Hey all! So I'm gonna be running a game for a few friends of mine, but I've run into a snag. For one of the contracts that this party will be fulfilling they have to kill a former member of one of their guilds, and I was hoping to run it as a DMPC since they will be travelling with this character for a while and it is going to be a bit of a betrayal. I want this fight to be fairly challenging, but the party is still pretty low level, 3 players level 2 or 3 at this point.
So my question is this, how many levels should I give this character so that it is a challenge, but won't curb stomp my players accidently? Any feedback would be appreciated as I've never run an encounter like this before. Thanks all!
It is hard to tell. I would think around level 8 or 9, however, if you build this character for combat and your party hasn't built their characters for it, then they could easily lose. If, however, your party are combat monsters then they might tear through a level 8 or 9 that isn't properly built. Magical items also change this.
Do you want to know the real trick to making boss fights challenging? Having phases. This means, if phase one is too simple, you ramp up the difficulty in phase two. Maybe in phase two they drink a position which gives them a special ability or something. If your players are struggling a bit, then you simply don't use the next phase.
What happens if they don't want to kill their former travelling companion?
Don't do levels, give them a statblock. Figure out roughly what CR (probably 4-5, depending on what level you choose for the PCs) you need them to be to challenge the players and build it out as a custom monster. Characters built as PCs are difficult to balance as encounters because the game just wasn't designed for it.
Level 8 or 9 would mince through 3 x level 3 players in a few turns. They'd zero a character every turn. A level 9 spellcaster will wipe the whole party in one spell.
Volo's Guide to Monster has lots of decent NPC statblocks. A good encounter for your party would be:
It is hard to tell. I would think around level 8 or 9, however, if you build this character for combat and your party hasn't built their characters for it, then they could easily lose. If, however, your party are combat monsters then they might tear through a level 8 or 9 that isn't properly built. Magical items also change this.
Do you want to know the real trick to making boss fights challenging? Having phases. This means, if phase one is too simple, you ramp up the difficulty in phase two. Maybe in phase two they drink a position which gives them a special ability or something. If your players are struggling a bit, then you simply don't use the next phase.
What happens if they don't want to kill their former travelling companion?
Thanks for the advice, I think phases could work really well now that I think about it.
Its an escort mission, where this guy is trying to capture the priestess in order to try and force her to lift his curse. The mission comes to this after a couple instances of hired thugs fail to kill the party essentially, forcing him to take matters into his own hands and blow his cover as their ally. If they lose the fight I plan to take the failing forward model, and turn it into a rescue arc where the party can gain influence with the church by saving her. If they win and don't want to kill him, I'll be honest I'm still working out the details here, but I'll turn him into a long term antagonist that they will run into every now and then, and I like the sound of a redemption arc for him, which would end up lifting his curse.
At this point I'm not putting in a way to avoid the fight necessarily, but I'm open to suggestions here as well.
Don't do levels, give them a statblock. Figure out roughly what CR (probably 4-5, depending on what level you choose for the PCs) you need them to be to challenge the players and build it out as a custom monster. Characters built as PCs are difficult to balance as encounters because the game just wasn't designed for it.
I was considering a statblock as well, I might still go with that, but I wanted to play around with the idea of pc levels to see if it would work.
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
I too vote for Statblock. Way easier to calculate. Also, be careful if you allow them to fight with magical gear or weapons as you are handing it to the party. Potions, on the other hand, are an excellent way to mitigate what the party can do and help an NPC survive a little longer or be more challenging in a fight... and when the fight is over, the party doesn't get the potion.
Potions, on the other hand, are an excellent way to mitigate what the party can do and help an NPC survive a little longer or be more challenging in a fight... and when the fight is over, the party doesn't get the potion.
Really good idea here. 👆 Potion of invisibility, plus potion of fly can be a great escape plan.
Yeah, it's like there's almost a reason you can't dump a character sheet into the encounter builder, like it's almost reflective of a game design principle or something.
First, while you use the term DMPC, it sounds like you're talking about functionally an NPC. A DMPC is a character the DM makes to play along with the party, level up etc. I don't see any real game functional use of the concept outside of DM vanity wanting to be part of the game (not talking about your approach just the general concept of a DMPC).
You already have a narrative planned for this character which makes him as a game device an NPC. Really, it's actually counterproductive to design an NPC through the constraints of the player's handbook. There's no rule in the game saying humanoid antagonists or allies of the PCs are required to be capable of reverse engineering through the PHB (that way actually leads to unnecessary meta). Take any of the NPC types in the PHB and add the features whether it be spell casting, martial options, sneakiness, or some combination of them all, until you have a suitable ally-to-challenge for the party. Sure the character maker is fun to play with, but for something like a BBEG it's often frustration to challenge building that the DM need not impose on themselves, you're binding yourself to unnecessary elements to put limitation on PCs.
A full blown character made in the maker isn't essentially "better" for a game than an embellished stat block. It's definitely less efficient design. Look at how challenging NPCs in the published adventures are presented ... you don't see class/subclass/levels for a reason ... it doesn't benefit game design.
Unfortunately this just doesn't really work well in 5e. PCs are glass cannons. If you give the DMPC a high enough level to survive more than one round of combat, then it will be one-shotting PCs each round.
One option is to create them as a reasonably low-level PC, maybe a little higher than the party, but then just give them a bunch of extra hp or lots of resistances.
Another is to give them minions. Maybe the minions attack the party and the DMPC turns on the players during the second or third round. Oooh, very dramatic. Or if this doesn't work for your plot, give the DMPC a summoning ability.
The last option is the polymorphed / disguised monster / final form option. The DMPC fights the PCs, but when defeated, you replace them with a standard monster stat block.
In any case, there's no easy rule for deciding how powerful to make them. The best official guidance I can give you is to check out the tables for calculating CR in the DMG. Even you create the character as a PC, you can use the rules about its average damage, hp, AC, and to hit bonus to calculate its CR as if it were a monster. Unfortunately CR is not a very intuitive or useful number either, but at least it might give you a way to compare against other fights that have been appropriately balanced for your players.
Guidance for creating unique humanoid enemies is unfortunately a gap in the 5e DMG. You kind of have to figure it out for yourself.
A Dungeon Master run Player Character is almost never a good idea. In particular if this is one the DM has used in other games. They come fully equipped with the best non-magical gear as well as whatever magic items their DM let them get their hands on. Just magical armor or weapons can change the nature of a low level game, and depending on just which magic items are around, can change even high level games. Multi-Classing and Feats also change the very nature of the game.
Betrayal is difficult to deal with. It inevitably leads to hard feelings, both for the characters involved and for the players against their DM. It hurts to be betrayed.
The DMPC would be constantly having to use Deception to cover their true motivations, and they are very unlikely to share the general Alignment of the party.
The kind of person who betrays others expects others to act just as they do, and will never trust anyone. They are incapable of seeing anyone as anything other than a potential enemy waiting to strike.
I think you can play a betrayal arc in a fun way without "cheating". Yes, include a few Deception checks. Do it every time the betrayer tells an out-and-out lie. Minimize how many times that happens. Roll Deception secretly (so secretly players don't know you're rolling: like have a table of pre-rolled d20s) and compare against passive Insight. Obviously give your bad guy a high Deception modifier, so they mass most checks with flying colors. But occasionally they fail as compared to your PCs highest one or two passive Insights. Tell these players secretly that they suspect the character of something, but give them strict instructions not to reveal that they know this as a result of Insight. Other players might suspect them of being paranoid, and ask them to prove their suspicions.
As a result, your deception might not play out as you had planned. But that's the fun of D&D. It's not a railroad. Anything can happen.
Thank you all for the feedback, this is why I love this community! You've given me a ton to think about, and I have a lot of new ideas for this character. I think after looking at it, I will go with a statblock after all, just wanted to play around with the idea. Thanks y'all!
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
You don't need to worry about this.
Social checks like Persuasion, Deception and Intimidation are there for the PCs to make. To run the best game you can, never, ever role checks for these type of things against the PCs. It doesn't even make sense.
Insight is the check that the PCs get to ask for, if they think that they are being deceived. They can attempt to check the veracity of someone's statement by making an insight check. You should never roll a Deception check for an NPC against PCs unless they call for an insight check. It screws player agency if you do:
"DM: The vampire with the blood all over his face is telling you that he didn't bite that shopkeeper." DM rolls Deception against passive insight. "You believe him."
Player: "But I don't."
DM: "No, you do."
No. Not ever. PCs always get to be in control of what they believe, and what they disbelieve. If you want to feed in clues to cause the PCs to distrust someone then you do so through RP. In a recent game my PCs began to suspect an NPC who was supposed to be helping in defence of a town, who wore all white, and after hours of fighting was still immaculate. They were totally right, and he betrayed them! And it came through logic and player agency - not through making a roll.
"The man opposite you says you're welcome in town." DM rolls deception. "You are not welcome in town."
I tell players who roll Insight checks that, "They seem to be telling the truth," whether they fail with a high roll or succeed. If they fail with a low roll, I say, "You're not sure." I don't tell them what they believe. If they choose to rely on a high roll, that's up to them. If they wish to remain suspicious, that's fine too.
In any case, failing a passive Insight check won't lead to telling someone they believe someone. You just don't say anything.
Not using passive Insight means the players are, to some extent, relying on how good a liar the DM is. Once they ask for the check, it depends on the NPC's Deception roll, but if you're a good liar they'll never ask. It's the same problem as never asking for players to make Persuasion checks. Their charisma stat becomes irrelevant, and its up to the player's charisma whether the DM will be persuaded, not up to the PC's charisma whether the NPC will be persuaded.
You could secretly roll Deception and role play the outcome, either lying well or badly. But for me at least, acting a bad lie is beyond my skill level. It will either be as cold-blooded as I can muster, or obviously transparent. What I suspect some people do is just consistently play high-Deception characters as consistently good liars, and low-Deception characters as obvious bumbling fools. But that doesn't account for the fact that spending a lot of time with an NPC should give you a lot of chances for them to slip up.
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
One way to do it is arrange for the NPC to not realize the betrayal will happen.
Set things up so that decision or realization comes later.
A high passive -Insight character, say 15, is still going to catch most deception about 1/3 to 1/2 the time. Safer for the NPC not to know.
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
You don't need to worry about this.
Social checks like Persuasion, Deception and Intimidation are there for the PCs to make. To run the best game you can, never, ever role checks for these type of things against the PCs. It doesn't even make sense.
Insight is the check that the PCs get to ask for, if they think that they are being deceived. They can attempt to check the veracity of someone's statement by making an insight check. You should never roll a Deception check for an NPC against PCs unless they call for an insight check. It screws player agency if you do:
"DM: The vampire with the blood all over his face is telling you that he didn't bite that shopkeeper." DM rolls Deception against passive insight. "You believe him."
Player: "But I don't."
DM: "No, you do."
No. Not ever. PCs always get to be in control of what they believe, and what they disbelieve. If you want to feed in clues to cause the PCs to distrust someone then you do so through RP. In a recent game my PCs began to suspect an NPC who was supposed to be helping in defence of a town, who wore all white, and after hours of fighting was still immaculate. They were totally right, and he betrayed them! And it came through logic and player agency - not through making a roll.
"The man opposite you says you're welcome in town." DM rolls deception. "You are not welcome in town."
Great RP right there.
Failing an insight check means you can't tell if the other party is lying, it doesn't mean you believe them. Passive insight should be used any time NPCs are lying to the PCs, otherwise it slows the game down immensely with players asking for insight checks every time anyone tells them anything.
Failing an insight check means you can't tell if the other party is lying, it doesn't mean you believe them. Passive insight should be used any time NPCs are lying to the PCs, otherwise it slows the game down immensely with players asking for insight checks every time anyone tells them anything.
Using passive insight means that the DM needs to make a deception roll every time an NPC lies to the PCs. It also leads to it being completely impossible for any NPC to maintain any kind of pretence without a massively high deception score. It's like requiring 5 PCs to all make DC15 stealth check without group stealth, and making one failure a fail for the group: even with +10 stealth for all PCs, one of them will probably fail. If you make constant rolls, you will get 1s and 2s, practically guaranteeing failure.
Let's assume a PC with a modest +1 Wisdom modifier, and any NPC with +1 to Charisma (Deception) checks. The NPC is likely to get busted automatically, with no decision making or thought by the PC, the either the first or second time they say something untrue. They will rarely manage 4 lies. 8 lies is highly unlikely. Even if you use some kind of "once per few minutes" system over individual lies, to do something as simple as pretend to be an ordinary wagon driver and take the PCs on a simple trip along the road, the PCs will undoubtedly figure out that they're being led into a trap within a few minutes. They won't know how they know this, they won't know why it's a lie, but they'll suddenly change their course of action for no reason. If you start adding bonuses for how convincing the lie is, you might as well auto pass them, ergo no check.
If you use this, essentially NPCs just can't deceive PCs, no matter how plausible the situation or deception is. Moreover, when the castellan says "Come right this way, the master is expecting you for dinner" and the PCs are utterly unaware as he leads them into a trap, you'd actually have the PCs auto-detect that it's a trap? And roll a Deception check for the NPC for just saying it? That approach to the game is baffling to me.
DM'ing can take a lot of skill in acting, clue dropping and you'll have a better game if you consider "how will this lying NPC mess up?"
If NPCs can use Deception against PCs, why not Persuasion?
In terms of slow down, I've never encountered players constantly asking for insight checks. If they do and it's weird, ask them why they want to make the roll. No reasonable reason? No roll, just like any other ability check.
Failing an insight check means you can't tell if the other party is lying, it doesn't mean you believe them. Passive insight should be used any time NPCs are lying to the PCs, otherwise it slows the game down immensely with players asking for insight checks every time anyone tells them anything.
Let's assume a PC with a modest +1 Wisdom modifier, and any NPC with +1 to Charisma (Deception) checks. The NPC is likely to get busted automatically, with no decision making or thought by the PC, the either the first or second time they say something untrue.
Sure, such an NPC is a bad liar, and not the kind of person you'd use as a double agent. You want someone with at least like +8 - +10 deception. Even then, with 2 - 3 lies there's a good chance you get busted.
Let's face it, that's realistic. You can't just repeatedly lie to high-tier clerics without them getting suspicious.
The best way to get away with deception is to lie as little as possible. Try to get away with one lie and not need any more.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Hey all! So I'm gonna be running a game for a few friends of mine, but I've run into a snag. For one of the contracts that this party will be fulfilling they have to kill a former member of one of their guilds, and I was hoping to run it as a DMPC since they will be travelling with this character for a while and it is going to be a bit of a betrayal. I want this fight to be fairly challenging, but the party is still pretty low level, 3 players level 2 or 3 at this point.
So my question is this, how many levels should I give this character so that it is a challenge, but won't curb stomp my players accidently? Any feedback would be appreciated as I've never run an encounter like this before. Thanks all!
It is hard to tell. I would think around level 8 or 9, however, if you build this character for combat and your party hasn't built their characters for it, then they could easily lose. If, however, your party are combat monsters then they might tear through a level 8 or 9 that isn't properly built. Magical items also change this.
Do you want to know the real trick to making boss fights challenging? Having phases. This means, if phase one is too simple, you ramp up the difficulty in phase two. Maybe in phase two they drink a position which gives them a special ability or something. If your players are struggling a bit, then you simply don't use the next phase.
What happens if they don't want to kill their former travelling companion?
Don't do levels, give them a statblock. Figure out roughly what CR (probably 4-5, depending on what level you choose for the PCs) you need them to be to challenge the players and build it out as a custom monster. Characters built as PCs are difficult to balance as encounters because the game just wasn't designed for it.
Level 8 or 9 would mince through 3 x level 3 players in a few turns. They'd zero a character every turn. A level 9 spellcaster will wipe the whole party in one spell.
Volo's Guide to Monster has lots of decent NPC statblocks. A good encounter for your party would be:
Master Thief
warlock of the archfey
archer
hobgoblin devastator (change the race if needed)
Thanks for the advice, I think phases could work really well now that I think about it.
Its an escort mission, where this guy is trying to capture the priestess in order to try and force her to lift his curse. The mission comes to this after a couple instances of hired thugs fail to kill the party essentially, forcing him to take matters into his own hands and blow his cover as their ally. If they lose the fight I plan to take the failing forward model, and turn it into a rescue arc where the party can gain influence with the church by saving her. If they win and don't want to kill him, I'll be honest I'm still working out the details here, but I'll turn him into a long term antagonist that they will run into every now and then, and I like the sound of a redemption arc for him, which would end up lifting his curse.
At this point I'm not putting in a way to avoid the fight necessarily, but I'm open to suggestions here as well.
I was considering a statblock as well, I might still go with that, but I wanted to play around with the idea of pc levels to see if it would work.
Just a word of warning. Betrayals can be tricky to pull off. Technically, you should be making regular deception checks for the NPC to keep up their act. If you have a character with a high passive insight, for example, it’s player might start wondering why that character didn’t notice something was off. I’m not saying don’t do it, just noting that your players may feel cheated if the NPC success because of dm fiat and plot armor, as opposed to doing it by the rules.
I'll make sure to keep that in mind
I too vote for Statblock. Way easier to calculate. Also, be careful if you allow them to fight with magical gear or weapons as you are handing it to the party. Potions, on the other hand, are an excellent way to mitigate what the party can do and help an NPC survive a little longer or be more challenging in a fight... and when the fight is over, the party doesn't get the potion.
Really good idea here. 👆
Potion of invisibility, plus potion of fly can be a great escape plan.
Yeah, it's like there's almost a reason you can't dump a character sheet into the encounter builder, like it's almost reflective of a game design principle or something.
First, while you use the term DMPC, it sounds like you're talking about functionally an NPC. A DMPC is a character the DM makes to play along with the party, level up etc. I don't see any real game functional use of the concept outside of DM vanity wanting to be part of the game (not talking about your approach just the general concept of a DMPC).
You already have a narrative planned for this character which makes him as a game device an NPC. Really, it's actually counterproductive to design an NPC through the constraints of the player's handbook. There's no rule in the game saying humanoid antagonists or allies of the PCs are required to be capable of reverse engineering through the PHB (that way actually leads to unnecessary meta). Take any of the NPC types in the PHB and add the features whether it be spell casting, martial options, sneakiness, or some combination of them all, until you have a suitable ally-to-challenge for the party. Sure the character maker is fun to play with, but for something like a BBEG it's often frustration to challenge building that the DM need not impose on themselves, you're binding yourself to unnecessary elements to put limitation on PCs.
A full blown character made in the maker isn't essentially "better" for a game than an embellished stat block. It's definitely less efficient design. Look at how challenging NPCs in the published adventures are presented ... you don't see class/subclass/levels for a reason ... it doesn't benefit game design.
Jander Sunstar is the thinking person's Drizzt, fight me.
Unfortunately this just doesn't really work well in 5e. PCs are glass cannons. If you give the DMPC a high enough level to survive more than one round of combat, then it will be one-shotting PCs each round.
One option is to create them as a reasonably low-level PC, maybe a little higher than the party, but then just give them a bunch of extra hp or lots of resistances.
Another is to give them minions. Maybe the minions attack the party and the DMPC turns on the players during the second or third round. Oooh, very dramatic. Or if this doesn't work for your plot, give the DMPC a summoning ability.
The last option is the polymorphed / disguised monster / final form option. The DMPC fights the PCs, but when defeated, you replace them with a standard monster stat block.
In any case, there's no easy rule for deciding how powerful to make them. The best official guidance I can give you is to check out the tables for calculating CR in the DMG. Even you create the character as a PC, you can use the rules about its average damage, hp, AC, and to hit bonus to calculate its CR as if it were a monster. Unfortunately CR is not a very intuitive or useful number either, but at least it might give you a way to compare against other fights that have been appropriately balanced for your players.
Guidance for creating unique humanoid enemies is unfortunately a gap in the 5e DMG. You kind of have to figure it out for yourself.
A Dungeon Master run Player Character is almost never a good idea. In particular if this is one the DM has used in other games. They come fully equipped with the best non-magical gear as well as whatever magic items their DM let them get their hands on. Just magical armor or weapons can change the nature of a low level game, and depending on just which magic items are around, can change even high level games. Multi-Classing and Feats also change the very nature of the game.
Betrayal is difficult to deal with. It inevitably leads to hard feelings, both for the characters involved and for the players against their DM. It hurts to be betrayed.
The DMPC would be constantly having to use Deception to cover their true motivations, and they are very unlikely to share the general Alignment of the party.
The kind of person who betrays others expects others to act just as they do, and will never trust anyone. They are incapable of seeing anyone as anything other than a potential enemy waiting to strike.
<Insert clever signature here>
I think you can play a betrayal arc in a fun way without "cheating". Yes, include a few Deception checks. Do it every time the betrayer tells an out-and-out lie. Minimize how many times that happens. Roll Deception secretly (so secretly players don't know you're rolling: like have a table of pre-rolled d20s) and compare against passive Insight. Obviously give your bad guy a high Deception modifier, so they mass most checks with flying colors. But occasionally they fail as compared to your PCs highest one or two passive Insights. Tell these players secretly that they suspect the character of something, but give them strict instructions not to reveal that they know this as a result of Insight. Other players might suspect them of being paranoid, and ask them to prove their suspicions.
As a result, your deception might not play out as you had planned. But that's the fun of D&D. It's not a railroad. Anything can happen.
Thank you all for the feedback, this is why I love this community! You've given me a ton to think about, and I have a lot of new ideas for this character. I think after looking at it, I will go with a statblock after all, just wanted to play around with the idea. Thanks y'all!
You don't need to worry about this.
Social checks like Persuasion, Deception and Intimidation are there for the PCs to make. To run the best game you can, never, ever role checks for these type of things against the PCs. It doesn't even make sense.
Insight is the check that the PCs get to ask for, if they think that they are being deceived. They can attempt to check the veracity of someone's statement by making an insight check. You should never roll a Deception check for an NPC against PCs unless they call for an insight check. It screws player agency if you do:
No. Not ever. PCs always get to be in control of what they believe, and what they disbelieve. If you want to feed in clues to cause the PCs to distrust someone then you do so through RP. In a recent game my PCs began to suspect an NPC who was supposed to be helping in defence of a town, who wore all white, and after hours of fighting was still immaculate. They were totally right, and he betrayed them! And it came through logic and player agency - not through making a roll.
"The man opposite you says you're welcome in town." DM rolls deception. "You are not welcome in town."
Great RP right there.
I tell players who roll Insight checks that, "They seem to be telling the truth," whether they fail with a high roll or succeed. If they fail with a low roll, I say, "You're not sure." I don't tell them what they believe. If they choose to rely on a high roll, that's up to them. If they wish to remain suspicious, that's fine too.
In any case, failing a passive Insight check won't lead to telling someone they believe someone. You just don't say anything.
Not using passive Insight means the players are, to some extent, relying on how good a liar the DM is. Once they ask for the check, it depends on the NPC's Deception roll, but if you're a good liar they'll never ask. It's the same problem as never asking for players to make Persuasion checks. Their charisma stat becomes irrelevant, and its up to the player's charisma whether the DM will be persuaded, not up to the PC's charisma whether the NPC will be persuaded.
You could secretly roll Deception and role play the outcome, either lying well or badly. But for me at least, acting a bad lie is beyond my skill level. It will either be as cold-blooded as I can muster, or obviously transparent. What I suspect some people do is just consistently play high-Deception characters as consistently good liars, and low-Deception characters as obvious bumbling fools. But that doesn't account for the fact that spending a lot of time with an NPC should give you a lot of chances for them to slip up.
One way to do it is arrange for the NPC to not realize the betrayal will happen.
Set things up so that decision or realization comes later.
A high passive -Insight character, say 15, is still going to catch most deception about 1/3 to 1/2 the time. Safer for the NPC not to know.
Failing an insight check means you can't tell if the other party is lying, it doesn't mean you believe them. Passive insight should be used any time NPCs are lying to the PCs, otherwise it slows the game down immensely with players asking for insight checks every time anyone tells them anything.
Using passive insight means that the DM needs to make a deception roll every time an NPC lies to the PCs. It also leads to it being completely impossible for any NPC to maintain any kind of pretence without a massively high deception score. It's like requiring 5 PCs to all make DC15 stealth check without group stealth, and making one failure a fail for the group: even with +10 stealth for all PCs, one of them will probably fail. If you make constant rolls, you will get 1s and 2s, practically guaranteeing failure.
Let's assume a PC with a modest +1 Wisdom modifier, and any NPC with +1 to Charisma (Deception) checks. The NPC is likely to get busted automatically, with no decision making or thought by the PC, the either the first or second time they say something untrue. They will rarely manage 4 lies. 8 lies is highly unlikely. Even if you use some kind of "once per few minutes" system over individual lies, to do something as simple as pretend to be an ordinary wagon driver and take the PCs on a simple trip along the road, the PCs will undoubtedly figure out that they're being led into a trap within a few minutes. They won't know how they know this, they won't know why it's a lie, but they'll suddenly change their course of action for no reason. If you start adding bonuses for how convincing the lie is, you might as well auto pass them, ergo no check.
If you use this, essentially NPCs just can't deceive PCs, no matter how plausible the situation or deception is. Moreover, when the castellan says "Come right this way, the master is expecting you for dinner" and the PCs are utterly unaware as he leads them into a trap, you'd actually have the PCs auto-detect that it's a trap? And roll a Deception check for the NPC for just saying it? That approach to the game is baffling to me.
DM'ing can take a lot of skill in acting, clue dropping and you'll have a better game if you consider "how will this lying NPC mess up?"
If NPCs can use Deception against PCs, why not Persuasion?
In terms of slow down, I've never encountered players constantly asking for insight checks. If they do and it's weird, ask them why they want to make the roll. No reasonable reason? No roll, just like any other ability check.
Sure, such an NPC is a bad liar, and not the kind of person you'd use as a double agent. You want someone with at least like +8 - +10 deception. Even then, with 2 - 3 lies there's a good chance you get busted.
Let's face it, that's realistic. You can't just repeatedly lie to high-tier clerics without them getting suspicious.
The best way to get away with deception is to lie as little as possible. Try to get away with one lie and not need any more.