So, in playing through the end of Mass Effect 2 I got to thinking about character death, how it can happen, and what it can give (yes, give) to a game. You see, the end of Mass Effect 2 was a lot more tense for me as a player than the end of Mass Effect 1, not because the scale was grander or anything (I honestly think 1 had the more dire situation) but because I knew that anyone, and in fact everyone, could die if I didn't play through the end sequence right. That's not a spoiler, if you go through the achievements for the game it is clearly advertised that people can die. The knowledge though makes things tense, am I picking the right person for this job? Do I really want to bring that person to watch my back, what if they die? Hell, what if I (the Main Char) dies?
You can get those same feelings in Table Top games, or any game really. What adds tension to the dire circumstances scenario is the fact that everyone knows that someone can die. The tension increases as the characters that people care about are put in the line of fire. Will they make it? Won't they?
That being said, lets look at some of the approaches most GMs seem to take to PCs and killing them, though I think I may have already shown where my view is.
View #1: I don't kill PCs
This is a fairly common view point, and it stems from the fact that either the GM just really doesn't like killing PCs, doesn't want to take a cherished PC from a player, or has their plot woven so tightly around those specific characters that they can't handle one of them dying and de-railing the entire game. Now what this gives to a game is consistent characters (unless the player ditches them), and that means you can build story elements off of them and for them confident that they'll be around when those hooks comes up. However, at the same time, you need to find reasons to keep players alive when the dice come down wrong, and you lose that tension of a tough situation when your players figure out that you won't kill them, and they will figure it out eventually.
View #2: Let the dice fall where they may.
This is just about as common as View #1, and in my experience happens most often from GMs who are backlashing from being fed up with people not dying. The view point is dangerous, but at the same time it is fair. Dice are the decision makers in games ultimately, in combat your life comes down to dice rolls, and GMing like this means that random chance can just reach out and kill a character. What this gives to a game is the tension I was talking about. Larger combats will be very tense as players know that a bad day with the dice can be the end for them, and thus almost any combat can be fatal. On the other hand, this means you need to be good at weaving your plot points around characters that can die at any moment, you can't rely on someone being alive at X or Y location to do A or B action. Also, if you're not careful you can end up wiping your entire game with one bad round or two for them. A few criticals on your part, a few botches on theirs, and it is game over.
View #3: I actively hunt PCs
I personally dislike this kind of GM, and the kind of game that comes from this kind of GMing. It is adversarial, GM vs PC, and I don't have fun doing that as a GM or a PC. That being said, if you do like it, then have at. This also adds tension to the game, death may not just be around any corner, it is actively hunting you down and you need to survive. On the other hand though, the GM is the adversary for the game, and that means you shouldn't get too attached to your characters because it may be on the way out soon. I honestly can't speak too much about this kind of GMing, as I avoid it. It simply isn't for me.
View #4: I'll kill a PC, but only if it is meaningful.
This method is kind of a blend of Views 1 and 2. The GM will kill PCs, but only at meaningful times of the game. The big plot fights, the noble sacrifice, etc. Players get a bit of protection in some of the fights, but know that when it matters they are risking everything. This way you can build plots off of your characters with some certainty they'll be there when the plots come up, but you don't sacrifice any tension for those big fights where the players know they can die. This works very well for games when you are trying to tell a specific story with characters playing a certain role in it.
So which to use?
Honestly, depending on the kind of game you are running determines this. I honestly can't recommend view #1 or 3 because I just don't like them. With 1, the meaning and point of combat can be removed, there is no danger because the character is going to live and the plot is going to go on regardless of what I do. I personally like the tension of knowing death may be right around the corner. For #3, well, as I said in it, I don't like having an adversarial relation with my GM or my players when I am GMing.
But out of #2 and #4, I'd honestly say it depends on what kind of game you are looking to run. If you are going for a more military type game, or one where it is about persevering against all odds, then go with #2. Death being right around every corner helps those games, and if you are doing a military game you have a structure in place to help keeping the plot moving. If on the other hand you are going for a more 'story' approach, and you really want the starting characters to have a good chance of being there at the end, then do #4. Just, you need to make sure that you play it right, don't let the players know when they are safe and when they're not, and be reasonable with when you protect them from death. If someone walks into a room full of 60 angry enemies who want to kill him, you shouldn't protect the character, especially if it is knowingly doing this. It is important to hide from your players that in some circumstances you won't kill them, otherwise they won't take those situations seriously. If you think they've started to figure it out, then stop protecting them as much for a bit and let them learn the hard way as to what is going on.
I can't stress this last part enough, so I'm giving it its own section. If you have started doing a game under View #2 for player death, do not switch to a method where you protect people if it has already killed someone or multiple characters. At least, not without talking to the players of those characters, if not the whole game. Why? Because it's not fair that everyone else gets protected some times, while their starting characters were just killed in a random fight. The precedent here is important, and it can cause problems if you just switch.