I have considered writing a MUD for years, and I have a lot of ideas my friends think are really cool (and that's how I'd hope to get anywhere -- word of mouth).

Thing is, there's one thing I have always wanted, that my friends and strangers hated: permanent death. Now, the emotional response I get to this is visceral revulsion, every time. I'm pretty sure I am the only person that wants this, or if I'm not, I'm a tiny minority.

Now, the reason I want it is because I want the actions of the players to matter. Unlike a lot of other MUDs, which have a set of static city-states and social institutions etc, I want the things my players do, should I get any, to actually change the situation. And that includes killing people.

If you kill someone, you didn't send them to time out, you killed them. What happens when you kill people? They go away. They don't come back in half an hour to smack talk you some more. They're gone. Forever.

By making death non-permanent, you make death not matter. It would be similar if a climax to a character's arc is getting a speeding ticket. It cheapens it. Non-permanent death cheapens death.

How can I: 1) Convince my players (and random people!) that this is actually a good idea?, or

2) Find some other way to make death and violence matter as much as it does in real life (except within the game, of course) sans character deletion? What alternatives are there out there?

  • 4
    \$\begingroup\$ The problem with permadeath is not only do your actions matter, but the actions of the guy who drove his car into a nearby power line also matter, as well as the actions of your ISP as they make mistakes upgrading a line causing an outage. Overall I like the second question. How do we make death matter in games without permadeath? \$\endgroup\$ Commented Jun 10, 2014 at 21:58
  • \$\begingroup\$ It doesn't really matter that this is a MUD, except potentially if you allow other players too much influence over whether or not a single player lives or not; have you played State of Decay? It deals with permadeath in an interesting fashion. \$\endgroup\$
    – user1430
    Commented Jun 10, 2014 at 21:58
  • \$\begingroup\$ Related: gamedev.stackexchange.com/questions/43801/… and gamedev.stackexchange.com/questions/14918/… \$\endgroup\$
    – user1430
    Commented Jun 10, 2014 at 21:59
  • 2
    \$\begingroup\$ It has been done, in fact one of the games I worked on had permanent death in a subscription-based game (Underlight). One extremely high-level player could permanently kill (Dream strike) another by sacrificing a massive portion of their stats if the other player's soul was "exposed". To my knowledge, it did actually happen one time. This was not something that an NPC could do (in fact, that game had no NPCs) - so it was all heavily RP'd by actual players. Kind of makes XP loss / corpse recovery on death look silly by comparison, and even that is considered too harsh by most MMOs these days. \$\endgroup\$ Commented Jun 11, 2014 at 2:28
  • \$\begingroup\$ In the early stages of World Of Warcraft, the in-world PVP had Honor Kills, and Dishonerable Kills. What about a system where NPC's, loot, luck, whatever is affected by Dishonerable kills? \$\endgroup\$ Commented Jun 11, 2014 at 16:37

7 Answers 7


When you want perma-death to matter as much for the victim as it matters in real-life, you also need equivalent consequences for the killer.

In just about any real-life society (present or historical), murder is the crime which receives the highest punishment that society is willing to administer. In any judical system of today, a murderer is either executed themself or imprisoned for the rest of their life. That means the consequence of intentionally taking someones life practically means losing your own life. Why is this? Because no society which tolerates murder can survive in the long-term. And so won't your game.

When you want murder to matter as much in your game as it does in real-life, the consequences for the murderer need to be equivalent. When a player destroys the progress of another player, they must be aware that as a consequence the same thing will likely happen to their own character.

You could model this by permanently turning the murderer into an outlaw. NPCs will try to capture and execute them (permanently, of course) and other players might be made aware of a murderer and be allowed to perma-kill them themself without suffering the same consequences (and maybe even receive a reward).

Make sure that there is no easy and reliable way to permanently escape prosecution. A player might be able to avoid the consequences temporarily, but when you add any loophole which allows them to permanently wipe their crime-record, you will encounter players who abuse this method.

Example: A player might be able to avoid criminal prosecution as long as they are a personal friend of the local figure of authority. But when that person changes or decides to no longer protect them, their amnesty is revoked and they can be prosecuted again.

  • 1
    \$\begingroup\$ This. If there is a real reason not to kill someone, then you're less likely to try in the first place. Thanks, Phillipp. \$\endgroup\$
    – Luke
    Commented Jun 11, 2014 at 18:38
  • 1
    \$\begingroup\$ If you made the consequence of murder equivalent to peramdeath (if I kill you, I get executed, or you weight the game so heavily against me I may as well be dead) then death has no point... you'll end up with people not killing each other because of the consequences. So they'd not bother fighting each other... So you'd end up with a game like Animal Crossing where death simply doesn't apply. \$\endgroup\$
    – Piku
    Commented Jun 14, 2014 at 9:32

I can't believe no one has mentioned successful MMOs centered around PVP permadeath: DayZ and Eve Online (or almost-permadeath in Eve's case). These are games that reward player killing, and have the entire gameplay experience built around that ability. If you are familiar with these games, you can stop reading and use them as case studies on how a permadeath MUD can work. I want to offer these as a counterpoint to the prevailing attitudes in the other answers, which seem to assume that player killing is a bad thing, even going so far as to suggest punishing the killer.

Here's a short description of the two games:

  • DayZ is an MMO FPS about individuals surviving a zombie apocalypse. Players must scavenge for food and weapons, can work together but also kill each other to steal supplies - there are no restrictions on what players can do to each other. Players are paranoid and terrified of each other, and there are plenty of anecdotes of slavery, hostage-taking and of course, killing for fun.
  • Eve Online is a space combat MMO where there is PVP and death can result in the loss of everything apart from character stats, but in a game where one must purchase ships at huge cost, the character means little. There is an NPC law-enforcement faction that attacks player killers which creates de-facto safe areas, but most gameplay takes place in lawless areas where piracy and open warfare between player factions are the norm. The huge range of ships and combined arms warfare encourages the formation of large player "corporations", but there are also many stories of political intrigue, shifting alliances, scams and espionage.

Giving players the ability to cause massive harm to each other leads to polarising opinions, and creates a large psychological factor to the gameplay. Permadeath alone makes the game take on a whole new tone, since even trivial choices (such as whether to quaff a potion) can have drastic consequences, so the game becomes more cerebral. Adding player killing to that mix makes things even more severe: not only do you have to weigh the benefits and costs of killing another player, you must also consider whether she considers killing you worthwhile, or even worse: whether she just wants to kill you for fun.

Another question to consider is: why is it that DayZ feels like the wild west, whereas Eve Online is like warring nations? In these games where there is so much player freedom (including the freedom to kill each other), tiny changes in gameplay have huge consequences. Things like whether players can kill each other in a split second, whether the combat favours small or big groups, whether there are safe areas, these all have a big impact on how players treat each other - suspicion, paranoia, the size and stability of groups. More learned folks than I have studied such phenomena, so I won't dwell here: find more reading online if you are interested. Bottom line is, PVP permadeath has huge implications for the game, and your level of control as a game designer; more often than not you'll feel like running a social experiment than crafting a world.

So to finally answer your questions:

  1. How can I convince people that PVP permadeath is good?

    You can start by pointing to examples of very successful games that have them. Keep in mind that they're not everyone's cup of tea, and different people have different thresholds for what they deem fair or acceptable. The Bartle Test is one way of looking at this: PVP permadeath greatly favours the Killers, which is necessarily at the expense of others (mostly Achievers), and different people like different things - not everyone is a Killer. A lot of people also won't find permadeath from other players acceptable or fair, but that's simply because people have different thresholds for fairness. One player's "unfair" is another's "that's just part of the game", whether that's an inconveniently-placed obstacle, a ruthless RNG, or player killing.

  2. How can I find some other way to make death and violence matter as much as it does in real life (without permadeath)?

    There's a range of punishments you can mete out, up to and including permadeath. Any kind of loss (XP, property or time) can be drastic and hence meaningful. Unfortunately the "death matters" that you want requires punishing the player, since only the fear of death (or great loss) can produce that kind of trepidation in normal gameplay, and the power-tripping satisfaction of having caused loss to another player characteristic of Killer types. Simply rewarding killers without punishing the killed won't produce these results. In the same sense, I don't see how punishing killers can be productive, since it removes the utilitarian motivation for killing and only leaves the purely sadistic.

    Finally I want to point out the futility in making a game "like real life". Real life has permadeath, but it also has no restarting. That is what really makes life and death matter in real life.

  • 1
    \$\begingroup\$ I wonder if you could create a game with no restarting. It'd need to draw the player in with a long "acclimatisation" period so they amass some sort of value they want to protect (which could simply be hours of time put into the game) before allowing death to happen. Since the reason permadeath and no restarts matters so much in real life is because nobody chooses to be born in the first place \$\endgroup\$
    – Piku
    Commented Jun 14, 2014 at 9:38
  • 1
    \$\begingroup\$ I have played Eve Online quite a bit, and I didn't consider that permadeath, so I didn't use it as an idea on my answer. That said, it does have a good system for player killing. The fact that the skills train in real time, and even they can be partially lost during death along with ships and implants makes it feel close to a permadeath game. The concept of 'safer' zones affords the sense of acclimatisation in Eve. If you die, and are low on funds, stay in safe areas for a little while before going in to dangerous areas again. (Not only possibility, but it is one way people play.) \$\endgroup\$
    – Aviose
    Commented Jun 17, 2014 at 15:12

The question is what kind of game you want to make. MMORPGs typically depend heavily on grinding a single character for weeks, making progressing the only gameplay element of it. In this case killing a character is nothing but a punch into the guts, there is no value in it, it just needlessly punishes players.

Well, first of all you could remove most of the grind (stat/gear wise) and make any unlocks account-wide (game changing & visual stuff), this would make perma-death much less punishing as you could just make a new character and reach the same level again quickly.

Further you could make death much more common and randomizing the content (for example picking a random spawn position every time), this would put much more emphasis on survival and make it feel a bit more like a roguelike, which is genre permadeath worked pretty very well for in the past.

Note that player behavior is heavily influenced by gameplay, if you don't want players to try to kill other players make it a unrewarding experience, or give bonuses for helping strangers and penalties for betraying them.


The inherent problem here is that your character's life is no longer in your hands. Gamers can be very competitive, and it only takes one person that decides to be a royal d*** to ruin the experience for everyone. Because of this, part of the question should be, "How do I ensure that players don't run wild perma - killing everyone they can?" Because of this, I would probably resolve this such that if a player beats another in a fight, they can take a non-trivial penalty to perma-kill the loser. This keeps the power in the hands of the strong, while preventing it from going rampant. It also helps to subtly push your players towards acting with respect towards other players, especially if the cost is relatively high.

With the prevalence of player killers in games regardless of the trivialities of death, many are turned off of perma-death because it is difficult to regulate. Penalties for doing it turn it into a self regulating system.

The only other feasible possibility would require a proportionally large selection of moderators, and that would require too much maintenance.


The key to a good permadeath system is to make the player feel in control, and not feel complete sense of loss when permadeath occurs.

Consider the analogy of a well-designed platformer. A well-designed platformer should not be trivial to complete: there should be some tricky jumps that require some skill in there, to get the player's adrenaline going. But when you fail a tricky jump, you should not feel like the game was unfair, you should feel like the fault was yours and be able to see it. And a well-designed platformer starts you over on your next life far enough away from the failure point that you need to work to get back there, but not so far that you have to repeat everything up to that point.

Preserving those sensibilities can scale to permanent death as well. Tuning this is pretty difficult, of course (especially when you start to have other players involved, and those other players can potentially contribute to the death of any other player characters). Generally, this means:

  • Designing the game system so the player has ample ability to avoid death through skillful play. Undetectable instant-kill traps are, consequently, not a very good idea. Similarly you want players to be able to learn from any failures to stay alive, so that less-skilled players can build up the requisite skill to survive such an encounter in the future. Make sure there are good "escape hatch" skills that every player has access to that can get them out of tricky situations, perhaps with appropriate foresight.

  • Provide additional venues of progression independent of individual characters that continue even if an individual character dies. Roguelike games tend to do this with their "inherited monster memory" features, and you do similar things. Perhaps new characters get to inherit selected gear from the old character, or selected personality traits or skills. Allow the recruitment of followers who can assume the role of your "main" character when your previous character dies.

Multiplayer contexts make it harder to build these kinds of systems (especially the first one). However, you can simply make the decision that players cannot directly engage in combat with others, if you want. This solves a large class of player-versus-player permadeath problems related to griefing (high-skill, high-level characters preying on the weak neophytes).

If you want to allow player-versus-player direct combat, make sure there is a significant cost and consequence to killing another player. The balance here should generally be weighted heavily towards the player who doesn't want to go around killing other players, so any player who does so is significantly disadvantaged.

(Generally, I feel like the idea of combining permadeath and player-versus-player combat is usually just a recipe for disaster, since it only enables "fun" for the griefers; co-operative play tends to be much more rewarding in my experience -- especially with permanent death. But that's just my opinion.)


Recovering a character's body from a state of death might leave permanent damage to its physiology / neurology:

  • XP/memory Loss (Diablo)
  • Loss of physical integrity (D&D; constitution loss on Raise Dead)
  • Loss of skills, intellect etc.

...I find the increased chances of raising a drooling idiot from the dead to be quite attractive as a gameplay dynamic, if not as a mental image. Not only does it make death count, it also has an element of puerile humour to it, which a MUD could tap into. What more could you ask?


There are some great points above, but I'm not seeing much along the lines of making a separation between player and characters in order to allow for deaths that can hurt but not necessarily unroll days/weeks/years of play.

For instance, let's say your game was about a zombie apocalypse. Instead of the player being their character, they could hire characters to fight zombies for their company. Success brings money and prestige for the company, allowing access to better weapons, more qualified mercs, etc. You could balance the situation in a way such that a veteran zombie slayer is somebody you do not want to lose - he's leveled up, he has unique skills that only come from surviving tough fights, etc. But maybe it only takes a short time (real-life) to get a rookie to that state.

This separation combined with the pain of death also affects other options available to players. For instance, your veteran is bitten during a rough battle. Do you chop of an arm hoping to avoid the infection? Burn him and start fresh? Spend a huge amount of your company's XP/prestige/money/whatever to get a cure? Just send him into a hopeless battle hoping for a blaze of glory that reflects well on your company? Etc.

And you get multiple factors contributing to the pain of death:

  • The progression from rookie to veteran is how you decide the value of keeping mercenaries alive. If you can go into two battles and have a kick-ass warrior, death isn't a very big deal. If it takes six months to max out, on the other hand, you have a pretty significant loss on a death.
  • The company rewards that are separate from the mercenary help temper the losses when death does inevitably happen. If successes increase a ton of company stats (money, prestige, maybe XP unlocking specially skills for future mercs, etc), death is less important. If successes only help a tiny bit, death is more important.

Balancing these two factors can be a living, ever-changing goal as you tweak rules to maximize enjoyment without necessarily sacrificing fear of death. If you want hard-core death rules, maybe it does take a player six months to get a single mercenary to an uber-high level. But the rewards to the company could be set up such that a high-level merc is providing tons of tertiary bonuses so that when he's dead, leveling up the next merc is only a five- or four-month process. Or maybe you have unlocked a new, higher level cap.

As an aside, I'm a huge fan of perma-death games like Nethack, but I always felt like it would be much better if each run-through left me with something extra based on my achievements. For instance, killing 100 orcs in a single game could mean unlocking a permanent +10% damage boost to orcs.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .