obediah said:
So having to reload when a character dies doesn't offer enough risk, but not having to do anything because the character pops back up like a fair game is okay?
Who said anything about "pops back up"?
He can spend weeks recovering. It can take resources to heal him. There can be any number of downsides short of actual death.
The key is to make it all interesting - the penalties need to be interesting, there should be interesting flavour consequences, possible indirect benefits etc.
If we're talking about a tactics game, I still think frequent reloading is not good (but less of a problem). In an RPG it's a huge immersion breaker, and adds nothing to the story.
I'm not necessarily suggesting that death should never be used. I am suggesting that it shouldn't (often) be used where it's going to trigger an automatic reload.
Death / near-death should be treated like any other feature: what does it contribute to the game? It needs to be reasonably credible (within the realm of dramatic license), so a character just getting back up after "death" is stupid.
A character happening to survive unconscious and badly injured is reasonable. This happening quite a few times becomes unlikely, but certainly not impossible - an RPG is probably intended to be an unlikely story.
Once you've dealt with the coherence issue, what does death give you? Just an opportunity to say "Shit - I failed. Let's reload." That's every bit as inspiring as failing an important quest and having the game arbitrarily end there with a big "You Failed" notice.
It's nearly always preferable to allow failure of the quest (with whatever horrible consequences), but give the player a way not to lose the game. Failing the quest can have all sorts of consequences - many of them bad for the player -, but requiring a reload is the least interesting way to handle things.
The same goes for death. If the design assumption is "The player reloads here.", then it's not great design.
Where you do include actual death, it should have interesting consequences beyond the player's personal reaction. Like any feature, it should contribute to the gameplay and to the story - or why did you include it? Coherence alone isn't a great argument.
It's quite possible to get tension without the negative consequence being death. It just needs to be something very undesirable for the party (to create tension), but not too undesirable for the player (by virtue of interesting consequences - so that he doesn't want to reload).
For 95% of players, reloading after a character death is fine. They've set their standard at getting through it with everyone alive, and they failed, so they try again.
They do it because that's what they're used to, and that's what's required to play the game. That doesn't make it an ideal situation.
Maybe it's out of style these days, but I when I grew up people enjoyed something that challenged their skills, rather than just sleepwalking their way to an easy victory.
It's possible to have challenging tasks with undesirable (for the party) medium/long term consequences for failure, without death.
In a totally linear game, not including death would clearly be a bad idea - since the "sleepwalk to victory" argument is reasonable.
A good RPG should not be like that - you could sleepwalk to the end of the game, but you might well have screwed everything up on the way and failed to achieve any of your aims (short of getting some, however inappropriate, ending).
Death is an inherently black/white concept, and therefore a rather blunt tool in one sense. I'm sure it could be used effectively in an RPG, but "reload trigger" usage is not effective. Challenge and tension can be created in better ways - ones which leave the player in the game world, not watching a load screen.
[Note also that encouraging re-loading necessarily encourages frequent saving. It's possible to be much more involved when the thought "When did I last save?" isn't constantly entering your head.]
Character death does add that extra something for ironman style games as well. Death as handled in KoToR makes the game feel more like a book.
For ironman situations, I have no problem with death - though it should still get interesting responses from the game world. The trouble is that the vast majority of players don't play ironman, so you need to cater for that.
You could add permanent death as an option (intended only for ironman types), but then do you put resources into handling it properly - i.e. alter reactions, dialogue, balance etc. based on death? That's a waste unless it's your main focus.
For squad tactics games, of course real death is a good thing.
I'll stick to my earlier statement - when the designer is assuming "The player reloads here", death is a bad thing in a non-linear / multi-character game.
If the design assumes (and supports and encourages) the player continuing past a death, then by all means have them dropping like flies.
As for KOTOR, I haven't played it. I'm assuming that (non)death is handled in a ludicrous fashion, which I don't support. It's easy to get real death right, but hard to get non-lethal consequences right. A bad implementation is probably worse than true death, but that's no argument against using a good one.
interesting and labor intensive ways to punish a player for a character death, which the people that reload to avoid death would still reload to avoid
Not necessarily though. Some people will reload, I quite agree, but certainly not everyone.
Nearly every game makes things interesting where the player character(s) is successful, and less interesting where he fails. This means there isn't just a direct incentive for the player to do what's best for his character, but also an indirect incentive - since success is interesting and failure is boring.
This is self-fullfilling, since designers know that players won't take the "failure" path, so designers put few resources into such paths, so they aren't too interesting, so players don't take them...
To prevent reloading on failure (in most people) you need to desensitize the player to it - you need to create a culture of (occasional) failure. This way the player expects things to go wrong, expects negative consequences (with some positive ones), and won't view it as a mistake as such - just an illustration that shit happens.
It's a similar situation to introducing the idea of important consequences. For example, in the Gothics, there are choices all over the place that open up some options and transparently close off others. It's not a shock when an option gets closed off (as it would be in e.g. Morrowind / Oblivion), since that's part of the way the world works.
The player expects it, so he accepts it.
If the "You can always succeed at everything" culture were dropped from games (or throughout one game), I'd expect there to be significantly less reloading. First because the player would consider failure as something that happens, but perhaps more importantly because he knows from experience that the designers have catered well for it.
There'd be no reason for the player to think "I've failed at X, so I'll be missing out on game content / reward Y."
Rather he'd be thinking "Damnit, I failed - I wonder what interesting places this will take me."
I do admit that the chocolate-milkers of this world will probably still reload, but there really is no helping some people.
"death's door" type situation. From AD&D, let your character fall between 0 and -10 and suffer some of these setbacks. Let him hit -10 and he's dead.
It's a thought, but what does it add (the actual death I mean)?
Presuming that the "death's door" situation has interesting and undesirable (for the party) consequences, why not leave it at that?
I'd have thought it's likely that the situation will already be tense - if it's a combat situation and one of your number has been incapacitated, the likelihood is that others will be under significant threat anyway.
Why not just assume that anyone below 0hp is at death's door, and that no enemy decides to kick/stab them to death where they lie?
The death's door situation just needs to be made interesting and given significant penalties, rather than some absurd instant recovery.
). 
