RPG Evolution: What if I just kill him?

Image courtesy of Pixabay.​

We've heard the term before: "murderhobo," player characters that lead a nomadic existence defined by indiscriminate violence and a lack of social ties. These PCs are not merely a choice but a symptom of a deeper cognitive disconnect: the assumption that an NPC's authority is derived exclusively from their hit point total or combat statistics.

Where do murderhobos come from? In my weekly Dungeons & Dragons library game that's open to the public, I've discovered they're out there, and I have a better of idea why.

The Digital Legacy​

My encounters with murderhobos usually happen after the second game, when a player has gotten comfortable enough to ask. In-game, an authority figure tells them what to do (usually a sheriff or judge), and the player asks a simple question: "What if I just kill him?"

The murderhobo does not emerge in a vacuum. Though it's existed before video games were common, it surely nowadays comes from the digital sandbox. In most computer role-playing games (CRPGs), players are conditioned to interact with the world through a binary interface. An NPC can be classified as a Static Quest Giver, often rendered unkillable by the game’s code—the aforementioned plot armor—or an Enemy, whose sole purpose is to be reduced to zero hit points for experience and loot. This binary creates an interesting dynamic where players only respect what the software forces them to respect.

When these players move to the tabletop, they often test the limits of the Dungeon Master’s world by attempting to kill low-level authority figures, such as town guards or local magistrates. If the DM has not established the institutional weight of these characters, players assume that a weak stat block implies a lack of importance.

The Hit Point Hallucination​

A core tenet of the murderhobo's logic is the Hit Point Hallucination: the belief that the capacity to take damage is the ultimate measure of an NPC’s worth. Historically, hit points have always been a nebulous abstraction. Since the original rules, they have represented a combination of physical durability, luck, and the will to survive. As the rules have evolved, particularly in the 2024 revision, this abstraction has been clarified to include "stamina, resilience, and endurance".

When a player looks at a king with 10 hit points and a dragon with 500, they often conclude that the dragon is important and the king is not. This ignores the reality that hit points do not determine effectiveness in battle until they reach zero. A king with 1 HP can still command an army to fire a thousand arrows. The fallacy lies in treating hit points as meat points rather than a narrative resource.

The core of the murderhobo problem is the belief that the player is the only "real" person in the world. In a collaborative narrative, the NPC's reality is maintained by the DM and the shared imagination of the group. If the party kills a beloved NPC, they are not just "clearing a mob"; they are destroying a piece of the shared world that everyone has invested in.

But to be clear, there has to be a world in the first place. A player can't be blamed for threatening every guard when they've been eyeballs deep in dungeon monsters for weeks. Conversely, if the game is hack-and-slash only, this may well be acceptable (there's a visceral thrill in playing a chaotic, violent campaign, though it probably doesn't last long) and as long as that's the goal of the campaign and everyone agrees, being openly violent can be fun.

Most times though, it's one player who is new to the game or gets bored with it, reacting to the lack of stimulus or the frustration that their character can cast miracles, but they have to listen to this (surely low level) NPC. Depending on how resilient the DM's campaign world is, they might not be wrong to think they can get away with it too. When players know that the world reacts logically to their actions—that killing a shopkeeper leads to the closure of the shop, the loss of a supply line, and a permanent Hostile attitude from the merchant's guild—they begin to see the world as a persistent entity rather than a disposable playground.

The Authority of the Shared Dream​

The murderhobo mindset can be a mismatch of expectations for those who have not yet learned to trust the DM or the world. Authority in a TTRPG is not a number; it is a story. It is the collective agreement that the king’s decree matters, that the merchant’s life has value, and that the party’s actions have weight. Conversely, not killing everything should matter; this means rewarding non-combat solutions, making NPCs feel like living people with goals and fears, and using the 2024 Influence rules to show that Charisma is just as powerful as a +1 longsword.

"Can't I just kill him?"

Sure, I respond with a smile. "But then you have to deal with the consequences." And that's usually enough to make the player change their mind.

Your Turn: How do you deal with murderhobos?

 

[Maxperson]

One thing I love doing with D&D and everyone else with rare exceptions seems to hate is... I treat life as cheap. Or, rather, life has a set value, and that really seems to bother some people. Objectively, a human(oid) life is worth about 2500 GP; that's the going rate for a 5th level casting of raise dead according to RAW spellcasting services, plus the 500 GP diamond necessary.

What does that mean? It means, to me, killing doesn't have a moral cost as long as you're willing to spend the monetary cost. Back in Anglo-Saxon England there was the concept of "weregild", or "man-gold", the amount of money you'd need to pay to a family in order to make up for a murder. That's far more literal in D&D.

"If you kill him, you'll become just like him!" doesn't hold water if you're willing to put your money where your mouth is... or, alternatively, only the poors need to care about killing people (which... isn't that far off from IRL I suppose). Kill the BBEG and resurrect his corpse in a prison if you have to. Or bring him back to life, read his mind, see if he'd do it again, and then kill him again if he doesn't repent. We don't have the same material conditions in D&D as we do in reality; doesn't it stand to reason that we'd have vastly different morality as well?

This all assumes that the raising will work. There's no such guarantee since the soul can refuse to return. Every killing has a chance of being permanent, and therefore has a moral cost. You KNOW before you commit the murder that the person may not ever come back to life, even if you pay the 2500gp.

 

[Clint_L]

Are we talking about randomly murdering people, though; or are we talking more about simply a violence-first approach to solving in-game challenges?

Example: when trying to get past a gate guard at night in order to get into an enemy town is the party's usual go-to option kill him, bribe him, sneak past him, charm him, or ? The 'kill him' option is what I mean by violence-first, and it's not at all unusual particularly in D&D where so much of the game is centered on combat and violence.

Innocent guard? School game I veto it. It’s a violation of our PG rule. Home game I’d just make it clear that the player is choosing to be a murderer, and if there weren’t extraordinary circumstances (greater good) then I’d remind them that I don’t DM psychopathic characters.

That’s hypothetical, though; it’s hard to imagine a player wanting to play like that in my home game.

 

[Maxperson]

Absolutely this. When the PCs have access to abilities and powers on a completely different scale from the rest of society, they cease to be bound by its rules, barring personal choices and moral compasses. The more you flatten the differences between NPCs and PCs, the less appealing the Gordian Knot murderhobo solution becomes, because you've got real risk of consequences.

The less appealing the game is, too. I don't do the murder hobo thing, but I can say that I also don't do games where my character is barely stronger than the butcher, the baker, and the candlestick maker. I want to play characters that can go toe to toe with demons, giants, and dragons

 

[MarkB]

This all assumes that the raising will work. There's no such guarantee since the soul can refuse to return. Every killing has a chance of being permanent, and therefore has a moral cost. You KNOW before you commit the murder that the person may not ever come back to life, even if you pay the 2500gp.

Yeah, my house-rule for most games is that, outside of PCs, it's very rare to find a soul both able and motivated to give up the afterlife and return to life. Revivify always works as it essentially pulls the soul back before it's fully moved on, but anything else is very unlikely to be effective on an NPC.

 

[EzekielRaiden]

It can work. It just depends of what flavour of evil they chose.

In my experience, the vast, VAST majority of people who want to play an "Evil" character actually do want to play specifically Chaotic Evil gleeful murderers. It's exceedingly rare to see anyone wanting to play a true "Noble Demon" archetype, where they do evil things but also have a real (if twisted) honor code and genuine lines they'd never cross, even if it would be advantageous to do so.

Maybe that can be my litmus test for folks asking to play evil characters in the future. "Name three things your evil character would never do, even if it were advantageous for them to do so. Please don't use any of the things that are on my Absolutely Not Ever Happening list." (That would mostly be things like unforgivable actions against children, SA, actively going out to place other humans into slavery, etc.) If they cannot name at least three things that they wouldn't be willing to do, even if doing so would be advantageous, then I'm sorry, they've got a Chaotic Evil character, not a "noble" LE character nor a pragmatic NE character.

Like i tried to explain in previous posts between soldier and murderer archetype. Soldier archetype is Frank Castle (Punisher), Murderer archetype is Joker from Dark Knight movie.

Punisher is, if we use d&d alignment system, at best extreme end of LN (and that's "softest" version of him), but usually LE. Party of LE Punisher like characters can work very well in long time campaign. Run quite few of them and they are fun to the point i actually prefer them over classic hero types.

What doesn't work is pary of Jokers. They are evil for evil's sake, they are chaotic and they just wanna see your world burn. That one gets real boring real fast. Their only goal is to to wreak as much chaos and mayhem for the sake of chaos and mayhem. There is no story, no drama, no conflict. There is also no space for character growth. They are flat, static, and just there to F up game world you as a DM spent time creating.

I can definitely see where you're coming from on this. My problem is the sheer number of people who pretend--perhaps even to themselves!--that they are playing The Punisher, but actually they're playing The Joker or even worse, just random acts without even a cunning, adaptive pseudo-plan behind them.

Wish-fulfillment, when coupled with an explicit imprimatur of "evil", has a tendency to warp behavior into something grotesque, unless the player finds such behavior actively distasteful. I'm one such player. I don't have it in me to play evil characters as PCs. I've tried. I always fail, usually sooner rather than later. I can only get away with it as a GM because I play a zillion characters and thus only spend a few minutes at a time "in their head" so to speak. It's a 30-second blast of freezing cold water, not an all-day, every-day arctic chill.

 

[Ralif Redhammer]

It can work. It just depends of what flavour of evil they chose.

Like i tried to explain in previous posts between soldier and murderer archetype. Soldier archetype is Frank Castle (Punisher), Murderer archetype is Joker from Dark Knight movie.

Punisher is, if we use d&d alignment system, at best extreme end of LN (and that's "softest" version of him), but usually LE. Party of LE Punisher like characters can work very well in long time campaign. Run quite few of them and they are fun to the point i actually prefer them over classic hero types.

What doesn't work is pary of Jokers. They are evil for evil's sake, they are chaotic and they just wanna see your world burn. That one gets real boring real fast. Their only goal is to to wreak as much chaos and mayhem for the sake of chaos and mayhem. There is no story, no drama, no conflict. There is also no space for character growth. They are flat, static, and just there to F up game world you as a DM spent time creating.

There are absolutely ways to do evil characters that are interesting, and work with a group. But even a Punisher-sort can create conflict when they decide to do something like executing the villain that the rest of the party planned on turning over to justice.

Thinking about that, maybe it's not so much the flavor of evil but whether the player takes into account the rest of the party when playing their character. Heck, even a super-LG character can create trouble at the table when they don't do that. The last LE character I played was greedy and a bit of coward. But he always went along with the group's decisions.

The less appealing the game is, too. I don't do the murder hobo thing, but I can say that I also don't do games where my character is barely stronger than the butcher, the baker, and the candlestick maker. I want to play characters that can go toe to toe with demons, giants, and dragons

DCC RPG has you starting out as the butcher, the baker, and the candlestick maker, and it's a hoot. But you don't stay there (provided you survive the funnel). I like games where, for the most part, you have something to fear from that group of town guards. Where you have to worry about consequences for your actions within the world and its societies. Even when you're able to fight dragons, there should still be some risk to the lesser threats.

In my experience, the vast, VAST majority of people who want to play an "Evil" character actually do want to play specifically Chaotic Evil gleeful murderers. It's exceedingly rare to see anyone wanting to play a true "Noble Demon" archetype, where they do evil things but also have a real (if twisted) honor code and genuine lines they'd never cross, even if it would be advantageous to do so.

Maybe that can be my litmus test for folks asking to play evil characters in the future. "Name three things your evil character would never do, even if it were advantageous for them to do so. Please don't use any of the things that are on my Absolutely Not Ever Happening list." (That would mostly be things like unforgivable actions against children, SA, actively going out to place other humans into slavery, etc.) If they cannot name at least three things that they wouldn't be willing to do, even if doing so would be advantageous, then I'm sorry, they've got a Chaotic Evil character, not a "noble" LE character nor a pragmatic NE character.


I can definitely see where you're coming from on this. My problem is the sheer number of people who pretend--perhaps even to themselves!--that they are playing The Punisher, but actually they're playing The Joker or even worse, just random acts without even a cunning, adaptive pseudo-plan behind them.

Wish-fulfillment, when coupled with an explicit imprimatur of "evil", has a tendency to warp behavior into something grotesque, unless the player finds such behavior actively distasteful. I'm one such player. I don't have it in me to play evil characters as PCs. I've tried. I always fail, usually sooner rather than later. I can only get away with it as a GM because I play a zillion characters and thus only spend a few minutes at a time "in their head" so to speak. It's a 30-second blast of freezing cold water, not an all-day, every-day arctic chill.

In my experience, yeah, to continue the theme, much of the time with evil characters, you get a Homelander.

 

[Maxperson]

Basically if a task doesn't require violence then you see the difference. Violence first party isn't going to randomly kill npcs on their way to deliver a message across a friendly town. The GTA style group probably would or atleast perform many more criminal activities such as grand theft horse. Maybe another way to say this is that one group uses violence to achieve an objective ,and the other uses violence to have fun.

The Venn diagrams between the two do overlap, you won't get an argument from me here, but there are differences.

Violence first parties stem from DMs having no consequences for all the murders and violence they do. Natural consequences to tons of violence tends to curb that strategy.

 

[Clint_L]

This discussion is, once again, why I hate the alignment system. Instead of worrying about whether a character is "evil" or not (whatever that means), I urge players to focus on building a believable character. When they resort to violence, what motivates them? When do they see it as justified? It's never justified "because I'm EVILLLLLL BWAHAHA!" Even Homelander doesn't do things to be evil; he's justified in his own eyes.

Randomly killing people for no good reasons seldom makes any kind of character sense. It's not something that normal people do. It's not even something that normal psychopaths do.

Incidentally, the one player in my home game who likes using alignment plays a paladin (because of course) and is by far the most inflexible when it comes to party decisions, but I like it because the way he sometimes throws a spanner in the works makes for good story and he's not doing it for the lols. And alignment does kinda work for someone who has an inflexible code.

 

[payn]

Randomly killing people for no good reasons seldom makes any kind of character sense. It's not something that normal people do. It's not even something that normal psychopaths do.

I think this occurs when the player just wants to engage the game portion of the RPG. They want stuff and are not afraid to be a random psychopath to get it for the game mechanic's sake. The classic kill a shopkeep for his stuff so you dont have to pay for it routine. Additionally, it might just be for the lolz. The player is bored and just wants to see what happens when the world burns. They are not invested in making a believable character, or engaging in a believable world.

 

[tetrasodium]

Violence first parties stem from DMs having no consequences for all the murders and violence they do. Natural consequences to tons of violence tends to curb that strategy.

You've misdirected that blame

Spoiler: System design plays a big part

The system undermining its own dm's efforts to have consequences doesn't stop there though, the pcs don't even need a shelter where they can rest up and recover like a personal keep or an inn because a ditch near a tree is squarely within what the design requires.

The only consequence of weight the gm has in their quiver that can be directed at a hypothetical NPC slaying murderhobo like the thread describes us gm fiat & that tends to reflect negativity on the gm when used to such a degree