I have been the DM of a wonderful group of Stack Exchange friends playing an urban fantasy campaign of D&D 5e for some time now, and a question came up in our group Discord that has got me thinking about how I should form a strategy for it.

One of my players, whose character is Sherri, has a very high Persuasion skill (+5) and has been itching for more chances to use it in-game. I'm admittedly often a bit bad at giving Sherri the chance to do so, because the player roleplays her so effectively in conversation that sometimes I just mentally skip over having them roll for whether Sherri's argument was effective. However, I want to get better about incorporating more of Sherri's Persuasion rolls and making sure she gets to use her charisma skills so that the player feels good about investing into those abilities mechanically.

The problem arises, however, when I consider what happens and how to work it into the narrative if the player roleplays an extremely effective/persuasive argument but then flubs the Persuasion roll.

For example, we recently had a situation where a kenku was spilling some information about the Evil Organization to the players and "borrowed" the cell phone of an NPC the party is attached to, then called them and mimicked that NPC's voice to lure them to a neutral location to have a conversation. At several points in the conversation the players were trying to get information out of the kenku, including making persuasive statements, where I absolutely could and should have had them roll Persuasion, Deception or Intimidation.

"Why don't we head inside to discuss this, dear?" Sherri suggests, and goes to unlock the door of the library.

"I say nothing until we have a deal. This is not from the Dreaming Brethren. This is from me." The bird blinks rapidly again. "Am in danger just for meeting you. Was not supposed to do it."

It didn't work out that way, but I had several planned routes for this encounter, some of which would have ended with the kenku fleeing if the social encounter had gone sideways or they had spooked her into thinking the party couldn't be trusted with her information. I had planned out several "persuasion points" that, if the party hit on at least one of them, would have persuaded the kenku to trust them - i.e. showing they are powerful enough to protect her, telling her she won't be harmed, not having any visible weapons, and so on. The party did well at gradually persuading the kenku that they were trustworthy and that the kenku wouldn't be harmed for bringing them what she knew. At any point during that process, however, one of the players could have roleplayed being very persuasive and hit on one of those planned "persuasion points"... but then rolled a 3 on Persuasion. I didn't have a plan for what to do if that happened. Does that mean...

  • ...the character stumbled over their intended words or made a Freudian slip?
  • ...the character's argument came out correctly but they accidentally did something threatening, like subconsciously drop a hand onto their weapon?
  • ...the NPC misread their intentions or misunderstood their argument?

How do I work into the narrative a good argument but a bad roll?


12 Answers 12


Options, Options, Options

A game master has numerous options for resolving the potential dissonance between a spoken arguments and contrary dice results - options that work with either part being poor.

That said, it seems relevant to reiterate my primary advice for game mastery (one of a half-dozen guidelines for all GMs and nearly all games): Never call for a dice roll if you are not prepared to deal with extreme results (critical success or critical failure).

Mechanical resolutions I have seen used, and their effects, described in no particular order:

  1. Ignore Dice: For really smooth-talking tables, this can be a viable and successful option, removing the onus of speaking from the characters and putting it entirely on the players. Benefits: This can speed up play, improve the quality of dialogue, increase immersion, and enhance character relationships.
    Drawbacks: This can also make social mechanics worthless (and can unbalance the related game mechanics), exclude less smooth players, make gaming more stressful for socially tired or introverted players, result in the dissonance of a Charisma 6 non-proficient Wizard that is the most beloved orator in the kingdom, and can be inconsistent as the Game Master's attitudes (and thus the attitudes of literally everyone else in the setting) will inevitably shift over time.
  2. Ignore Role Play: For tables with low social skills, language issues, clashing cultural expectations, or a problem with GM bias this can be a successful resolution.
    Benefits: This is generally fair (the only place to fudge the outcomes is the DC), outcomes are usually consistent, the results are pretty fast to resolve ("I make friends with the knight; 17"), and the resolution is consistent with the associated game mechanics.
    Drawbacks: This can feel stifling for avid role players, generally makes social encounters less memorable and less engaging, and will often make the entire campaign less memorable.
  3. Reward Persuasive RP with Improved Rolls: Using the game mechanics to make a persuasive RP more likely to succeed (lower the DC, grant Advantage, etc.). In these cases, failures are going to be the result of a misplaced word or idea causing the other party to react without considering the logic of the argument.
    Benefits: Most game systems have mechanics for these kinds of 'ad hoc' modifications, this rewards heavy RP players without stressing the light RP players, and it is easier to be consistent and fair than with only RP.
    Drawbacks: The primary drawback is having to improvise a reason the argument didn't succeed - fortunately, the internet has shown us dozens of reasons people will ignore a persuasive argument (clash with fundamental beliefs, inattention, triggering other memories, found the color of the speaker's clothing offensive, thinking about their morning pastry, etc.) and the other players can help with that justification.
  4. Resolve the Dice, THEN Role Play: By resolving the mechanics, the player can alter their narration to fit the established mechanical resolution.
    Benefits: This can be a lot of fun, especially for players that enjoy describing how their characters fail, and removes the potential dissonance.
    Drawbacks: Unfortunately, not everyone can readily describe why their clever plans did not work. Similarly, some players do not respond well to other people describing their character's failures, which can turn into a social friction point. This can also discourage planning, and social engagement, since the outcomes are known before the fast talking starts.

How Does This Help You?

The four resolution techniques provide a framework for resolving social interactions.

Based upon your descriptions, I would recommend you use either the third or fourth methods.

Generally a character should fail in a manner consistent with their abilities.
When a strong character fails to lift something, it is generally described as the result of overconfidence (used a showy but inefficient technique), laziness (didn't actually try), or surprise information (the object was super-heavy or secured in place). When a weak character fails to lift something, it is generally enough to highlight that the character is weak.
Similar guidelines can be applied to social failures. If an inept character fails, despite a good idea, then they probably didn't convey their good idea or the other party ignored them because of the character's poor social presence. If a suave character fails then it should generally be because they were lazy, overconfident, or caught by unknown information.

Example 1: Terry Silvertongue gives an impassioned speech about homes and righteousness to inspire the knightly order to do battle. It fails. So why did it fail? Terry's rousing speech may have been delivered while lounging at a banquet table, sipping wine - showing a laziness that the knights resent and is at odds with the text of the speech. Or Terry may have presented the speech in a poor fashion (via poetry or song) or at a poor time (while the knights were drunk) because Terry was too confident that it would work. Or Terry may have chosen a bad argument for his audience - outsiders with mercenary attitudes, unwilling to risk life and limb for the 'greater good' or strangers that are often rude to the knights.
Example 2: Mortimer the Artist of Confidence tries to bluff a nobleman, and his scandalous date, into believing that the adventurers are agents employed by other nobility to handle issues surrounding a plot device that the nobleman is aware of. Mortimer fails. Why? The nobleman became incensed at the implication that agents of a mere viscountess would have any influence over his own behavior, and that these cads might bring scandal upon himself and his forbidden love - the surprise information that the nobleman's pride was harmed by the details of the ruse.

  • 4
    \$\begingroup\$ This is a great answer that summarizes my options. I agree with you that I might switch over to Option 3 or Option 4 as a way to incorporate and reward social mechanics without stifling RP. \$\endgroup\$
    – Sciborg
    Apr 3, 2022 at 1:50
  • 3
    \$\begingroup\$ +1 for options 3/4 but -0.5 for removing player agency in describing the failure. "Terry's rousing speech may have been delivered while lounging at a banquet table, sipping wine - showing a laziness that the knights resent" is likely to not go down well unless that is how the player's described their character. \$\endgroup\$ Apr 3, 2022 at 10:38
  • 5
    \$\begingroup\$ @AncientSwordRage Yeah, I'm not happy with that example but it fit with how a lazy presentation can sabotage the good words. And the GM always needs to keep character style and presentation in mind when describing a character's actions - doubly so for player characters since that goes back to player agency. However, harping on that is a separate question. \$\endgroup\$
    – ValhallaGH
    Apr 3, 2022 at 13:25
  • 4
    \$\begingroup\$ @AncientSwordRage sounds like @ValhallaGH rolled low when persuading you with an otherwise excellent post! \$\endgroup\$
    – ti7
    Apr 3, 2022 at 16:52
  • 4
    \$\begingroup\$ @ValhallaGH I agree with the "lazy presentation" example, but I much prefer to present it as a factor that the PC & player didn't know beforehand (and likely won't know afterwards). Maybe this is a specific word or two that unfortunately reminds the NPC of a cultural ethos to favour the group over outsiders; maybe it's similar phrasing that a corrupt local leader regularly used before they were exposed, so it sounds to NPC like this charismatic lying; maybe it just reminds them of their ex and puts them in a bad mood. Either way, the PCs were just unlucky in a way they couldn't foresee. \$\endgroup\$ Apr 4, 2022 at 13:38

"I am not my character."

One of my players occasionally says, "My character is smarter than I am, and I need a hint." And she's right; that's an issue! A similar situation is the player who's very good at solving puzzles, but happens to be running an INT 7 dwarf.

Ultimately, there isn't a really good answer to this sort of thing. You either have to allow the player to have their way regardless of their character's ability scores, or you have to let the ability scores hold sway. (And remember, you don't have to be 100% consistent with which one you pick.)

Roll before the scene goes too far.

One option is to let the conversation get far enough that you can see what the player is doing -- whether they're being charming, deceptive, or intimidating -- and the gist of what their angle is (which should dictate the difficulty of the roll). Once you have that down, you can call for a roll, before the player gets too deep into an extended discussion that's going to spotlight their personal abilities, and then finish the scene based on the outcome of the roll.

There's a few problems with that, though -- first and foremost, it can wreck a good scene to halt it in the middle to throw some dice and suddenly demand that the player switch off the charisma and make a mess of it. Worse, some players won't even be interested in continuing the scene after they know the outcome, and there's nothing good about having them suddenly cut off a scene and leave because they know it's futile.

You could give the player Inspiration for good roleplaying, which they can then use to get advantage on the roll, but you might find other players getting grumpy that they never get inspiration because they're just not brilliant actors, and anyway the player could still flub the roll. It's difficult.

Failure shouldn't mean the PC looks like a fool.

One thing that might help is a piece of advice I try to apply to my games when describing outcomes: Success is usually due to the PC's skill, and failure is usually due to the situation or the opponent's skill.

In combat, it means when a PC misses, don't describe how the player character just totally misjudges the strike and buries their sword in a wall; that feels unheroic. Instead, explain how the swordmaster deflects the blow with a practiced twist of his wrist or how a flash of lightning dazzles the PC at just the wrong moment, making her attack go wide.

On a skill check, the STR 20 barbarian didn't just become inexplicably weak and unable to open a heavy gate; instead, the gate is rusted in place or jammed by debris or locked shut by a latching mechanism. The failure isn't the player character's failure of strength, but the external situation that makes it impossible. (This also helps when you want to avoid retries -- once somebody attempts to lift the gate, the result changes the universe to reflect that roll; subsequent attempts to lift the gate will always fail because it's rusted in place, but maybe somebody has some acid on hand to clean it with...)

In social situations, you can apply the same concept. When a player makes an excellent argument but then rolls badly, it's not that the player character has suffered a failure of skill; instead the dice create a new fact about the world that makes the argument futile, not matter how good it is. The guard totally believes that this is in fact the Count of Cagliostro, but my greatest pardons, the duke has ordered that nobody is to be allowed inside without his personal approval, no matter their rank. The king is unswayed by your plea for help not because it was a poorly constructed argument, but because he has no help to give; the army is busy dealing with bandits on the border with Guilder and it would take days to get them back here, even if he were willing to let the bandits burn the borderlands in the meantime. The wizard finds your threats completely credible, but he can't divulge the identity of his contact because they always wear a bat-like mask and speak in a harsh half-whisper.

You can also go the "skill and strength" route here. You make an excellent speech, but the king is just so used to hearing well argued rhetoric that he isn't easily swayed. The wizard believes you'll try to beat him up if he won't help, but he trusts that his arcane power is enough to destroy you where you stand should you try it. The guard hears "do you know who I am" twenty times a day, and your ruse is simply doomed because it's the wrong way to approach him. The key is that it's not the PC making a mess of the attempt (mumbling incoherent threats, stumbling over his lie, accidentally calling the king's wife a hippo), but rather the opponent being too savvy for it.

It won't work for every situation, but it certainly covers a lot of scenarios -- and a nice benefit of this method is it can give the players a new plot thread to pursue. Instead of trying to find other ways to deal with the initial problem, it gives them a new goal. They now need to get the Duke's permission, locate and hire some mercenaries, or find the bat-man and chase him across moonlit rooftops.

  • \$\begingroup\$ and failure is due to the situation or the opponent's skill. - Except on a nat1 in combat. That's when you can just fully misjudge your swing despite knowing what you're doing with a blade, on a 1 in 20 chance. \$\endgroup\$ Apr 3, 2022 at 5:27
  • 3
    \$\begingroup\$ Potentially. I have no real problem with doing it that way, but it does lean into the problem with critical failures, where high level fighters are more likely to have a flubbed roll in any given turn so your melee experts are the ones most likely to "misjudge the swing" or whatever, contrary to what the narrative would seem to support. \$\endgroup\$ Apr 3, 2022 at 5:32
  • 3
    \$\begingroup\$ You don't have to play Yakety Sax for every nat1. But RAW, it would have missed regardless of what the target did, so that's a good argument for narratively having it bounce off a resilient part of their armor or just fully miss. High level fighters are also trying to get a lot done, not perfect the accuracy of each attack. And against mobile enemies, trying to guess ahead of what they're going to do next. "Swing and a miss, but then your next attack..." Without saying anything about looking dumb doing it. Good point that you don't want that for the supposed experts. \$\endgroup\$ Apr 3, 2022 at 5:40
  • 5
    \$\begingroup\$ This is a great answer — I love that you're working to show the players as competent and capable. I have one suggestion: In the section on social situations, you could include an example where it's not an external fact that caused the failure (analogous to the flash of lightning or rusted gate) but rather the opposition's skill is just better (analogous to the more skilled swordsman parrying an attack). Maybe a persuasion check fails because the other person is too motivated by their cushy position, or too politically savvy to be swayed by a speech. \$\endgroup\$ Apr 4, 2022 at 0:16
  • 4
    \$\begingroup\$ @ChrisBouchard You're right, I neglected to discuss that. I was thinking "Oh, everyone does that anyway" but then realized that isn't really following the metaphor, where the social version of "you just shoot badly" is "you said something stupid instead of making an eloquent argument". \$\endgroup\$ Apr 5, 2022 at 21:30

There are some excellent answers here relating to this specific situation, but I'm going to present a simple one that's more generally applicable to any and all situations where a character should definitely be able to achieve a task, but for whatever reason a roll was called for and failed:

Succeed at a cost.

The player technically achieves the thing they wanted, but ultimately failed in their unstated aim to not mess something else up in doing so.

  • They found the important clue on the desk, but due to a terrible Investigation roll they caused such a commotion that they can't erase their tracks, and someone nearby has been alerted.
  • The invisible rogue under the effects of Silence snuck past the guard post, but due to a botched Stealth check (even with advantage! That deserves recognition) they were so overcautious that significant time has elapsed and they won't be able to reach their objective and make it safely back out in time.

And in the case of the suspicious Kenku:

  • There's a compelling argument that the party are strong and able to shield their informant from harm. With their weapons and magic on display, Sherri is very persuasive. Too persuasive. The Kenku's confidence is bolstered and they tell what they know. Then later let slip to some of their more trusted colleagues at Evil Org that there's a new power in town that they should really align themselves with. Before long, rumour has spread around of an elite strike force ready to strike, and the party find themselves dodging more attention than they'd bargained for.
  • The kenku is understandably paranoid and highly strung. There's an exasperating amount of back-and-forth before Sherri can finally get across that she means no harm. Between the occasionally raised voices and peculiarities of Kenku communication, a third party has come to see what all the fuss is about. The party has to get rid of them, by further persuasion or a show of force, before they get what they came for.
  • The Kenku is convinced they won't come to any harm, but it's obvious now how desperate the party is for the information they have. They're able to provide the information as promised, but they have a request before they're willing to. Safe passage for them or a loved one, further service (that Sherri accidentally let slip the party is well equipped to do) or just cold hard cash.
  • 7
    \$\begingroup\$ As the player who plays Sherri, I quite like this answer. It strikes a nice balance between the RP reasons for succeeding and the external factors for failure (bad roll, represented in the universe). Incidentally, the kenku did wind up requesting safe passage (which we provided). \$\endgroup\$
    – Mithical
    Apr 4, 2022 at 8:51
  • 2
    \$\begingroup\$ I'm also in this game and I like this answer. \$\endgroup\$ Apr 4, 2022 at 9:49
  • 5
    \$\begingroup\$ It's worth noting that this is pretty much exactly what other systems suggest in this scenario. In FATE, for instance, if you get bad dice your character may still get what they want, but gets it at some serious cost. In other words, loads of people have played with this exact system for handling failure, and in my experience it works really really well. \$\endgroup\$
    – ymbirtt
    Apr 4, 2022 at 10:36

Choose between the rolls and the roleplaying

This is an unfortunate situation I think we have all been in from time to time. The incredible roleplayer makes an extremely convincing argument, then rolls terribly. Or the inverse situation: a character roleplays a terrible argument that would never in a million years convince anyone, then rolls a natural 20.

The question you need to ask yourself is:
Should social interactions be adjudicated by rolling dice, or simply by roleplaying.

There are benefits for both options.

If you choose to have social interaction be entirely based on a player's roleplay, then less charismatic and convincing players will have a hard time playing charismatic characters, which is a bad thing since games like D&D exist to help people pretend to do things they can’t in real life.

On the other hand, if you make roleplay entirely based off of the die roll, then you are basically encouraging players not to roleplay.

If all of your players are strong roleplayers, and all want to play this way, then I would suggest removing the die roll entirely. Sherri’s good persuasion score is no longer a number to be added to a die, it is permission for her player to roleplay her persuasively.

If most of your players are not as strong in roleplaying as Sherri’s player, going this route would likely make them feel unhappy. In that case, you would be better off going purely with die rolls

Mix the Two

Alternatively, you can mix the two options.
There are a few different ways to do this. The first and most common (in my experience) is to grant advantage on charisma rolls after a great roleplayed speech. This helps to make success more likely without entirely removing the roll of the dice.

You could also have players roll first, and then state their argument. If Sherri rolls a 3, then the player roleplays her stumbling or forgetting what she wanted to say. If Sherri rolls a 20, then the player can roleplay her fluently and convincingly.

  • 3
    \$\begingroup\$ To this, I strike a medium. If my players RP things extremely well, I may opt to just forego the role and let them succeed. If the RP is iffy, I may call for a roll to see how it goes. If the player doesn't want to RP it at all, they can tell me what they are going for and then just roll. \$\endgroup\$ Apr 2, 2022 at 22:25
  • 2
    \$\begingroup\$ This is a really helpful answer, thank you. I think my thoughts were leaning in this direction and this helped to clarify that I've generally been a "player skill/roleplaying over character skill/dice RNG" kind of DM. I particularly like the line that "Sherri’s good persuasion score is no longer a number to be added to a die, it is permission for her player to roleplay her persuasively." \$\endgroup\$
    – Sciborg
    Apr 2, 2022 at 23:15
  • \$\begingroup\$ How have players felt at the table when you've done this? I can imagine reactions being different based on bonuses for good rp where someone else doesn't even give themselves the opportunity for a bonus. Does your experience help with that potential? \$\endgroup\$
    – NotArch
    Apr 2, 2022 at 23:51
  • \$\begingroup\$ @NautArch I give multiple suggestions. Which are you asking about? \$\endgroup\$ Apr 3, 2022 at 0:34
  • 2
    \$\begingroup\$ @guildsbounty if that were the answer I would honestly -1 it. It incentivises roleplaying not according to skill and glorifes persuasion / bluff / intimidation over all other checks. Would you forego a nature roll if your player describes in great detail the real world habits of a great african swallow? \$\endgroup\$
    – Lause
    Apr 4, 2022 at 10:34

Consider two things

  • What outcomes you want to reach
  • How you want the players to feel about the roll.


Sometimes the character is doing something sufficiently straightforward it should just succeed. If the player has their character put forth a justification that would be persuasive based only on the facts, that doesn't rely on the character seeming trustworthy, then it should usually just succeed, just the same as if the player wants the character to do a simple action that would be an automatic success for anyone.

However, sometimes you do want to roll even if you don't really need to to determine the result. To keep some tension in the scene and remind the players that things can go wrong. To give the player who deployed the rhetoric a feeling of triumph, that they succeeded against a set difficulty, not just the GM's decision.

The Trick

In that case, you pretty much want the action to succeed because you already decided that, and you want the player to feel like they were awesomely successful (especially if persuasion is something their character is mechanically good at), but you also want the player to feel like their character failed anyway (else they'll stop feeling like the dice determine the outcome).

From there it's easy to think, you need an unfortunate set-back which wasn't a bad mistake on the character's part. E.g, the NPC is persuaded by the argument but raises one of your other pre-planned objections which the PCs can then address. Or is persuaded but gets cold feet at the last second. You need some improvising to find something that works but you usually can.

There's surprisingly much flexibility in determining the extent of a failure which you can use to your advantage here. It works in reverse too if you let the character roll to do something impossible, have them feel epic and look glorious while not actually achieving the impossible. You can do this in most scenes but be careful to keep an instinct for when the negative consequences do need to really bite to keep things realistic and when you can fudge it to fit the narrative that feels right.


You need a more productive concept of failure, so you can compromise with it.

This is what's tripping you up:

some of which would have ended with the kenku fleeing if the social encounter had gone sideways

When the output of a scene is "does the kenku stay YES or NO" you can very easily find yourself in this situation, where you don't want to say NO but the dice aren't saying YES, but this is a false dilemma. If you have a concept of failure that's more meaningful than just NO, you can compromise to less than just YES when the dice say failure but the setup for the roll seems really good.

It's alright, you don't really get taught well how to frame a dice failure in that way. The main thing dice come out for is combat, where failing your roll just looks like "nothing happens", but actually that's just from one combatant's perspective. What happens is you pay a price for failure and that price is "everyone else gets a go before it's your turn again", but that price lives out in the order of combat, not on the roll itself, so no one ever tells you to think of it like that.

The key to framing a failure you can compromise with is that in a believable world, everything exists for more reasons than just to interact with the player characters. You don't have to get thumbtacks and string and plot out your entire NPC roster on your bedroom wall or anything, but if you're bringing someone in as the focus of a non-combat scene, you should know what they want. When you know what they want, you can give them what they want on a hard failure and compromise with what they want when the dice say failure but the setup for the roll seems really good. You can fail forward.

Failing Forward, As In Motion

You might have a concept of "fail forward" that amounts to "say YES but be wibbly about it" but what "fail forward" means is that things change as a result of failure. In your scene, if the kenku panics and bolts, nothing has changed. You went through all that trouble for the kenku to steal a phone and set up a meet, but at the end of it all, things are just where they were before - the Evil Organization exists, and the PCs have no inroads to it.

(This is probably another reason why you don't want to say NO, because you set up this entire scene and demanded your players participate in it, and after all that what's it going to feel like to get nothing?)

Now, you don't always need to have a particularly progressive concept of failure - sometimes it's fine to just keep going the way you're going, especially if the players are acting on their own initiative to e.g. try and get some free armor out of the blacksmith. As long as you know what your NPCs exist for more than to interact with the player characters -- what reason they have to keep doing what they're doing -- you can just have them keep going the way they were, the player argument very good but not stronger than their own motivations. (Future's looking grim, blacksmith's gotta eat, there is no "smiling bard" discount, if you won't buy it at this price there'll be someone else along who will.)

But when your first impulse is to set up a YES or NO scene, how do you come up with a failure that changes things? Once again, consider what the NPC exists for, and how that can do something to negatively impact the PCs' lives, more than just wasting their time to do nothing.

Potential Ways Out

Now, exactly what the NPC in this case should do is something you're much better equipped to answer than I am. It's going to depend on all the campaign prep you haven't shared and which would probably be too long to share.

But, looking at just what you've provided -- the kenku wants to make a deal with the PCs but is paranoid and easily spooked -- here's how I'd use either of those as inspiration for a failure that changes things that I can compromise with.

If the kenku wants to make a deal, a way this can work to the PCs' detriment is if they make a bad deal for the PCs - like, the PCs have to pick some number of the following onerous conditions (but you can claw that number back as a compromise for a good setup):

  • "An unbroken watch, for security." One PC must keep a constant watch over the kenku while they sleep. Whoever does gains a level of exhaustion.

  • "I must have these wards, for protection." Every day the PCs have to burn some modestly powerful spells on the kenku.

  • "They are too easily compromised." The PCs must break off relations with another useful NPC they have been working with.

  • "You need some skin in the game." (Extremely scenario dependent and this should count for 2/only come in on total failure.) A PC must be captured by the Evil Organization (and their player can run someone else in the meantime) so the kenku knows the PCs will get serious about stopping them.

If the kenku is paranoid and easily spooked, they have an extraction plan - a repo squad on call to help them escape.

  • On a total failure, they would call in the squad and try to fight their way out. Unless the PCs don't fight back they'll at least have a few captures of their own to learn something from.

  • As a compromise for a good setup, they panic and call in the squad accidentally, then beg for forgiveness and hide. The fight is easier, but the squad will fight to rescue them. (Oh, now they're saying they don't want to be rescued? Obviously they were charmed, and they can fix that back at base.)

  • On an outstanding success? Maybe they offer to call in the repo squad and suckersmack them to hand the PCs an easy win.

  • \$\begingroup\$ This is great advice, but on a first read I'm not seeing how this helps the OP narrate the extremely persuasive argument failing. Let me know if I've missed something? \$\endgroup\$ Apr 22, 2022 at 10:21
  • \$\begingroup\$ I've edited it to more visibly forward the use of compromise in narrative. Is this clearer? \$\endgroup\$
    – Glazius
    Apr 22, 2022 at 14:29
  • \$\begingroup\$ That's much better! \$\endgroup\$ Apr 22, 2022 at 14:39

Know your tools for persuasion.

As a DM you have several tools in your toolbox to adjust how an encounter goes. You have the DC of any particular check. You can give a check advantage, or disadvantage. You can decide if a check happens or doesn't. You can decide how many chances they get, or how many times they need to succeed.

All of these are the sort of things you can tweak based on how persuasive a player is. A very persuasive argument might get a lower DC, or advantage. If you want a particular outcome you can give them multiple points to roll. Conversely, you can give disadvantage, a high DC, or make them need to roll multiple times for a large success.

Know what you can offer for failure, doing ok, or success.

Once you have some idea of your tools, you need to know for the encounter what you can offer them. You need to move the plot forward, and so whether they fail or succeed you should be making something interesting happen, and so be aware of what could likely happen. I make sure all events are active and good for

For example, I might make a mental model of the Kenku conversation like this.

Failure consequences- the Kenku calls a death squad from the evil org. The Kenku reveals she has evidence of a past crime and uses it to blackmail them into investigating the organization. The Kenku runs back to their secret lover in the evil org to spill all.

Base information to share. The evil org is luring orphans in like the kenku by acting as a family. The evil org has started vanishing people and this worries the kenku. The evil org can help the players with experimental biotech problems.

Success information to share. The Kenku knows a secret path inside the evil org. The kenku knows the public identity of a secret member, their lover. They suspect the org is abducting children to help de-age the leader, as they seem to have been getting younger.

Some information you can give regardless of what they say, and you can adjust what happens if they fail or succeed. If you are feeling friendly you might give them a second roll to save the situation. If not, the plot can advance as something interesting happens.

Know the NPC's incentives and objections and make them talk.

NPCs who just won't talk because they're distrustful are annoying. Players tend to not like them. Talkative NPCs are more fun. You should be aware of what the NPCs want and why they're talking. For that, know what their incentives are, and what their objections are.

For the Kenku say, their incentive may be getting the party to investigate the evil org abducting orphans and they want the leader dead. Their objection may be they don't want to get their lover in trouble and don't want to be killed. Rather than having them simply be distrustful, give them a concrete reason to not talk about some information. They can talk on other matters, but not this.

Separate approach from charisma.

When you're at that key moment, unless you've agreed that roleplaying can grant you much greater success, reward them for the approach, not for the roleplaying. If they target an incentive or an objective and fix it, then they'll do much better.

If you want to show a failure, if the failure is big it should be big because they hit an incentive or objective. For example, if they said "Tell us everything or we'll kill your lover." Then that's something she could go to call a death squad on them over. If they just say "We'll give you gold to tell us all." And they fail, you might just not tell them special info.


I'm no expert on DMing, but the best advice I ever got is to

Read the Room

This may be a difficult technique for anyone who is playing over the internet, but you've got to remember that the game is fake, and the people controlling the characters are real. If the person proposes a brilliant argument that confounds everyone in the room, you don't need to ask them to roll a charisma check; just have their argument succeed.

In the same way, if a caring and kind person has their character try to lift up an NPC to save them from a collapsing dungeon, don't have them roll a strength check in the first place; just let them do it. The point of these games are to have fun, not to stamp down your friends!


Don't take this mindset the wrong way. Dismal failures can be what makes a DnD run special! Just make sure these failures are lighthearted and fun, not devastating or mean.


If you do choose to take my advice, follow a 'no-takebacks' rule. If you do have someone roll a charisma check and it fails, unless if it's an emergency (so to speak), don't let them flout the bad roll. If you just say "Ah, screw it, you succeed", then it makes it seem like your just showing favoritism and being mean for every bad roll after that.

This can be damaging to your group, so make sure to avoid it by asking yourself "If they were to fail, would the group (not the game, the group) be worse for it?" Just keeping this in mind has made a lot of my dungeons so much better!

(One last thing)

I only really answered the main, overlying question of the bunch. For the other ones, I would say its best to determine them based on the roll.

ex: Suppose Bob rolls a 3 for charisma, while Alice rolls a 9. Both are failures, but the NPC still perceives Bob as more hostile because Bob rolled lower, and therefore landed further from his goal than Alice.

As a sort of rule-of-thumb, the differences in the scores on the dice should correlate with the differences of what happen in the situation.

I am new to this particular Stack Exchange, but I hope my answer was helpful! :)

  • 1
    \$\begingroup\$ The OP is explicitly looking for ways to let the player flex a skill. Choosing to not make them roll doesn't work with that \$\endgroup\$ Apr 3, 2022 at 10:13
  • \$\begingroup\$ @AncientSwordRage I disagree. The point of this technique is to choose your moment and see what's at stake BEFORE you make them roll. If you follow the 'no-takebacks rule' and ask yourself the question beforehand, like I said in my answer, it can incorporate itself within the standard DnD run. \$\endgroup\$ Apr 3, 2022 at 15:15

For extended social interactions I prefer not to rely on a single attribute check. Bluffing a guard would be a single roll, but for a complex negotiation I tend to assign a number of cumulative successes/failures. That way a capable character won't fail as a result of a single poor dice roll.

As the character makes their arguments they make rolls until they accumulate enough successes to convince the NPC to do what they want (or enough failures that the NPC refuses).

For example, you have determined that three successes are required to convince the kenku. Sherri, might start off by promising the party is unarmed and won't harm the kenku, but fails the first roll.

You narrate how the creature is unconvinced and scared and tries to leave saying they should not have come.

Sherri perhaps changes tack and intimidates the kenku somehow. She passes the roll and the kenku is cowed into staying (the failed roll has not ended the encounter, and Sherri now has one success and one failure). You narrate that the kenku is impressed by the party's strength.

Sherri promises to use their strength to protect their informant. She succeeds on her roll and the kenku believes her.

Finally she appeals to the greater good or whatever and gets her third and final required success.

Sherri's player has got to role play the encounter, Sherri has got to actually use her proficiencies, and the first bad roll did not result in an unsatisfying and potentially inconsistent result.

  • \$\begingroup\$ I do not see why this has down votes. This is a reasonable approach by applying multiple throws to balance out a single bad roll. This is done elsewhere with the big one being you need to fail multiple death saving throws before actually dying. \$\endgroup\$
    – Anketam
    Apr 5, 2022 at 20:15

Roleplay the receiving end

Communication is always a two-side business: one side talks and the other listens. Even the most elaborate and well-thought argument exposed by the most eloquent orator will not work if the other party fails to understand (which may be unlikely if the orator was very very good) or doesn't want to accept the idea.

So, even though the player has role-played an excellent persuasion, nothing prevents the DM from deciding it just fell on deaf ears: the NPC is just not persuaded. The NPC may have prejudices towards the character or his ideas or the points of his argument, or he may have better reasons not to accept the character's requests, or he may simply not feel like giving the characters what they want. A player-NPC conversation can easily be brought to the outcome the DM expects, because the players can't (and shouldn't) control what the NPCs do or think.

When it comes to PC-PC interactions, instead, I think that no rolls should be done. The players should stick to their characters though: for instance, if there is a human warrior who hates dwarves with a heart, when the dwarf ally (whom they hardly tolerate to work with) suggests a good plan, the warrior should be very reluctant to accept the suggestion; if the warrior's player gladly welcomes it he just did bad RP, if he refuses it prejudicially he's being faithful to the character - and dice rolls shouldn't be necessary.

It's a different thing if the act of persuasion is magically enhanced. If a caster uses Charm, it's the spell that it's doing the persuasion, just as if someone tried to lift a heavy object using a machine: either the machine or the spell works or it does not, no role-playing can change that.

  • 2
    \$\begingroup\$ FYI, what you're saying isn't necessarily wrong - and definitely not bad. Take a gander at the citation expectations for rpg.se. Subjective answers are more than okay, but it does require some more thought and ability to translate your related experience to OP's problems. \$\endgroup\$
    – NotArch
    Apr 4, 2022 at 17:00
  • \$\begingroup\$ Thank you @NautArch. While I didn't point it out, my advice comes from personal experience (even though I was usually a player: this is what I learned discussing with a couple DMs). Anyway, it's still highly subjective; I'm not sure if just saying "this is how we did it" is acceptable. \$\endgroup\$
    – Simone
    Apr 4, 2022 at 21:22

Don't roll if the outcome is not in doubt.

This is pretty simple. If the argument is good enough that people would accept it on merits alone they do not need to be persuaded. If someone offers a guard 3 gold coins to do something they won't even get in trouble for, why would they.. ever?... say no? They'd need to really hate the guy doing the offering to not take his money.

Rolling dice when something is impossible or should just happen automatically leads to odd situations and should typically be avoided.

If the outcome of the roll and the roleplaying do not match, adjust reality to fit the roll.

This can be done in a lot of different ways. Perhaps the player at the table has a silver tongue and roleplays a convincing and well delivered persuasion, but then rolls a 1 and is playing a barbarian with -2 charisma mod for a total of -1 on a DC 15 check.

You could

  • Tell the player his barbarian grunts, farts, delivers the same words with completely different intonation, sounds like he doesn't mean it, and the man gets the vague impression that the barbarian is interested in his wife in a non-platonic manner. This is the 'same words, but different soft skills' approach.

  • Tell the player that his barbarian tries to say all those things, but the words get muddled. You then decide how the words get muddled and what he actually says. If you've ever said things that absolutely were not what you meant to say, there should be ample inspiration to draw on. Maybe there's a mid-conversation tirade on Etruscans and how they can't be trusted, maybe the barbarian gives the impression of being drunk or high, maybe he starts talking about loyalty to the kingdom and it comes off as a threat (i'm secret police) rather than inspirational. This is the 'you decide what your character tries to do, but the DM decides how that pans out based on rolls - and that applies to social situations, too' approach.

  • Alter the NPC based on the roll. Sure, the barbarian delivers this whole speech. But he's tall and wearing furs, which makes him Gallic in this individual's eyes. And thus his entire being and words are innately suspect, which adds a darker cast to this entire interaction. So while the barbarian is quite persuasive, he's working uphill instead of on level ground - almost nothing he says will make a dent now. This is the 'roleplaying > dice rolls, but the outcome will still stand' approach.

There's also other ways to handle it. The shopkeeper agrees with you, but then his wife comes in and turns out she's the decisionmaker in their marriage. Etc etc. A million ways to either fit the situation to the dice roll without changing the roleplay, or alter the roleplaying, with varying levels of granularity in terms of player control of their character actions and verisimilitude of the world setting.

  • \$\begingroup\$ The root issue here is that the player wants to roll more. In your other suggestions, there's not a lot relatable to the OPs specific situation that doesn't take away from player agency \$\endgroup\$ Apr 7, 2022 at 10:16

It's up to you whether to call for a roll in the first place. Dice exist to adjudicate situations where your intuition is insufficient. If you decide people are persuaded, they are. You can make a different decision for each individual, if you want. No roll needed.

  • 3
    \$\begingroup\$ The player explicitly wants more opportunities to roll though, so that's not in question. The roll happens - how to narrate it? \$\endgroup\$ Apr 17, 2022 at 8:28
  • \$\begingroup\$ What the player wants is irrelevant. You decide if a roll is required. If they roll, you decide the results. \$\endgroup\$
    – srcs
    Apr 17, 2022 at 21:31

You must log in to answer this question.

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