.

Whilst I agree we need to focus on more endeavours than repeating the 'game' trick over and over again, there are many games that I did enjoy, sans a human opponent. Old adventure games, for instance, are very much like ballet. There is the story and setting to maintain it, but we do not expect it all to make sense - it is meant to keep the 'raison' going. Something like Syberia included puzzles you may not enjoy, but I rather did. A game like Riven is an example the other way - I hated the game for being so unplayable. These days I love it for what it is; a puzzle-based game appreciatable now that I know how.

Videogames do not suck 'as games'... do not get me wrong, I still proclaim we need to expand the field and our understanding of it. Because many people like these different things. But that does not take away I like some videogames, such as Beyond Good & Evil. Sure, the narrative and gameplay do not make sense. But neither does a ballet. In your own type of writing, we need to accept that medium for what it is as well. If Beyond Good & Evil sucks 'as a game' that perhaps this definition of 'game' is unworkable. If it sucks 'as something else' then I may agree; if it sucks 'as a notgame' then I will agree also. But as it stands it is an enjoyable experience to me - I happen to like the challenge.

I hope you see my point - as much as you have a dislike of what videogames have become, it is not directly right to pull the term 'games' underneath from them.

[Michaël Samyn]

Maybe "suck" is too strong a word.

(But as Erik noted, it's good for effect. )