This is effectively the reverse of How can I end combat quickly when the outcome is inevitable?

Lame duck scenarios in game design occur when a player cannot win, but the game isn't over yet. They're obviously undesirable, and good game design would seek to minimize it.

I'm interested to see how role-playing games avoid these scenarios, because they seem especially vulnerable to it. After all, even if the odds are against the players are insurmountable, it can take several rounds before the player's HP actually reaches zero, and they still have to "decide" what to do in those rounds (inverted commas because they're actually dead, they just haven't died yet).

Examples of the scenarios I'm thinking of:

  • The scenario given in this question. After the player attempts to shoot the men on horseback, it becomes a 5-on-1 fight. The DM still said "roll dice", but that seems pointless, since realistically the player's not making it out alive.
  • Five players attempt to attack five monsters, which does not sound like a stupid thing to do, but the players roll an ungodly amount of misses while the monsters roll an ungodly amount of critical hits, and kill three players.

Basically, anything which in a video game would make the player bring up the "load game" button.

How do role-playing games handle and/or avoid these situations? Some guesses:

  • DM decides to go easy on the players. After the critical hits kill three players the DM encourages the survivors to run, and then lets them run without pursuing (even if it makes zero sense story-wise). (Sounds contrived)
  • DM decides to directly influence the dice. "Wait a moment this is horribly unlucky for you, we'll pretend this dice actually rolled __ and you survive." (Sounds even more contrived)
  • DM says "derp, I guess you guys are all dead, next game?" (This might be beyond the realm of reason)
  • DM engineers a real "load game" button, restarting the fight and effectively bringing the three slain players back to life. (But if this is permissible, wouldn't the quest always succeed?)
  • 28
    \$\begingroup\$ I haven't encountered any RPG system with "always moving to the right" fight. All of them allowed myriad actions, like surrender, flee, hide, suicide, diversion and so on. Could you tell us which game systems gave you impression that "continue fighting, being already dead" is a thing? \$\endgroup\$
    – Mołot
    Commented Jan 29, 2020 at 7:53
  • 3
    \$\begingroup\$ This sounds more like a Kobayashi Maru scenario, than a Lame Duck one. \$\endgroup\$
    – aaron9eee
    Commented Jan 29, 2020 at 8:13
  • 7
    \$\begingroup\$ "DM decides to directly influence the dice." - that is, I believe, fairly common practice. The DM tends to roll behind a screen of some kind, preventing the players from seeing the results. This prevents players being able to use direct numbers to determine monster stats, and allows the DM to fudge the numbers one way or the other to keep challenges interesting \$\endgroup\$ Commented Jan 29, 2020 at 15:33
  • 4
    \$\begingroup\$ Uhhhm @Allure YOUR OWN LINK regarding what is a Lame duck scenario states that none of your examples are... \$\endgroup\$
    – Hobbamok
    Commented Jan 30, 2020 at 13:22
  • \$\begingroup\$ @LogicianWithAHat I also think that not letting players know you're doing it is the whole reason it works. I was baffled when the question suggested you should announce a dice fudge. My DM doesn't ever let on when he's going easy dice-wise, I can't imagine why he would. \$\endgroup\$ Commented Jan 31, 2020 at 15:58

16 Answers 16


Beware of mislabeling situations as lame duck

In most games players have a vast array of options available to them. Generally speaking it is best to let players try their darnedest to survive.

After the player attempts to shoot the men on horseback, it becomes a 5-on-1 fight. The DM still said "roll dice", but that seems pointless, since realistically the player's not making it out alive.

Well, what can the players do now? Surrender with their hands up? Plead for their lives? Offer up their money and items?

Five players attempt to attack five monsters, which does not sound like a stupid thing to do, but the players roll an ungodly amount of misses while the monsters roll an ungodly amount of critical hits, and kill three players.

What can the remaining 2 players do now? Try to run from the monsters? Can they revive their friends? Can they engineer a clever escape, then ask the local church to dispatch paladins and clerics to recover the corpses of their friends?

That isn't to say that the players necessarily succeed in their attempts, or that you use GM magic to make the enemies allow them to succeed. Failure is valid roleplay.

In both situations there is another roleplaying option, which seems to be assumed to be the default: fight until you die. That is a valid choice, and if a player chooses to do that then you should let them.

In a real lame duck situation...

The 50ft tall magic immune archdemon has the wizard dangling by the foot over a vast lake of infernal lava. The wizard's death is guaranteed.

So the wizard dies. What else can you say? Play out the death a little, give them a couple of chances to try their spells or wiggle like a worm, let them feel powerless because they are, their death is assured.

But remember, this situation happened because of a long string of meaningful choices. In the last 2 seconds, yes it's a lame duck situation. But there were numerous, perhaps even countless, choices that players made which lead to this situation. Perhaps 1 second earlier it wasn't a lame duck situation.

Sometimes PCs die, that's valid roleplay

What you do after is up to your table. At my table the other players can try to resurrect the character, and/or the player could move on to playing a new character. Even if the entire party is killed, perhaps they wake up in a makeshift hospital in the local church, they were rescued and cared for but now they are in debt.

Fudging rolls or rewinding the game removes player autonomy, and consequences

It seems to me that the lame duck mentality you are discussing stems from the idea of "winning" or "losing" at a roleplaying game. If you shift your mentality a little to "having a fun game" and "participating in an interesting story", then suddenly having the whole party die after falling into lava might not be such a bad conclusion.

In order to have these experiences it is important to empower players. If you fudge rolls or rewind games, you are telling your players "I have a story that I am going to tell you". Every time you do this your players become less like "players" and more like "pawns".

Instead, if you empower them, you shift the responsibility for the entire game off your shoulders, and everyone can share the burden. If the party falls into lava, then it's everyone's fault.

  • 19
    \$\begingroup\$ "I would never fudge rolls or rewind the game because to me that removes player autonomy and consequences" -- this seems worthy of being the headline here. Of four possible approaches the question envisions, three are essentially "fudge rolls" in one form or another, removing consequences for player actions, and thus are all undesirable ways to deal with the situation. I mean, I suppose each DM has the right to choose to do that or not, but I think there are clear objective reasons why doing so is counter-productive. \$\endgroup\$ Commented Jan 29, 2020 at 18:04
  • 6
    \$\begingroup\$ some games have PC death as a point but don't stop the flow with it. Like Paranoia... \$\endgroup\$
    – Trish
    Commented Jan 29, 2020 at 18:30
  • 5
    \$\begingroup\$ I particularly liked the "surrender" idea. After all, the NPC is not just a computer generated bunch of pixels which has been programmed to kill. A good DM should be able to see, if he can (and wants to) fit a scenario where players are defeated, but not necessarily outright killed. Just because players are slaughtering monsters doesn't mean their opponent NPCs can't be made interestingly humane (or have ulterior motives apart from killing). \$\endgroup\$
    – Gnudiff
    Commented Jan 30, 2020 at 11:39
  • 2
    \$\begingroup\$ "Sometimes PCs die, that's valid roleplay" - and depending on the system (and level of players), death may not even be a significant impediment to progress! \$\endgroup\$
    – Michael W.
    Commented Jan 30, 2020 at 17:01
  • 3
    \$\begingroup\$ @Falco: if you're in a situation where you would fudge the roll, just don't roll. If the only "interesting" outcome left is for the players to successfully persuade attackers to accept their surrender, then just rule that the surrender has succeeded. Use die rolls only when there's a meaningful chance of both success and failure, and when you want to let that chance play out. \$\endgroup\$ Commented Jan 31, 2020 at 17:33

None of these scenarios are Lame Duck scenarios

In order for a scenario to be a Lame Duck scenario, it must be impossible, not just improbable, that you win. A fight in D&D 5th edition with five mounted combatants against your first-level rogue might seem impossible, but it's not. If you crit enough and they miss enough, you'll come out victorious. Or if you do something clever. Or if they do something particularly stupid.

That said, sufficiently bad odds in a game with a clear and immutable goal can still be usefully treated as a Lame Duck scenario, even if it technically isn't one.

Lame Duck scenarios in RPGs can only occur where there is a clearly defined immutable goal

Most of the time in RPGs, the players' goals are not explicitly stated. Furthermore, most of the time, these goals are subject to change as circumstances evolve. "We are gonna get to and loot the treasure vault" might be impossible once it's clear that the party has accidentally bored into a lava vent and there's no way to get to the treasure before the lava flows in and wrecks everything, but "We are gonna escape the rising lava with our lives" works perfectly fine as a back up.

Nor does the inevitability of death ensure a Lame Duck scenario! In most RPG settings, everyone dies sooner or later, even the immortals if there are any and you wait long enough. Nevertheless, people find meaning in their lives. Deciding what your character does when they are eminently going to die in the immediate future and there is nothing they can do about it has been a very fruitful artistic tool in the past.

How to deal with Lame Duck scenarios

That said, you can still have Lame Duck scenarios in RPGs.

Two-sided games

If the RPG is two-sided, like the typical game of adversarial D&D, it is useful to just wrap things up rather than actually playing it out, similar to how one might resign in chess. If all the players agree, they should generally be allowed to just say "ok, we lose" or whatever and do whatever wrapping up is needed from there. Then the game can continue from the failure state, if that's a thing, or if not then new characters can be made and the game can begin again, or the players might go do something else.

Similarly, in two-sided games with a DM, the DM should be able to go "ok, you win" and wrap things up from there rather than playing things out, continuing from the success state, if that's a thing, or if not then the same as when the game's ended any other way.

Just like in traditional games, a two-sided game's Lame Duck scenarios are basically a trivial case, easily solved by the convention of resignations.

Many-sided games

There are also many-sided RPGs with Lame Duck scenarios possible. For example, this very much includes early D&D -- Warhammer and Chainmail and other such table-top wargames are often played with many sides in a single game (and indeed this was almost always done with early D&D/wargame crossover games) and almost always get to a point where it is literally impossible for one or more of the sides to accomplish their victory objectives without the game being over (for example, if they lack an appropriate type of unit to hold a King of the Hill type point thingy).

In these games, there's no easy solution to Lame Duck situations. People typically just suck it up, make up a new objective to do while they lose, and then do that. If they have a grudge from their loss and the ability to do so, they might try to play kingmaker.

With particularly immature, bad-sport players, it is sometimes necessary to call off the game at this point because they throw a fit or refuse to keep playing or otherwise make continuing more trouble than its worth. In those cases a modified version of resignation, where the bad sport calls off the game and then everybody else talks out what would probably have happened, seems to be preferred. Indeed, several groups I've been in have had this as the expected outcome before play began, to cater to certain participants' needs.

  • 4
    \$\begingroup\$ I wouldn’t say there is any such thing as “the typical game of adversarial D&D,” since D&D isn’t typically adversarial. The DM controls the PC’s adversaries, true, but the DM controls everything outside of the PCs. And it’s neither recommended nor typical for the DM themselves to be opposed to the PCs. Rather it seems more like you’re commenting on the particular stories that D&D often tells, and is most well-suited to telling: quests and adventures, where the PCs are agents of action who set out to accomplish some goal. But that’s not the same as “two-sided” or “adversarial.” \$\endgroup\$
    – KRyan
    Commented Jan 29, 2020 at 14:40
  • 1
    \$\begingroup\$ @KRyan I don't mean that the typical game of D&D is adversarial. I mean that the typical game of adversarial D&D is two-sided. I don't think a Lame Duck scenario makes as much sense in a non-adversarial game of D&D, since there's not really, in my opinion at least, a cogent enough definition of 'winning' or 'losing' there. \$\endgroup\$ Commented Jan 29, 2020 at 18:12

With a good GM, roleplaying games if anything are more robust against lame duck scenarios than board games. That's because unlike board games, in which the rules of the game are the only course of resolution possible, in a role playing game there is always the option to change paradigm and find a solution using a different approach. Fighting doesn't work? You can try to talk, or maybe run and hide.

A good GM also realises when the players start getting stuck in one approach which is fruitless, and will communicate that to the players out of character to find what solution would work for them.

As for the specific situation you mention of a fight against insurmountable odds, you mention that "Letting people run" might make zero sense. I will turn that around and ask you: in what situation does chasing after fleeing foes make real-life sense?

If you wanted to chase people off, congratulations, you did that. If you wanted to get their money. Congratulations, you have several corpses to loot already. Why chase after the others and risk your precious hit points? If you are a monster looking for food, congratulations, there is some right in front of you now. Why chaser after the other food still fighting back and risking that already dead food getting stolen by scavengers? Looking for a McGuffin and a fleeing player holds it? Sure, you will give chase, but why keep wasting your precious hit points on killing the holder when you can intimidate them into giving it up instead?

Doing any kind of fudging or retconning as a GM hurts the play long-term because it breaks the connection of actions and consequences. There is no reason to fear death if you know the GM will rather reboot the world than let you die.

The chance for a Total Party Kill (or even just a single PC kill) needs to be present for the players to make meaningful decisions, both in and out of combat. The choice to keep fighting in a hopeless situation and its consequences are the responsibility of the players, as long as the GM plays fair. This doesn't mean giving people only balanced encounters. It means giving them a fair warning that there is peril ahead, and giving them a fair chance to avoid it by making it clear, both in and out of character, that not fighting to the death is a viable option.

  • \$\begingroup\$ In what situation does chasing after fleeing foes make real-life sense real life example: you're one of the Navy SEALs who stormed Osama bin Laden's compound. You find him, but he attempts to flee. Do you give chase? I don't know how realistic this example is in an RPG, but it certainly seems believable that there'll be times it doesn't make sense to not give chase. \$\endgroup\$
    – Allure
    Commented Jan 29, 2020 at 8:48
  • 11
    \$\begingroup\$ @Allure If a PC has worked themselves up to the notoriety of Osama bin Laden and gets targeted by assassination squads like that, then running away not being an option is a fair consequence. Also note: The SEALs were after bin Laden. Any of his bodyguards running away would not have been chased. \$\endgroup\$ Commented Jan 29, 2020 at 8:59
  • \$\begingroup\$ @AlienAtSystem what if, e.g., the SEALs were trying to operate covertly and the bodyguards were going to raise the alarm? Or perhaps the enemies are just genocidal (e.g. Nazis dealing with escaping Jews), or frustrated (e.g. if they'd been getting harassed by guerrillas for months, they might chase a guerrilla who's trying to escape). \$\endgroup\$
    – Allure
    Commented Feb 1, 2020 at 13:43
  • \$\begingroup\$ @Allure You are a soldier, told by your superiors to kill some civilians. They manage to flee into a forest as you arrive at their home. It's raining, it's night, you've never been in that forest before, you're just two people and neither of you brought a compass. Do you give chase? Because that's what the Nazi scenario would look like on the ground. As for the other examples, those would still be fair peril, and you're overestimating how eager people are to get themselves killed for a cause. \$\endgroup\$ Commented Feb 1, 2020 at 13:57
  • \$\begingroup\$ @AlienAtSystem I'd probably still chase them until they get to the forest, which could be some distance away, especially if I'm armed and they're not (and they're running from me, i.e. I know I can beat them). It's also possible the enemies are not rational from the human sense of the word, e.g. the Others in Game of Thrones. There are always more possibilities, e.g. you are supposed to be guarding something and the PCs stole it, do you give chase? \$\endgroup\$
    – Allure
    Commented Feb 1, 2020 at 20:44

How do role-playing games handle and/or avoid these situations?

You don't necessarily avoid them. Especially in traditional, old school play.

DM says "derp, I guess you guys are all dead, next game?" (This might be beyond the realm of reason)

This parenthetical assessment is incorrect. In fact, there is a widely recognized acronym, TPK, for "Total Party Kill", to describe the fairly common situation in which everyone in the party is indeed dead.

Now, some DMs do fudge dice or results, or change the monster count or strength on the fly, or decide by fiat for monsters to run away or take prisoners, or something like that.

But the main thing that I see missing in the question is the possibility for PCs to run away from a fight. PCs are not committed to slugging it out with every single monster they see. And they are not necessarily locked in a battle to the death once an encounter begins. In old-school circles, this has become known as something of a common blind spot for new players, especially those who transition from video game to tabletop RPGs.

Here's Gary Gygax addressing the issue in the 1E AD&D DMG (1979, p. 61):

Combat is a common pursuit in the vast majority of adventures, and the participants in the campaign deserve a chance to exercise intelligent choice during such confrontations. As hit points dwindle they can opt to break off the encounter and attempt to flee.

In fact, in that 1E ruleset, avoiding is actually the first item in the list of possible actions within an encounter round (same page):

AD&D Actions

Following that, there are quite extensive rules for flight in the "Pursuit and Evasion of Pursuit" section (1E DMG, p. 67-69).

For a 5E take, consider Keith Ammann's book and blog, The Monsters Know What They're Doing; his standard essay format always lays out a protocol for when any monster will run away from a flight, and how. In fact (echoing the 1E action list), it's his first basic premise!

Every creature wants, first and foremost, to survive. If it’s seriously wounded (by my definition, reduced to 40 percent of its maximum hit points or fewer—you may prefer a different threshold), it will try to flee. Exceptions are (a) fanatics or (b) intelligent beings who believe they’ll be hunted down and killed if they do flee.

More specifically on the "lame duck" issue, Gygax in the 1E DMG has a sample of play that ends with a gnome PC, separated from the party, who gets surprised by a pack of wandering ghouls. The DM makes a decision to hand-wave the fact that the gnome is clearly going to get killed, and not bother playing it out (p. 100):

Groans as a 1 comes up indicating surprise. The DM then rolls 3 attacks for the ghoul that grabbed at the busy gnome, and one claw attack does 2 hit points of damage and paralyzes the hapless character, whereupon the DM judges that the other 3 would rend him to bits. However, the DM does NOT tell the players what has happened, despite impassioned pleas and urgent demands...

Last observation on the traditional expectation for TPKs; in 2004 on the Pied Piper forums, Gary Gygax outlined rules by which he boosted the abilities of the infamous kobolds on the first level of Castle Greyhawk every time they scored a TPK on players:

  • 1st TPK brought 12 more kobolds
  • 2nd TPK gave them armor class of 6
  • 3rd (near) TPK gave them all +1 HP
  • 4th TPK added +1 damage
  • 5th TPK added 4 2nd level and 2 3rd level kobolds
  • 6th TPK gave them tactical manouvering and a 4th level leader
  • 7th TPK upped AC to 5
  • 8th TPK gave them unshakable morale

So: The original co-creator of the game thought that running away should be the top option to PCs in many cases, was willing to hand-wave away doomed PCs without rolling dice for them, and reasonably expected a whole lot of TPKs.

  • 3
    \$\begingroup\$ +1 There are so many options PCs have before they even get into combat that PCs need to take responsibility for bad decisions. Buffing NPCs when the party is TPKed seems counter intuitive but it forces the PCs to think instead of brute forcing a problem. \$\endgroup\$ Commented Feb 1, 2020 at 2:47
  • 2
    \$\begingroup\$ @jgn Thanks for the +1. :-) It's interesting that to me that rule seems intuitively correct; either side equitably gets to level-up when they beat the other. \$\endgroup\$ Commented Feb 1, 2020 at 7:43

You can win in an RPG, but you can't win an RPG

How do role-playing games handle and/or avoid these situations?

By divorcing the objective/s of the game from conventional notions of "winning" and "losing." In D&D 5th Edition, for example, it's possible to clearly lose a single fight. Dragon crits on a breath attack, party is mostly or entirely slain, and the combat is lost. No arguments there.

Thing is, the game of D&D isn't that one combat--it's, potentially, an entire series of encounters (combat and otherwise) both in the past and future. It's entirely reasonable for a wiped party to all make new characters and continuing playing, and to consider that group to be playing the same game even though they're a new adventuring party.

D&D even tells us this in the Player's Handbook introduction on page 5:

"The Dungeons & Dragons roleplaying game is about storytelling in worlds of swords and sorcery."

It's not about "winning combats," though that element is certainly part of the game's appeal. Further on, we even get a clearer statement that separates D&D from a binary "win"/"loss" state.

"There’s no winning and losing in the Dungeons & Dragons game—at least, not the way those terms are usually understood. Together, the DM and the players create an exciting story of bold adventurers who confront deadly perils. Sometimes an adventurer might come to a grisly end, torn apart by ferocious monsters or done in by a nefarious villain. Even so, the other adventurers can search for powerful magic to revive their fallen comrade, or the player might choose to create a new character to carry on. The group might fail to complete an adventure successfully, but if everyone had a good time and created a memorable story, they all win."

When it comes to a lame duck encounter--say, a combat--there are certainly techniques and rules that can be deployed to avoid dragging out the inevitable. Generally, if it's truly impossible (as opposed to merely improbable) for something to happen, there's no need to use resolution mechanics at all. The GM states that the inevitable thing happens, and the game continues. (Sometimes, that inevitable thing might be "someone dies" or "you all get captured.") But, in terms of a lame duck scenarios for the game as a whole: they don't exist, because RPGs don't have win/loss states the way other games (chess, football, "Jeopardy") might.


Failure is an option

I think the main issue you are running into is thinking of scenarios where combat happens as a binary, where winning is Good and losing is Bad. Winning a fight is almost always the goal, but losing a fight does not necessarily mean that the game is over. There are two different parts of that situation that need to be looked at from a different point of view; the fight itself, and the results of failure.

Fighting Tooth and Nail

Let's say your PCs get into a fight that they aren't prepared for. Someone jumps the gun, attacks a much larger group, and suddenly you have a fight where the PCs are out numbered or outmatched. Things are looking pretty bleak, right?Well, only if the fight is against a horde of mindless monsters or something similar. Versus intelligent, humanoid enemies you can easily come up with a response other than "Murder those guys to the last man".

Your lone PC shoots an arrow at a group of knights and aggros them? Maybe the leader challenges the PC to an honor duel and they fight one-on-one. Or the knights beat the PC into submission and arrests them for assault. Or the knights beat the PC and strip them of their gear and leave them naked and bloody on the side of the road. There are multiple steps of increasing aggression that you can go through before you land on outright PC death.

Death is Cheap

Now, sometimes you might end up in a situation where you have a party wipe on accident. Maybe you didn't expect the PCs to actually charge that cloud giant, or you underestimated how hard your boss fight was. Or plain old bad luck struck and there was nothing the party could do about it. Running into a deadly situation like that should ideally be a rare occurrence, but when it happens you are still not out of options.

In a game like D&D where resurrection is possible, you can use that to both bring the party back and introduce new plot hooks. A friendly church finds the PCs' bodies and raises them from the dead, but only because their deity requires adventures for a Very Important Quest. Or just skip the middle man and have the deity itself raise them and send them out on a mission. If you really want to change the flavor of the game, have the people that killed the PCs revive them for a similar reason. The only thing worst than being wiped out by a bunch of kobolds is being raised by their shaman and put under a geas to work for him.

If your game is more realistic and coming back from the dead is a no go, you can go for the classic "blackout and captured" trope. The party loses the fight and gets wiped but instead of dying they are all just beaten unconscious. When they wake up they are prisoners of whatever bad guys beat them and have to go from there. Maybe they stage a prison break, or get forced into working for bad guys, or just get rescued by a different group.

Win or Lose, Keep it Interesting

The main point I am making is that there really isn't a difference between winning and losing in a fight, other than the direction your story goes. The only "lame duck" scenario you really need to avoid is a hard stop to the story. For anything else success and failure are just paths that you can take in any given situation. As long as it moves the story along then losing a fight, or even dying, isn't actually a bad thing.

  • 1
    \$\begingroup\$ Another post death solution is for the characters to awaken in an after-life. It might then require adventures for them to get back to where they were - but with the advantage of the experience acquired along the road. \$\endgroup\$
    – Isaac
    Commented Jan 30, 2020 at 12:46
  • 1
    \$\begingroup\$ @IsaacGood idea. I was going to suggest something similar, even if it is just a one-shot session where your party gets to play around as ghosts or in the afterlife. But this answer kind of got bigger than I planned even without that. \$\endgroup\$
    – D.Spetz
    Commented Jan 30, 2020 at 12:57

Since you ask how 'games' solve this problem, I'll give one specific example: 13th Age has a rule for fleeing (p.166). This is a tactic for the whole party, not individuals. If taken, the party are assumed to safely escape any losing fight, taking their casualties with them. The cost is a 'campaign loss' where the overall situation the party are facing worsens. This rule is optional, but using it negates the possibility of the kind of lame duck scenarios you suggest. The rule concludes 'The point of this rule is to encourage daring attacks and to make retreating interesting on the level of story rather than tactics.'


(I'll be talking mostly in D&D terms, as it's the game I'm most familiar with)

There is no such thing as a real lame duck scenario in role playing games in my experience, unless you lock your head on a single goal and refuse to shift. There's almost no situation that is truly impossible.

The first question that needs to be asked when talking about whether a scenario is truly lame duck is "Do the PCs have any chance to achieve their current goals?" If the answer is no, or possibly even extremely small chance, the second question is "Can the PCs change their goals to make something work?" If the answer to THAT is no, the third question isn't one the PCs get involved in, but this one is on the GM entirely "Can I change the way this ends, or do we let it end?"

One of my DMs before I became a DM liked to tell the story of one of his parties encountering a little girl. A quest giver to a very low level party that was actually a high (relatively speaking) level demon in disguise. The PCs caught on and attacked in what should have been suicide, but enough "half damage on a save" and radiant/holy damage, which the demon was vulnerable to, along with critical hit rules let the party not only survive, but kill off what was intended to be a mid game big bad way earlier on than the DM ever thought possible.

I was playing a game of Rolemaster as a Plasma High Elemental mage - our rogue thought it was a good idea to shoot the sleeping fire drake (level 10) with our level 1 party sneaking through. The thing did enough damage to kill most of us in a hit or two, but the way critical hits work in that game I got a lucky one the hit literally said "Your viscious blast crushes the foe's neck and shatters its spine, Foe drops and dies in 3 rounds." It was down, dying and unable to really fight with a broken neck.

Your 5 on 1 scenario? That's a lot like the Little Girl - they shouldn't be able to win but the way the rules work it's possible, just requires a lot of good luck on the player's side, and a lot of bad luck on the side of the mobs, but it can happen. And remember you are the GM - while rules lawyers may disagree, the rules serve you, not the other way around. You are the ultimate god of this world. One of the reasons for GMs to have a GM screen is so players do not see the rolls we make, the dice we roll introduce chance, but we don't actually have to live by the number that shows up. I've turned 20s into 19s before as the DM to further the group's fun, I've done the reverse to further the suspense and make fights closer than they otherwise would be, I've fudged damage rolls to leave the PCs with life when they would have died. You can always adjust your rolls.

Even before we look at what the dice rolls - what are the stats of these men on horseback - You may have set them or be taking them from some monster manual, but they are still your mobs in your world. The last fight my group did they were getting pretty badly beaten up by a group of golems that reformed if they weren't killed right - first I fudged the rules on how they had to be killed, giving more options than I would have under what was originally there, but still all but 1 of them were killed right, and they had really beaten up the party - so when the last one reformed I had it reform with its max HP cut in half. If the scenario is truly so close to unwinnable that it doesn't seem worth trying, change the scenario, drop the AC and HP of your men on horseback, weaken their weapons to do less damage. You can always change the monsters to make things more possible.

Even with all that, why do they have to keep fighting, why do they have to be killed? The player can surrender, the men on horseback can win knock the player unconscious and they wake up tied to a chair or in a cell somewhere, maybe in a hospital since they were just beaten nearly to death. There's always an out that doesn't end the game. You talked about shooting, introduce some kind of failure in their guns or crossbows, make them jam or the string break, so the fight switches to close range weapons letting your PC fight a retreating battle. Give your PC somewhere to flee the horses won't fit, so the men need to dismount or give up the chase. You can always adjust the scenario. Give the PC some allies, your scenario sounds somewhat western, how many times in westerns has the hero standing up to the outlaws caused the town to band together, once one person resists, make that happen here too, suddenly your lone man now has a militia at his back.

And even after all this the players can still lose, consequence is part of the fun of RPGs. And sometimes going down in a blaze of glory with a TPW is just as fun, if not more fun, than winning.


Flexibility after a loss

Computer games tend to be railroaded to a particular story with a predetermined direction, where that fight must be won to continue and losing the fight breaks the script and so is an end-game outcome.

On the other hand, role playing games can certainly have a satisfying story that includes that loss. Yes, both the players and the characters wanted to win, that was not what happened - that doesn't mean that the world ends and the story breaks down. Heroic deaths are a common trope in myths, legends, and other stories, and it works much better if it happens as a 'natural' outcome instead of a scripted event like in some computer games. The traditional three-act story structure relies on experiencing serious setbacks, and a lost fight can be an effective event for this.

In essence, a satisfying game means a substantially different thing in role playing games than in computer games, and always 'winning' is not necessary if the story can adapt or be adapted to properly include any loss (including very substantial losses) in the narrative. A computer game following a pre-made script can't do that, a good game master can.


I can tell you how I handled two of them recently. These of course aren't the only option. I'm not even sure its the best option, but the player in question did complement me for the handling of the first one at the end of the session (the second was deft enough they had no way of knowing it was unplanned).

The Bobby Ewing Approach

I'm running a 1-person (1-sidekick) campaign of the Dragon of Icespire Peak adventure. Depending on player choices, its possible on your very first outing at level 1 to end up in a confrontation with a creature that has the 3rd most HP of any in the entire game, and has 3 attacks per turn, each quite capable of 1-shotting a player. Certain of it, if it crits. The only realistic way out alive RAW at lvl 1 is to either abandon the quest entirely, or negotiate with it (and it isn't obvious that this creature is even conversant. Most such aren't, although I made sure to show it talking in Common as a big fat hint).

Upon encountering it, my noobie party (/wife) immediately attacked. I played it out (both dead in 2 turns, with the mob hardly scratched, even though its initiative was lowest). I then said that this was the character's vision of what would probably happen if she chose that course of action, and restarted the encounter. Call it the Bobby Ewing approach.

I suppose I could have really let it kill and eat the party. However, they were so early in the campaign (literally just started), that would have just rerolled the same char and tried again, and really why waste all that time to no different effect than just making it a premonition? My .. er .. player hates losing enough that the act of going through it once is more than enough to get her attention. I could see where some other player personalities might not learn anything if you don't make them do that extra work though.

You might think that having this happen on the first encounter is a special case, but so far my experience is that the very first encounter at level one is the most likely time to have a party wipe.

Unexpected Help

I had another encounter at level 2 start to go pear shaped due to an early crit (downing the companion in a 2-person party obviously halves the potential damage output, which is super harsh in round 2), and made a split-second decision to have a nearby mage have heard the noise and come running in to help. Roughly the same logic applies. It was still super early in the campaign, they knew the mage was nearby, and the intervention made perfect RP sense (even if it wasn't mentioned as an option in the materials).


RPGs don't leave the GM helpless.

A computer is helpless against its programming. If you're playing a computer RPG and on a combat map, and you're losing, you have to choose from the options programmed for you. If there's no retreat, all you may be left with is to get lucky and win or die and reload.

A GM is a human being. They're not following programming. They have the responsibility and the power to-

Let's start there.

Player Roles, or: How to Win at RPGs

An RPG of the type you likely think of when you think of RPG is played by players with very asymmetric powers and desires. Most players are regular players, each controlling usually one character in a world, the Player Character. They make decisions about what their character thinks, feels, does, and wants, but the outcome of their actions is often in doubt. One player is the game master, usually called GM but the name varies based on system. They're responsible for presenting the rest of the game setting: the whole world and everyone else in it, friends and foes, gods and devils, and they have the power to make vast, unilateral declarations like "the town gate is locked" or "these two kingdoms are at war" or "the sun explodes".

You can only win an RPG if you're a player character, and usually it's more of a self-directed goal? You're trying to get your character to somewhere satisfying to you, even if the entire world is against you. Death is usually something you don't want to happen, but there are also plenty of little things your character wants. Fame, fortune, love, respect. The role of the GM is to play the world believably, and giving you opposition to overcome that makes sense to the place and time you're at in the world is part of that.

It might seem like there's a mirror to the player's goals, that the GM can win too by stopping them, but that's not the case. The GM may play people dedicated to stopping you, running them honestly to the extent of their capabilities, but if they in their role as GM really wanted to stop you? They can blow up the sun.

Everyone's there to have fun playing an RPG, but you can have fun playing something even if you don't win. You can have fun playing pickup basketball if you don't win, but if you play pickup basketball and just try to have fun but not win, that might not work out so well for the fun of your teammates and opponents who are trying to win. So it's expected that everyone plays to win. It's "being a good sport".

In RPGs it's expected on a similar "being a good sport" level that the GM will play fairly but that the players will win... at least in the short term. Most of the obstacles a GM presents are not intended to be tests of life or death but rather to use up the players' resources and thus curtail their ambitions. So how do you get to the point where you're at a test of life and death that the players are failing, and what do you do then? Well...

Dice Hate Me, or: Disclaiming Power and Responsibility

Successfully playing a game where you cannot win but must present a credible challenge is something you might expect out of a true virtuoso player. Not everybody who picks up an RPG book to GM is one of those virtuosos, which is why most RPGs include some random element. The random element, usually dice, is given the responsibility and thus the power to decide what happens under certain circumstances. "You have a 70% chance to climb this cliff." "I'm climbing this cliff!" "Oh, good job, you made it!" feels weird as a conversation between just a player and a GM - how is the GM deciding that? Are they really being fair? But with a dice roll in the middle of it it feels fairer all around. There's a risk of failure that's visible to everyone, and it feels good when it doesn't happen!

And of course, when death, the ultimate end to all ambitions, is on the line, many RPGs go heavy in the dice and impose a lot of structure on the play - you know, "the combat system" - so that the chance of failure is there but the odds can be rigged in the PCs' favor.

As an aside, this is why most GMs don't just offer up a "load game" option, where time rewinds to before the combat and the party has another go. They could, of course, and if there was some misreading of the rules or something they sometimes do. And GMs always have the power to do that! But doing that means overruling the dice, which are supposed to be the fair arbiter here, outside the GM's fancy. What sort of credible opposition breaks the rules to give their opponents the victory?

This is different from something else which would probably show up in a computer game as "undo last action". Since it's assumed that players aren't deliberately trying to die, a GM can pause before a player e.g. opens fire on forty horse riders charging them down on an open plain and ask - wait a second. This is forty horse riders charging you down on an open plain. How are you planning to survive this? Often it'll turn out that a player had a different view of the circumstances than the GM - that they could hide to the side in the cloud of dust the horses were kicking up and take potshots which would be taken for more arbalest fire - so pausing to reflect and reframe things in this way is just making sure everybody has a clear picture of what's going on, not denying the dice the power they're supposed to have over life and death.

Oh, right. We were talking about how you gave the dice the power over life and death and now they're trying to kill the PCs. What's up with that and what do you do?

The Long Odds, or: Failure Is (Not) An Option

The sad truth is, a lot of RPG system and scenario design has a common problem: the players are "expected to win", so the designer may not spend a lot of time providing ways to deal with what happens when the players don't, or even just throw lethal obstacles in the players' path and expect that the players can just have the dice fall their way.

Like: a fiend from hell, bound to your destruction, that can teleport faster than you can run. It throws a frustrating wrinkle into a tactical combat, sure! But after a run of bad luck, even if the system has a formal numbers-based offramp for escaping from a fight, that whole "teleport faster than you can run" thing is probably going to seal the PCs' fate.

So a skill a GM has to develop is to recognize what they're leaving up to chance. Anything that's possible could happen! If you throw that fiend into combat with the PCs you are leaving it up to chance whether they live or die! Do you want that? But what can the GM do?

Remember, the role of the GM is to play the world believably. Something that wants nothing more than to kill the PCs is fine as long as it fails, but there's the chance it will succeed. In that case what you need is something the creature wants more than to kill the PCs, and a way for the PCs to get that if it's not out there already. The riders want to escape the town more than they want to kill, so they'll be satisfied with leaving a PC for dead and riding off. The fiend from Hell has been compelled into killing the characters by a contract, and it wants a better deal - maybe it binds the PCs to tear up its original contract, or demands a pact from them.

But again, if a system isn't designed to support a more graceful out from combat than "one side or the other is dead", even playing the world in this way can seem like you're making an exception for the players that the system never covered.

Systems That Cover, or: Our New Offramp

Here are three modern engines that support an alternate out from combat than "one side or the other is dead".

Powered by the Apocalypse/Forged in the Dark: "What combat system?"

Both of these engines support players taking actions at a dramatic pace. Characters can fight for their lives, sure, but they're not doing it in any special kind of rules structure. If things look bad they can just try to run or talk it out with no special interruption.

Both also generally don't have the GM rolling dice to determine what their characters do. The players roll dice to adjudicate their actions, and on player failure the GM is given a free hand to impose negative consequences - so the GM keeps a tighter hold on the power of life and death.

Fate: Pull the Ripcord

Players in Fate campaigns have a generally free hand to demand the GM give them an exit from combat on their own terms. The Fate conflict system can be used to resolve all sorts of things, not just fights to the death, but within it any player can choose to concede unless they're staring lethal damage in the face.

Conceding means that the opposition accomplishes their goals but the characters keep control over their own fate, and the GM gets to decide how that happens, not the dice. If just conceding would leave them in too bad of a position, or they have reason to want to leave combat without conceding, players can also stop fighting and run, kicking off a different structured play element.

Torchbearer: Setting Stakes

Torchbearer's conflict resolution system doesn't have an out. What it does have is a variety of ins. It is a game that tries to evoke an earlier, deadlier style of play, where Cord the Fighting-Man could just drop dead in the dungeon and with a minimum of mourning and reasoning, Bord, brother of Cord, would be right back in the fray next play session, but for all of that death doesn't have to be on the line.

A Torchbearer conflict starts with the PCs setting goals, and their lives are on the line only if they decide they'll be fighting to the death instead of trying to drive off, capture, run from, argue with, or outwit their enemies. (That's not strictly true. Characters who are already injured can still risk "a fatal injury" in any physical conflict, but again, they know that going in.) PCs aren't necessarily expected to win the non-fatal conflicts, either; one of the most important mechanics of the game is that suffering setbacks based on your personal quirks motivates you to do more during your downtime in the dungeon, and it's easiest to suffer setbacks when you already have terrible odds and they're not likely to hurt any worse.

Even in a fatal conflict, though, PCs still have something to fight for when things look bad, because the winners of any conflict owe the losers a compromise depending on how close it was, and the compromises for fatal conflicts are spelled out in the rules. A close fatal fight can result in the losers just being left for dead instead of dead outright, but even in something that seems like a straight-up stomp, Cord the Fighting-Man can at least try to force a minor compromise and hide a pouch of gems and the family axe in a place only Bord, brother of Cord, would think to look.


These aren't necessarily lame duck scenarios in all circumstances. I remember a campaign when our 5 man party of level 2 characters got attacked by a bunch of spiders, due to going where we shouldn't have really, despite the GM hinting that it was a bad idea. We got to the point where 4 of us were down, paralyzed by spider venom and we already thinking of new characters, but our paladin managed to roll super well, make all his con saves and take down the remaining spiders with 1 hp left and 2 crits in a row. It was so epic, that we still talk about it years later.

Don't deny your players the chance of having these moments.

  • \$\begingroup\$ Welcome to RPG.SE! Take the tour if you haven't already, and check out the help center for more guidance. \$\endgroup\$
    – V2Blast
    Commented Jan 31, 2020 at 8:57

DM says "derp, I guess you guys are all dead, next game?"

This doesn't have to be the end of the game. There's the concept of the decoy protagonist. If your heroes are dead, your players might re-engage into the same story with different characters, turning the recently dead (or dying) characters into NPCs.

This can work well, and there's even an interesting example from literature. In Worm, a serial web novel, there is a fight with one of the most powerful monsters in the entire setting, where lots of named characters die. The author literally rolled dice to determine who dies, and even the main character was not safe. Had the main character died, a side character would have been promoted to become the main character.

  • 1
    \$\begingroup\$ Those decoy protagonists could also be on a different timeline and arrive just in time to intervene and rescue the original characters! \$\endgroup\$ Commented Feb 27, 2020 at 7:49

The specific issue described here is (in the context of typical tabletop RPGs) not worth worrying about.

For example, let's say I'm running one D&D campaign per year.

How often will Total Party Kills occur? Maybe once per two years.

How often will there be a clearly unsurvivable situation leading up to the TPK? Once per five years? Mostly the PCs have a chance to escape or surrender, or they have the hope of turning things around with a series of lucky dice rolls, or they all die at once in a massive explosion.

If the hopeless situation does occur, how long is there going to be between "the battle has become completely impossible to survive" and "everyone is dead"? Probably only a few minutes. And in this time you can at least try to give your character a heroic death, and come up with some cool last words.

There is a big difference between the issue of "the party has already won" (frequent, because most battles are easy for the PCs, and boring) and the issue of "the party has already lost" (rare, miserable, but not boring).


Death is exciting

RPG's can avoid the described "lame duck scenario" because death is not a finality.

For example in D&D a character can be resurrected. Letting a character die can be the beginning of an epic campaign, not necessarily the end. There is the possibility of resurection but if not, death is a free ticket to another plane of existence.

Forgotten realms; what happens to a human when they die?

Most humans believe the souls of the recently deceased are spirited away to the Fugue Plane, where they wander the great City of Judgment, often unaware they are dead. The servants of the gods come to collect such souls and, if they are worthy, they are taken to their awaited afterlife in the deity’s domain.

How cool is that? Your character wakes up in a strange land and is approached by Gods and Devils!

Before it gets to that point however, there are a number of options for a player and a DM.

First for the player, they can choose to surrender. Like in life if you did not have the energy or patience to keep fighting, you might raise a white flag and render yourself to the whims of the enemy, hoping for mercy or an opportunity to bargain.

Bargaining with the enemy you completely avoid the "lame duck scenario" as you've started a whole new quest thread; losing the Excalibur sword, taking on a large debt or obligation.

If a player presented with the option to surrender decides to continue to fight, then it must be exciting enough for them, they imagine a possibility that they can win, even if it is not apparent to others (the DM).

Should the player be defeated, in D&D they are first (in most cases) knocked unconscious. What can proceed from that point is very exciting!

The DM has a range of creative options for what could happen next. Perhaps the characters were tied up and sold as slaves. Or they wake up after a long rest at a sacrificial alter to an evil god. Maybe there was a misunderstanding and the players wake up to a feast offered by their supposed enemy.

In an RPG rather than looking at character death as a finality or a "lame scenario", look at it as an opportunity for creative fiction.


By diversification of outcomes

Lame duck scenarios in game design occur when a player cannot win, but the game isn't over yet.

Removing one of the possible outcomes is not a problem. Leaving only one possible outcome is. It's the inevitability of one particular result that makes playing any game meaningless.

Therefore, "a player cannot win" is a red herring. This is not a problem at all, moreover, there are people that think that playing D&D to win is the most boring way to play.

What is a problem tho, is dooming your players to one particular result, regardless of their effort. That's why we say player agency is important, while railroading is usually bad.

Naturally, if your game of D&D becomes a a win-lose game, "players cannot win" means exactly that — they are doomed to one particular outcome, so their efforts are meaningless. But in this case, the bigger problem is that a TRPG game turns into a win-lose game. Even a game of chess can end in a draw. Even more so, role-playing games don't mean to be binary.

To put it in a nutshell, you can not "win" or "lose" in a role-playing game. What is "lose"? A character's death? But death is inevitable, while heroic death can be valuable. What is "win"? Killing all creatures of a particular alignment, patriality or race? It's not "win", it's genocide. In this sense, tabletop role-playing games like D&D is not a "game" at all, so the "lame duck" concept can't be applied to them at all.


You must log in to answer this question.

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