I sense in your question (as well as the fact that you’ve not yet rewarded one of the many excellent answers) a desire to strike that delicate balance between retaining risk of death in gameplay with an equal desire not to disappoint players with a meaningless demise.
I think what you’re really after is an approach that works broadly,
becoming an automatic part of gameplay, not a custom fix for a
specific character death.
First let’s analyze the problem. For a game to be a game, victory must be possible, but for victory to feel meaningful, loosing must also be possible. In role-playing, loosing means dying. If the players feel they can’t die, they can never really feel they’ve won. This is the issue of winning vs. loosing. But role-playing is complex. There are more issues at play.
- Investment in Character -- Whether the players created a detailed backstory or not, the longer they play a character, the
more attached they become. It’s important to honor the player’s
investment in their characters.
- Staying in the Game -- This one’s obvious, but important to keep in mind. When a character dies, the player doesn’t get to play
anymore. Whether dying in an RPG, or going bankrupt early in a game
of Monopoly, everybody else get’s to keep playing, but your fun is
- Preservation of Narrative -- More than gameplay, RPG’s have a story. When a character dies an abrupt and meaningless death, it
can break the ongoing plot - both in terms of that character’s arc
and potentially the greater story line, which is annoying for
WHY I THINK THE OTHER SUGGESTIONS DON’T WORK
Purchasing Life -- Several answers suggested approaches for allowing players to buy back their character’s life such as sacrificing magic items, paying local clerics for resurrection, or reducing stats to appease the gods. Put simply, this is commerce. Whatever brings the character back becomes a kind of ‘death money’ that players can save up and use again and again. This is an effective way to mitigate death, in fact, it’s too effective. Really, it removes death entirely and replaces it with resource management. As long as the party as a whole keeps a reserve of their death money, they cannot die.
Death is not Death -- There where a couple of creative ideas about allowing players to continue using their characters after death in the form of undead spirits or visitors to the afterlife. In truth, I loved these ideas, but they don’t address your question because by their nature they are subplots that cannot repeat. Also, these subplots are as likely to disrupt the campaign’s main plot line as augment it.
-- Some suggested translating fails from death to incapacitation such as a coma, perhaps requiring additional saves. While not a bad notion, it doesn’t address the victim’s desire to stay in the game. Further, since the party must care for the fallen comrade, it tends to take away their ability to proceed with the campaign disrupting their play. And in the end, the player might just die anyway. Like Death is Not Death above, this would get old if it happened more than once.
Multiple lives -- Perhaps borrowed from the world of computer games, the idea is to give each character multiple lives. I think this just kicks the problem down the road. In terms of a meaningless luck-based death, it doesn’t really matter that a character might have cheated death before. Once they’re down to their last life, the issue returns. Worst, those PC’s with several lives in the bank are likely to engage in wild, risk-taking behavior potentially disrupting play.
Backup Characters -- Having a backup character ready to go solves the issue of staying in the game, but it doesn’t address preserving the narrative and investment in character.
Almost any approach mentioned above will work well a few times only or
fast becomes a crass workaround - a familiar ‘trick’ players can and
will use to confidently cheat death, effectively taking death out of
SO WHAT WOULD I SUGGEST?
Luck Points -- Each character has 6 luck points that they can spend whenever they wish on whatever they want (within play, not character creation or level advancement). Each luck point modifies a die roll by 1 (or 5% on a percentage roll) in the player’s favor. Luck points do regenerate, but slowly - say 1/day or 1/adventure. Further, luck points do not increase with level and cannot be magically replaced or augmented.
I’ve used this a lot and predictably many players tend to bank them for saving throw rolls. Other common uses are assuring a successful to hit or damage roll at a critical moment. What I like about them is that players choose when and how to use them and since there aren’t many of them, it keeps death and danger a possibility.
Play Two Characters -- Kind of like combining backup characters with multiple lives, the idea is that each player actively play two characters during the campaign. I think at first blush, this notion rubs a lot of game masters wrong, but I found it works really well.
Firstly, it seems to increase role play. Both in backstory and play, players must address how these characters know and interact with each other. Unconsciously players will choose one of their characters to be dominant which leads to interesting ideas of relationships - master/apprentice, employer/bodyguard, siblings, spouses, or just friends. Also, it’s good for rounding out a party in terms of character classes.
Secondly, it nicely addresses sudden death. If one character dies, the player can keep playing with a character that is theirs, is established in a role playing sense, and is part of the campaign’s narrative. Yes the player was invested in the departed character, but also in the surviving character. GM’s may then allow players to create a new backup character (or not) as they see fit. On the odd chance both of a player's characters were taken down at same time, the GM could give the player a break and let them keep one of the characters.
There are a few things to keep in mind with this approach. If you have a lot of players, this may create too many characters and overly lengthen combats. Inexperienced players may find two characters a lot to keep track of. And parties who typically display poor teamwork might factionalize into pairs. For the most part however, I’ve found the two character approach very effective.