So yesterday I was DM’ing our weekly D&D 5e session. For practical reasons it was a one-shot. At some point, one of my players, a Lvl 10 Bard, ends up in the BBEG’s lair alone (an adult red dragon, CR 17). Partially his fault for ending up alone (because he tends to act impulsively), but that’s not really relevant here.

So the player wanted to persuade (+13) the dragon saying “I come in peace". I argued that, seeing as there are explosions happening all around caused by the party (and some suicidal Molotov-cocktail Kobolds), there is no way he’d fall for that. He then asked me to let the dragon counterroll insight (which, for a dragon, is only +1). Even when I gave the dragon advantage on his insight roll (20), and the player disadvantage on his persuasion roll (21), the player still won (yup that’s the effect of a +13).

So I thought to myself “fair I’ll give him one round, so he can try to get out” and let the dragon fly up. But one round later, the player did not try to get out nor try to call out to the rest of the party, instead smugly believing he nailed it. So then in this second round, the dragon said “did you really believe I fell for that?” and breathed some fire over him. The player was however not amused, and said this wasn’t supposed to happen because he won his persuasion roll. We continued the game anyway (“ok ok you’re the DM”), but kept on discussing about it until the end of the session; he really felt he was treated unjust.

I am really wondering if I acted wrong, and how I could have dealt with it in a better way. I particularly don’t like the ensuing discussion afterwards, as it slows down the game a lot (and makes me hesitant of putting them in nasty scenarios even when it’s partially their own fault).

Some thoughts:

  • an adult red dragon is a smart and arrogant tyrant. I felt it to be really in character to let him believe for 1 round that he was persuaded.

  • I found the role playing for the persuasion rather weak. I’d have been more keen to follow his plan, if he’d say something of interest for the dragon, like “I can spy for you on the nearby town”.

  • From a game mechanic point of view, should it be possible to persuade your way out of boss battles? That seems very flat and boring, but with his +13 persuasion he’s gonna succeed most of the time

  • Does the roll difference matter? He only won by one more than the dragon’s roll (which was one of my arguments why he’d realise very fast anyway)

I’d love to get some insight on how you’d have dealt with this.

  • 33
    \$\begingroup\$ Also, as an aside, if the Bard was most definitely not coming in peace, as seems to be the case from your description, then that should have been a Deception roll, not a Persuasion roll. Of course, a mechanically minded player who has proficiency and/or expertise in Persuasion (likely, if they have +13) but not Deception will always, in my experience, try to argue that it should be a Persuasion roll, but if they're actually trying to deceive the NPC/Monster, don't fall for that; it's a Deception roll. Don't let the player deceive you. \$\endgroup\$
    – NathanS
    Commented Aug 27, 2020 at 7:42
  • 2
    \$\begingroup\$ What was the reason the party had to fight the dragon? \$\endgroup\$
    – enkryptor
    Commented Aug 27, 2020 at 7:55
  • 2
    \$\begingroup\$ @NathanS Fair point about Deception, didn't realise that (though he has +13 on deception as well, wouldn't have made a difference :-D) \$\endgroup\$ Commented Aug 27, 2020 at 8:16
  • 3
    \$\begingroup\$ @enkryptor It's the Clash at Kobold Cauldron adventure (free module in Foundry VTT). They are hired by a city to find the distillery that creates a highly addictive liquor that totally wrecked the now-full-of-addicts-city, and deal with it. The dragon is the owner of the distillery, and this is all part of his scheme to get all their gold 'legally' and completely dominate the region \$\endgroup\$ Commented Aug 27, 2020 at 9:05
  • 3
    \$\begingroup\$ Similar recent question: What to do if a player expected much more than DM gave? \$\endgroup\$
    – user56480
    Commented Aug 27, 2020 at 20:26

8 Answers 8


You are the DM, you decide the rolls

The problem was when this happened:

He’d then ask me to let the dragon counterroll insight (which, for a dragon, is only +1).

Why is the player dictating to you when rolls happen? As the DM, you are the only one who should call for rolls. Even if the player presumptuously rolls for persuasion, for example, it doesn't actually count unless you say it does.

Remember that dice rolls are to determine an uncertain outcome. If the outcome is certain (i.e. "there is no way he'd fall for that"), then there's no need to roll.

From PHB, p. 174:

The DM calls for an ability check when the character or monster attempts an action (other than an attack) that has a chance of failure. When the outcome is uncertain, the dice determine the results.

Although the above quote mentions failure, the same is true of only rolling if there's a chance of success. From DMG, p. 237 (borrowed from Szega's answer):

When deciding whether to use a roll, ask yourself two questions: Is a task so easy and so free of conflict and stress that there should be no chance of failure? Is a task so inappropriate or impossible - such as hitting the moon with an arrow - that it can't work? If the answer to both of these questions is no, some kind of roll is appropriate.

  • 2
    \$\begingroup\$ +1. This was my first thought as I was reading the question. \$\endgroup\$ Commented Aug 28, 2020 at 18:56
  • \$\begingroup\$ It's true that the player calling for a roll is counter to the RAW, but that's not really related to the problem that happened here. The problem was caused by miscommunication over what the roll meant: The player thought success meant deescalation from combat, while the GM thought it meant buying a little extra time. \$\endgroup\$
    – GMJoe
    Commented Aug 28, 2020 at 20:50
  • \$\begingroup\$ @GMJoe I disagree; the fact that there was a roll at all set up that miscommunication. If the player had understood that the DM should be the one to call for a roll if the outcome was uncertain, then there would have been no roll to have any miscommunication about. \$\endgroup\$
    – NathanS
    Commented Aug 28, 2020 at 21:46
  • \$\begingroup\$ @NathanS By the same token, if there had been no BBEG, neither GM nor player would have called for a roll, thus preventing the problem. Yes, players calling for rolls can be a problem, and yes, it happened here - but it's not the direct cause of the player getting upset and arguing with the GM, or an upset argument would happen every time a player called for a roll. \$\endgroup\$
    – GMJoe
    Commented Aug 30, 2020 at 22:41
  • \$\begingroup\$ I still disagree. Your argument does not convince me. I still believe that the player calling for a roll was the direct cause. I don't see much point in us continuing this. \$\endgroup\$
    – NathanS
    Commented Aug 30, 2020 at 23:18

Your problem is that there was a mismatch between what you thought possible and what the player thought possible and how you both communicated these to each other.

1. What is the intent?

What I am missing from the player's description of the action is the intent. What was the bard trying to achieve with this? Start a diplomatic conversation? Attempt to get closer and surprise the dragon? Buy time for the rest of the party to arrive? If the player does not say, and it is not clear, the GM should ask for the intent behind the action. And here it matters.

One, depending on the intent it can be either persuasion or deception. The player likely asked for persuasion because the bard has a high modifier for it, but as the GM you can overrule their choice (actually a player cannot even make such a choice, but that is another matter). But I advise you to inform the player of this before the attempt: "If you try that you will have to roll deception, not persuasion. Do you still want to go ahead?".

Two, it will indicate what outcome the player is shooting for. Which leads us to:

2. Should there be a roll?

When deciding whether to use a roll, ask yourself two questions: Is a task so easy and so free of conflict and stress that there should be no chance of failure? Is a task so inappropriate or impossible - such as hitting the moon with an arrow - that it can't work? If the answer to both of these questions is no, some kind of roll is appropriate. (DMG 237)

Given the circumstances, is there any chance that the dragon will believe him? Even if he might, will it change their actions? An arrogant and powerful entity might just think "I do not care whether you come in peace, Dinner.". If so, there should be no rolling at all.

There is another way to think of this. The question is, would you want to have either outcome in this game? Never ask for a roll if one outcome is such that you think it would break the tone, pacing or consistency of the game. Should this be able to happen (or fail)? You can then rationalize it any way you deem appropriate.

In your situation I would say that it would have been reasonable to not even ask for a roll, if the desired outcome was to make the dragon friendly or passive. This was likely what the player wanted.

3. Communication of the outcome

You thought that just talking the dragon into submission was out of the question, but you felt that giving the bard a head start would be a reasonable outcome of the roll. Tell them so.

In most cases the character is able to assess the situation in-game. Here they might have heard of the attitude of this dragon, or just know legends of red dragons in general or might be able to gauge their reaction in the moment. You can just say "The dragon is obviously aware of the havoc in their lair and seems angry. You would need an exceptional excuse to calm them, but you can buy yourself a few moments if you fast-talk and confuse them."

Your problem stems from just deciding this in your head. By telling the player to roll, you lead them to believe that what they intended (but also have not said clearly) was possible depending on the roll.

About your additional questions:

On talking out boss battles. There are no special rules for "bosses", calling it a boss just describes your intention when creating the encounter. As such, reasonable discussion might be an option. You should keep an open mind, but that does not mean to just allow any and all attempts at parley.

On roll differences. They generally do not matter, you pass or fail (see PHB 174 on ability checks and DC-s). There are some monster abilities that specifically call out that "failing by 5 or more" will have more severe effects. However, you as the DM can improvise such a rule on the spot if you wish. Communication of the possible outcomes is still important, though.

  • 2
    \$\begingroup\$ Note it is also worth encouraging players to roll insight checks themselves to see if the enemy is still hostile. enemies can lie. Intelligent enemies may not act hostile even if they are still hostile. \$\endgroup\$
    – John
    Commented Aug 27, 2020 at 11:46
  • 6
    \$\begingroup\$ @John That is out of scope here and would undermine the point, even if the principle is prudent. Also players do not decide to roll anything. I personally abhor the "I wanna roll Insight" approach and use Deception vs. passive Insight for that kind of thing. \$\endgroup\$
    – Szega
    Commented Aug 27, 2020 at 12:51
  • \$\begingroup\$ +1 for being the only answer to mention clarity of communication between player and GM as the root cause of the issue. The other answers' fixation on the player calling for the roll is mistaking a symptom for the disease: The problem wasn't caused by the player calling for a roll, but by his and the GM's failure to discuss what the roll represented in-fiction. \$\endgroup\$
    – GMJoe
    Commented Aug 28, 2020 at 20:48

The first problem, which many people have pointed out, is that the circumstance of allowing a roll should have been dictated by you - not by the player. Asking the dragon to roll a contested insight against their persuasion, or even deception, was not within their power as a player.

However, let's say the circumstances are that you do want to allow a roll - in this case to give the player a fair chance to at least ruse the dragon long enough to escape. That is after all what you said you wanted to do.

A successful roll does not make the impossible possible - nor is it an auto-win button

You as the DM get to decide the outcome of the roll - and the outcome you decided on was 'the dragon is distracted enough that it isn't really thinking clearly, so you could use this chance to get away'. The player can argue about it, but you are the final arbiter of that.

The one thing you could do differently is to narrate the results of their actions on the dragon and how effective their ruse actually is, so that it can better inform their action - make it clear that the fast-talking Bard has bought himself some time but only that, and that the numerous explosions and noise around them are going to clue the dragon in pretty soon.


The answer to the topline question is "yes", encounters can have solutions other than combat

One of the biggest, unique strengths of TTRPGs is that anything the players can imagine, and their characters can attempt, could happen. If, in your determination as a DM, a player's proposed course of action is not literally impossible, it's definitely something you can allow. (Side note: you can also allow things that are literally impossible, but that's a different matter).

This exact sort of situation has happened to me at the table many, many times: my players seem to love outmaneuvering boss-style combats. It's surprising the first couple of times, but at this point I nearly always have backup encounters ready in case players try to short-circuit the boss combat.

My key guideline is that the stakes of situations over the course of a given chapter of an adventure should be roughly the same, regardless of how players approach encounters. They can talk instead of fighting, but the talking strategy should bring similar risks: if losing the combat means the characters would be imprisoned, doing a poor job of chatting up the NPC might have the same result (or one that is similarly meaningful within the plot and narrative of the game). Or perhaps persuading an intended boss out of a combat encounter involves the players agreeing to undertake a quest for that NPC, which happens to involve its own boss encounter.

I find this easiest to do with NPCs that have more fleshed out personalities and, crucially, goals and obstacles. The PCs can try to work with the NPC, but the NPC generally isn't motivated to work with (or for!) the PCs.

This player's complaints are valid, but it was a muddy situation

As others have already pointed out, players don't get to ask for rolls (per the rules, most people in my experience slip up and ask for rolls directly, which is not a huge deal). The first issue with the encounter as described was allowing the player to choose to roll, combined with allowing them to describe the roll the NPC should make as well.

This is problematic, but not insurmountable. It might have been best to tell the player that they can try to persuade the dragon, but then not call for a roll (and certainly not a contested roll). If you want to keep a story or encounter on the rails, then rolls are inappropriate-- the randomness that makes them fun means you can't guarantee outcomes very well.

However, once you had called for the rolls, it became far less reasonable to override them. That's the valid part of the complaint-- the persuasion succeeded, because that's what those roll results mean. The roll was not the bard trying to appear to succeed.

This is mostly about expectations, and how they are set by events at the table:

  • Had you decided that the persuasion was impossible, you could have let the player roleplay things out and see their efforts be unsuccessful in the process.
  • Had you decided that a red dragon would not really care if an interloper came in peace or not (which is very in-character for a red dragon-- people are, at best, servants or snacks), you could have narrated that the dragon seemed unimpressed (or have given the dragon dialogue to that effect).
  • If the roll were allowed and successful, the dragon could have become violently angry later in the conversation (also very red dragon-like), and a narration like "the dragon begins shifting around, seeming impatient and less and less impressed with your peaceful intentions" would remove the surprise factor when the dragon roasted the bard.

None of these is necessarily better than the others, or than other options not listed. Problems tend to crop up when crossing between different approaches, though. The surprise nullification of a high roll, and an explicit victory in a contested check, is almost certainly what the player took issue with.

Degrees of success can matter, if you choose

This is written out in the DMG, though I think it's tangential here. The issue seems to be that you were largely set on the PC failing with that particular approach, but the dice did not support that outcome. More options which are based on dice roll results are similarly flawed when you're aiming at a specific result. Those options are also less helpful when a PC is statted to have a bonus like +13: they will simply get high roll results often when using that skill.

Bonus section: non-combat encounters don't have to be flat and boring

You can make conversation-based, adversarial encounters richer and more interesting than a single roll, if you'd like. This can be less formal (like a series of checks based around overcoming each of an NPC's objections, for example) or more formal (a full-blown conversational combat system, complete with mechanics, "hit points", rounds, and so on).

The former is standard D&D, while the latter is a homebrew situation. Even if your players love a non-traditional playstyle, D&D can suit that style.


In DnD, as the GM, you decide when the dice are rolled, and what is at stake when the roll is made. This is doubly true in situations that are not covered by formal rules, such as the limits of skill checks to impact the world. However, you are playing the game for everyone's enjoyment, so you also have the responsibility to make the decisions in a way that works for your group.

This may mean the dragon is subject to persuasion, possibly regardless of what the Monster Manual or similar sources say. Some players are knowledgeable and passionate about monster lore being played "correctly" while others are unaware or uninterested in following creature stereotypes. Many players appreciate having an alternative option to deal with enemies, often a "pacifist route".

It might also mean the dragon is not subject to persuasion. Perhaps the players just want a fight – I know I do, in many cases, and I would honestly be annoyed if a promising boss fight or even the tense build-up to it was ruined by someone insisting on lengthy back-and-fro negotiation. For in-universe reasons, the dragon might indeed be an arrogant warlord type who doesn't care about lesser species trying to talk to it.

The bottom line? The GM decides, but should do it such that the group is happy with the decision.

Defer to the players

What I would do personally is ask my players what they wanted to see next. Do they want to see a diplomatic conclusion or at least a hopeful attempt at one? Then let there be rolls, and improvise based on the results. Do they want a tense boss fight? Then no need to roll, the negotiations are doomed to fail.

  • 2
    \$\begingroup\$ "The GM decides, but should do it such that the group is happy with the decision": I think that both GM and the party have to be happy with the decision. Moreover, I think that the GM's decision have to be realistic: referring to this particular situation, the bard could try to deceive the dragon, but is it likely that it happens? \$\endgroup\$
    – Eddymage
    Commented Aug 27, 2020 at 7:47
  • \$\begingroup\$ @Eddymage The GM is a part of the group. As for realism... it's a dragon, in a game heavily influenced by genre fiction. I honestly don't see the problem. \$\endgroup\$
    – kviiri
    Commented Aug 27, 2020 at 7:55
  • \$\begingroup\$ Ok, sorry, I misunderstood, I thought that with "group" you meant "party". I used the word realistic, but what I meant is "coherent" in the fantasy-world setting. But that's OT. \$\endgroup\$
    – Eddymage
    Commented Aug 27, 2020 at 7:59

You could have just made it a DC 30 check, it didnt have to be a contested check at all, it being a contested check was unnecessary. Either that or made it a DC 40 check, so it was sure it wouldn't have happened.

On the other hand...

You maybe should instead of using rolls, you make it fully about their argument for why they aren't aggressive. This makes it so your player/s will think more about their approach next time that they go to do something. Make sure that they are aware that if theres a BBEG nearby, that they don't think that they can cheap out of an encounter through a single contest.

Again, there was no reason to make it a contest, a check would have been better.

At least that's how I see it.

  • 2
    \$\begingroup\$ This is a good point, contested checks are never a good idea against NPC characters because they are not designed for it \$\endgroup\$
    – SeriousBri
    Commented Aug 27, 2020 at 9:28
  • 4
    \$\begingroup\$ @SeriousBri only if you are not designing them well. if I build a merchant or politician NPC they have pretty strong insight, deception and/or persuasion. if a roll is impossible just say it is impossible don't waste the players time. \$\endgroup\$
    – John
    Commented Aug 27, 2020 at 11:49
  • 1
    \$\begingroup\$ @john I mean the book ones, like said dragon. \$\endgroup\$
    – SeriousBri
    Commented Aug 27, 2020 at 13:32

Ask yourself, 'How did winning the contest help the player?' One of the main reasons dragons have such a high CR is their ability to fly. They can easily fly into range of the party, attack, and fly out of range in a single round. If the player never tried rolling at all, then the dragon taking flight would be completely expected. Maybe the dragon was planning on immediately frying the player with its breath weapon, so taking flight is preferable, but either option is certainly an aggressive act a reasonable dragon could take in that situation. What is important is that the player be rewarded for succeeding. That does not mean the reward has to be unreasonably huge. Imagine if after the player won the contest, you said:

'The dragon breaths in gallons of air, preparing to unleash a wave of fire/frost/whatever upon you. But for the briefest of moments, the air catches in his throat. He instead bellows in a voice that has shattered spirits, "You expect me to believe your word!" How do you respond?

The player has been rewarded in 3 very minor ways. First, he has gained the opportunity to talk his way out of the encounter, where before his only option was to fight. Even if there is no way he can succeed outside of a natural 20, at least he now has that chance. Second, the knowledge that the dragon tends to start combat with his breath weapon is very valuable. Even if he must ultimately fight his way out, he now knows that finding cover should be a priority. Third, the player may be able to capitalize on the dragon's hesitation. This can mechanically be represented as a +2 bonus to his initiative roll, or being able to act in the surprize round if you are feeling generous.

These rewards are incredibly small. The first has only a 5% chance of paying dividend, and most players would probably not even recognize the latter 2. What is important is that a reward for success exists; the size of the reward is almost irrelevant.

P.S. I would have made the Persuasion role a DC30 check. Contests, where both sides role, are inherently more likely to favor the player because they can be won by the player rolling well or the opposition rolling poorly. DC checks on the other hand can only be won by the former. This is just a very minor mistake, but I felt it was worth mentioning. I would have left this last point as just a comment, but my reputation is not high enough yet.

P.P.S. I have noticed a lot of people saying it is wrong for the player to ask for a contest instead of a DC check. I respectfully disagree. It is completely appropriate for a player to ask for a contest instead of a DC check or vice versa. Even asking for a saving throw would be okay, although that is not particularly relevant to this situation. The counter balance to this is that the DM is completely within their rights to deny such a request. In addition, it would be wildly inappropriate for a player to argue against any ruling by the DM, beyond a sentence or two to plead their case.

  • 3
    \$\begingroup\$ "Contests, where both sides role, are inherently more likely to favor the player because they can be won by the player rolling well or the opposition rolling poorly." I feel like i'm missing something obvious here, but can't they also be lost by either the player rolling poorly or the opposition rolling well? The distribution of (d20 - d20) is symmetrical. \$\endgroup\$
    – Mark Wells
    Commented Aug 28, 2020 at 6:54
  • \$\begingroup\$ And, welcome to RPG.SE. Please take the tour when you have time, and thanks for providing a detailed and thoughtful answer. \$\endgroup\$
    – Mark Wells
    Commented Aug 28, 2020 at 6:54
  • \$\begingroup\$ @Mark Wells If a win was worth 1 point, a lose worth -1 points, and tie worth 0 points, you would be correct. However, when a contest that ends in a tie the status quo is preserved. In most situations this is beneficial to the players, so a tie would actually have a value >0. \$\endgroup\$
    – E Tam
    Commented Aug 31, 2020 at 5:46

I posit that a red dragon would still dump fire on the head of someone coming before it unbidden and unwelcomed in a time the dragon's home is being assaulted even if the dragon believed the target was coming in peace.

I think giving the bard the round was gracious above and beyond the call of properly GMing an evil besieged red dragon!

Persuasion does not deflect a charging bull or an irritated angry red dragon. I would say a red dragon in the situation you describe would probably eat one of its top advisors should they present themselves unbidden; intention be darned!

Persuasion could avert a BBEG Battle...when that aversion is possible.


You must log in to answer this question.

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