henke on 2/12/2010 at 17:46
Quote Posted by Pyrian
I'm really not clear on what point, if any, is supposedly being made?
Quote Posted by Papy
I'm not sure what is the subject of this thread.
Ok, I was thinking about how games, at their core, are about winning, and doing the absolutely best you can. But for a narrative to be exciting the protagonist can't always be right. He can't be infallible. And how, because of this, the game and it's narrative can be at odds with each other.
That said, I'm not sure I can figure out or put in words the finer points of this theory.
In fact I thought it was obvious by the "Discuss" at the bottom of my post that I was hoping TTLG would do the better part of the legwork in figuring this stuff out. :p
And interesting points have indeed been raised. Pathose and Aja and demagogue, I like what you guys are saying. Sandbox games. That's where it's at! Trying to enforce linear narratives on a non-linear medium and if the player doesn't do what the game wants him to do he'll feel like he's not follwing the script correctly. But put him in an open world with no story and his mistakes and accomplishments will feel more real because they're the result of his actions. They're not just happening because they're in the script.
Pyrian on 2/12/2010 at 17:46
EDIT: Okay, that's a rather different and IMO more interesting point than what I got out of your original posts. I'll try to write more about it later today. The rest of this post was based on previous stuff and not directly relevant to the topic of narrative strength. 'Cause I agree that the player always winning is not a strong narrative, and that most of the alternatives don't work either, and would like to extend that a bit.
Quote Posted by henke
Quickloading is a huge imparment against letting a protagonist have flaws. All too often if players have discovered that they've fucked up, not done something in the best way it could've been done, they will quickload and do it over instead of just living with their mistakes.
I don't buy that this is the fault of quickloading. Instead, it's a question of training. We're accustomed to failstates being endgame. Take the Deus Ex example; if you get shot down in the hotel, the natural thing to do is hit quickload. That's what you've been trained by the game (and every other FPS) to do when you get gunned down. It's not a refusal to accept that you've lost in-game, it's literally a failure to realize that it's even an option. At least, that's how I reacted. Any other time in that game (and most others) you get gunned down, you have no choice but to reload from an earlier save.
In fact, I'm a bit unconvinced that it's even a problem per se. Let's say I use twice as much ammo as necessary shooting my way out of a situation, and a few medpacks as well. Do I continue with my "flawed" play or do I reload? Generally speaking I will continue, although I certainly have the option to redo that section until I get it perfect (or close enough). Even if I use virtually all my ammo and health I'll try to push on, but I typically won't quicksave until I've gotten back on my feet on the principle that if I die I can re-do the section that was primarily responsible.
Some people play very differently from that, and will reload from virtually any flawed performance. (Typically these same people also play on higher difficulty levels...) And y'know what? That's okay. Some people will even consult walkthroughs before they start in an attempt to get some optimal path on their first (and, in these cases, usually only) playthrough. And y'know what? That's okay too.
Far Cry 2 was pretty fantastic in this regard. Buddy dies? You've got others. You go down? A buddy will pull you out. ...Better be careful until you can reset your backup buddy! If you don't want to abuse loading, you don't have to.
Games with percentage rolls are another matter altogether. Lost a round of gambling? Reload and try again. Didn't make the 15% chance to persuade an NPC? Reload and try again. I think that mechanic encourages reloading TOO much.
Sulphur on 2/12/2010 at 19:42
I'm not particularly for a sandbox game where you 'create' your own narrative because, fuck, there isn't the freedom to do so. You can do stuff like toast a couple of civilians and have an epic chase across the city and then hop onto a speedboat and whip off the beach and turn around and launch your boat into a helicopter and jump out at the last second and parachute away to safety, and that's all so very cool...
...but empty at the same time. You can run and hide until a heat meter cools down, so what? There's a reason why so many sandbox games either suck at narrative or have primary missions that put you on a linear course to completion. They're sandboxes, and any narrative you 'create', including all those cool gameplay story vignettes, doesn't hang together one whit.
It's arguably harder to do a credible protagonist in a sandbox game because while you may choose to run over a pedestrian or hide like a nancy boy in a street corner hoping the police don't find you, none of your actions can have any effect on the game's primary narrative.
You want to be able to play a flawed protagonist, a person who can be weak, who can fuck things up, who has the choice to, and the game's story carries on despite it? Play Heavy Rain. I'm serious. It ticks all those boxes and it's not a sandbox open-world game.
vH: SH2 isn't it; SH2 is commentary on the blind trust you place in the protagonist first and foremost. You don't choose for James to be a dick, that's forced upon you -- you only learn that after it's too late. The moment near the end where you learn the truth about James radically changes everything you thought the game was about from that point on.
SubJeff on 2/12/2010 at 23:42
I disagree with the title of the OP, totally - games can and do have flawed protagonists.
In fact the most interesting protagonists are flawed because the flaws are what makes them human, and therefore real, to us. Garrett is a good character because he is so flawed, not in spite of it.
This is a very different concept to a protagonist who, whilst under our control, only ever does the right thing. Plots tend not to lead us down a bad path and even games like GTA try to have some sort of moral rationale for your actions as per the plot - within the framework of the setting. There is always someone who you think is a bigger dick than you, who needs offing more than you, who has lied or double crossed you/your crew/some boss and so is morally lesser than you. So yeah, I agree with that point. I guess devs can't really stomach the idea of us being the bad guy, not really and certainly not by their design (though you can sandbox it all you want). And if they did do it they'd run the risk of the game(s) being criticised on moral grounds, more so than they are now. Imagine that!
But a non-flawed protagonist is like a guitar solo that just repeats itself thrice; no matter how technically or artistically awesome each part of it is, it's boring.
demagogue on 3/12/2010 at 02:30
Yeah but his point wasn't that your character couldn't have flaws that come out like in cut scenes, but that games usually don't allow character-flaw actions in the actual game-play to have any meaning in the game's story or plot-progression. It's not often recognized as you make progress towards anything.
And I'm acknowledging that sandboxing is a bit of a cheap way out of it, if you want to play a game with a strong story. I still like my example. There's the terrorist incident and you keep your head down and don't interfere and play the "coward"... Then when you get home your wife is like "I heard about the incident. Thank God you kept out of it and are ok," so you're "rewarded" in one way, but you go to the office the next day and your colleagues wonder "Why the hell didn't you do something? You're a trained officer and have a responsibility..." and are "punished" in another, so it comes out sort of a wash in the plot.
Deus Ex is one example where they usually tried to handle you fucking up the plot line, so you could kill a major character early on, or go in the woman's bathroom, and the plot would roll with it, or it would let you know if it didn't like you being too lethal. Though it could be clunky with that sometimes. (Bioshock tried some of that too, though arguably even more limited & clunky).
I also recognize the infinite-plot-tree problem with this; you can't account for every little thing, and you still need a plot that goes *somewhere*. But games could do more to make you feel like you're showing your character in your game-play, not just leaving it all to cutscenes and your only role is pew pew pew.
Phatose on 3/12/2010 at 03:16
In linear, narrative driven games options for failure typically aren't worth creating. Same principle at work as what's driving games towards greater accessibility - not worth spending development time and money on things most players will not see.
Players have been trained for years that objectives are to be completed, unquestioningly - to the extent where it's not uncommon for players to happily go kill someone for no reason other then a little objective message popped up. Failure to complete those objects results, at best, in the player being punished down the line for it. Typically, failure simply isn't an option - either you get an instant game over, or you can not proceed until you complete the given objective.
So, if you program an acceptable failure, most players will do everything to avoid it - and if they happen to fail and it lets them continue, many if not most will reload an old save figuring that failure will result in some kind of punishment down the line - missed items/achievements/power ups, or harder missions later. You end up spending a lot of time creating a path which nearly no one will actually see.
Add on to that the reality that some objectives have to be "Failure is not an option" or the narrative can't work, and you end up with a situation where sometimes you can fail and sometimes you can't and it's not always clear, and players are going to tend to try hard not to fail just in case - again, you end up with players who won't see your hard work.
So they don't get built. Failure remains not an option, reinforcing the notion in gamers.
To really be able to pull it off, you'd need to make failure clearly an option when it is, and you'd need to make sure it's crystal clear to the player they're not going to get punished down the line for it. And you need to do so before they go for the 'reload data' button. Sounds like a right royal bitch to do without coming across as ridiculously forced, at best.
Sulphur on 3/12/2010 at 06:21
Quote Posted by demagogue
And I'm acknowledging that sandboxing is a bit of a cheap way out of it, if you want to play a game with a strong story. I still like my example. There's the terrorist incident and you keep your head down and don't interfere and play the "coward"... Then when you get home your wife is like "I heard about the incident. Thank God you kept out of it and are ok," so you're "rewarded" in one way, but you go to the office the next day and your colleagues wonder "Why the hell didn't you do something? You're a trained officer and have a responsibility..." and are "punished" in another, so it comes out sort of a wash in the plot.
I love the example, but the problem with it is that sandbox games don't allow for that most of the time.
The scenario is that you're roaming the city and a side-mission becomes available - the terrorist incident, for example. You trigger it, then choose not to do it, and go back home. The game gives you a failure screen and the mission becomes available again. There's zero acknowledgement from the game or from any characters about your choice because the game isn't designed to use that information.
Sandbox games are about the player creating his own distractions, and none of these affect the game's overarching plot in any meaningful way.
Sulphur on 3/12/2010 at 06:49
Quote Posted by Phatose
In linear, narrative driven games options for failure typically aren't worth creating. Same principle at work as what's driving games towards greater accessibility - not worth spending development time and money on things most players will not see.
Players have been trained for years that objectives are to be completed, unquestioningly - to the extent where it's not uncommon for players to happily go kill someone for no reason other then a little objective message popped up. Failure to complete those objects results, at best, in the player being punished down the line for it. Typically, failure simply isn't an option - either you get an instant game over, or you can not proceed until you complete the given objective.
One of the key things to ensure with acceptable failure is to not give a player that game over or 'mission failed' screen if he fucks up. A mission debrief right after, for example, pushes the game forward and lets the player know he can continue regardless.
The crux of the matter is, like everyone's said so far, that people are conditioned to reaching the win state in the end, because that's the whole point of any game. So you're going to have some people reload anyway to reach that optimum state.
So the best thing to do is: change the objective of the game around. Is the player supposed to win the game, or see it through to the end? Are you supposed to pop people in the head like a skilled robot, or make choices that you can fuck up here or there?
If it's the latter, the game needs to make this clear via feedback. In this type of game, player skill should ideally not be what it focuses on. The focus needs to be on clear choices, and not the total number of headshots per level. The game needs to remove the mission failed/game over screens.
Once that's in place, instead of multiple short-term goals in a bulleted list, give the player one long-term goal and make him work towards it: you chose to do the mission this way, you fucked up, but that's okay because there's still another way we can capture the enemy/kill the traitor/save the world, etc.
There will still be an ultimate failure state if the long-term goal isn't won, obviously - it's up to the developer how 'losing' the long-term goal plays out in the end. But the road to the end can be littered with wins and losses, each of which, in turn, have had an effect on the options available and the state of the gameworld at the end.
Thirith on 3/12/2010 at 12:32
Sulphur, your post reminds me of the Wing Commander series and its branching mission tree. For an old game that was focused mostly on glossy presentation they handled this pretty neatly IMO.
demagogue on 3/12/2010 at 15:11
Quote Posted by Sulphur
I love the example, but the problem with it is that sandbox games don't allow for that most of the time.
I wasn't clear with my first sentence. I meant it to mean "I acknowledge that you can't really do this in sandbox gaming, which is why I liked my first example, which isn't sandbox but has a plot tree-structure that takes this kind of stuff into account..."
I noted that Deus Ex did it a bit, and worried about covering every plot-trees branch, but thought that was no excuse not to account for some stuff like this. So basically I was and am agreeing with the whole gist of your last two posts.
