D&D 5E Save Our Villians

[FormerlyHemlock]

That sounds like a perfectly rational approach for the villain to take to me.

One issue with D&D generally is that, as you discovered, there isn't a good framework in place for a DM to manage actions which take out of the PCs' sight. On the one hand, there's generally no way for PCs themselves to know what is happening out of their visual radius (unless it is really loud). On the other hand, it's kind of un-fun for the DM and the players if the crucial decisions are all being made by the DM without player involvement, and the players only get to see the end result of "after ten minutes of relative quiet, the three invisible mind flayers who have been sneaking up on you for the past five minutes Mind Blast everybody, and then four wights charge in the door while ten hobgoblins pour in fire with bows." Since the players didn't participate at all in constructing the ambush (except for taking ten minutes) you run the risk of it feeling like unfair metagaming on the DM's part.

I like your approach of controlling the whole complex, leaving perimeter forces intact while interior forces mass to counterattack. Another approach that I've used is to simply tell the players (not PCs) that I'll be rolling randomly to determine how many enemies arrive on any given round, out of the total number of enemies in the area. (E.g. out of 24 umber hulks, if I don't know what part of the neogi ship they're in, I'll say that "1d4 umber hulks arrive per turn".) This keeps the focus on the PCs, including my mental focus as DM, and gives them a measure of predictability at the metagame level while still keeping the pressure up. It's inferior to the detailed approach you used, but it's cheaper mentally for me and so I've used it more often. If I had better tools for managing a whole dungeon at once I would use the detailed approach more often.

 

[Uller]

The random approach may actually be a _more_ realistic approach than the DM orchestrating the actions of NPCs off camera. I do usually let my players know what was happening off camera after it is no longer relevant...mainly because I want them to understand why things were happening as they were and as far as I'm concerned, if the players don't know about it, it didn't really happen...so I think of it like watching the extras or the directory commentary on a DVD after you watch the movie...so you can understand why things happened.

Anyway...in a real life crisis, human beings don't behave rationally and with perfect knowledge. Even ones that have the information they need to act often act in ways that are unpredictable and even counter to their interests or the interests of those around them. The best example I can think of is I saw a video years ago of a gas station cashier continuing to wait on customers (and customers standing in line) as the magazine rack burst into flames. Some people fled the store. Some looked at the flames and got in line. The cashier just kept on taking money...

Many many years ago I was tasked with guarding an empty tent in a park in the middle of the night, with no weapons and no way to communicate. I resolved that if anyone came to take the tent I would simply let them and simply walk back to my barracks and tell someone in the morning....

So some randomness in what NPCs do, especially in the absence of leader, probably is the most sensible thing to do (and maybe easier).

 

[FormerlyHemlock]

The random approach may actually be a _more_ realistic approach than the DM orchestrating the actions of NPCs off camera. I do usually let my players know what was happening off camera after it is no longer relevant...mainly because I want them to understand why things were happening as they were and as far as I'm concerned, if the players don't know about it, it didn't really happen...so I think of it like watching the extras or the directory commentary on a DVD after you watch the movie...so you can understand why things happened.

Anyway...in a real life crisis, human beings don't behave rationally and with perfect knowledge. Even ones that have the information they need to act often act in ways that are unpredictable and even counter to their interests or the interests of those around them. The best example I can think of is I saw a video years ago of a gas station cashier continuing to wait on customers (and customers standing in line) as the magazine rack burst into flames. Some people fled the store. Some looked at the flames and got in line. The cashier just kept on taking money...

Many many years ago I was tasked with guarding an empty tent in a park in the middle of the night, with no weapons and no way to communicate. I resolved that if anyone came to take the tent I would simply let them and simply walk back to my barracks and tell someone in the morning....

So some randomness in what NPCs do, especially in the absence of leader, probably is the most sensible thing to do (and maybe easier).

Sure, and so when I talk about tools for orchestrating NPCs offscreen I intend to include a measure of randomness. Something like, "When this NPC sends the alarm to this other NPC, it takes him 1d6 rounds to get everyone within shouting distance to understand the situation, and then he moves on while the guys he just alerted strap on their armor and shields and grab their weapons (depends on NPC equipment--may take up to 10 minutes for heavy armor) and then moves towards a point of contact identified by the first NPC, which may or may not be where PCs currently are, but if they hear anything on the way over they may redirect to the new point of conflict. That is, assuming they are guards/combatants, otherwise they just keep their heads down and/or flee." That's precisely why I need tools to use that approach, because it's way too much information to keep in my head.

I really like your idea of a "director's cut" version that you show the players afterwards. I think if I spliced the player's timeline into my tool's version of what is happening ("this is what was happening when you guys finished casting Animate Dead on all ten corpses, after five minutes; you can see that five wights have already arrived and so has the wizard, and all of them rolled at least 16 on their Stealth so that's why you didn't hear them") I would feel a lot less shy about giving the players nasty surprises from off-camera, because I could show them the plays afterward. The amount of work it would take to get there is daunting but it's a great idea to consider.

 

[Psikerlord#]

I think the easiest way to keep BBEGs alive is to add more monsters. Lots of lower level fodder to keep the party from getting at the BBEG. But if your party is the super stealthy focus fire type, you might not be able to avoid those kinds of ambushes too easily. Lots of closed or locked doors defeats familiar scouts in my experience. Not sure about shadow monks!

I use the legendary template for every "solos", and also allow legendary resistance to give a save when normally there is none (eg against sleep for example).

Honestly over the years I have found the best bad guys happen naturally. From time to time a bad guy will escape, partly just by luck. This is the guy who becomes a recurring baddy. Don't plan too far ahead. The party will earn a nemesis in time. Don't try to force it (too much).

 

[FormerlyHemlock]

It just occurred to me that you can do the "director's cut" thing without any technology at all. Here's how:

As the players are sitting there after the combat, talking about what they're going to do next, roll a d6. If the d6 comes up a six, toss down an index card face-down which says ("3 Mind flayers arrive and stage outside"). The players know something is going on but don't know what. If they move out then, they'll discover that there are three mind flayers outside and they can fight them--but if the players keep arguing with each other, as players are wont to do, once per real-time minute or so roll a d6 again, and each time it comes up 6 toss down another index card ("Wizard arrives, casts Invisibility IV on the three Mind Flayers"; "4 Wights arrive"; "A dozen hobgoblins arrive"; "Offensive begins").

I've used at least the roll-a-die part of this before to put pressure on players in time-sensitive situations without completely requiring a table time = game time equivalence, and it seemed to work well, in that when they ran out of time (I rolled a 20 on the d20) the players were really bummed about being too late, but in a way that didn't make me feel like I'd just arbitrarily killed the PC and without any wheedling by the players for extra time. One player kind of blamed another for taking too long to decide on a course of action, which is pretty much the reaction you'd hope for in that situation--the world progresses even while you're standing there thinking, it's not just DM fiat. (He got over the blame thing though, don't worry.) Maybe the die roll wasn't necessary, maybe they would have felt the same kind of tension if I'd just been like, "Hurry up guys... okay, time's up, he's dead now" but it felt like leaving it partly up to chance helped the players feel like the outcome was fair.

Anyway, I think having four index cards on the table that the players can turn over, detailing what has been happening for the past ten minutes, would probably feel more fair than me just narrating it. The index cards are a physical artifact in the real world, and they've been accumulating there on the table for ten minutes... totally not arbitrary.

I'm going to try to find a way to work this technique into my next session that features some kind of dungeon or other time-sensitive environment.

 

[Rhenny]

Lots of good advice in this thread. Here's some more.

If you want a villain to get away, give it spellcasting levels for movement spells (expeditious retreat, dimension door, plane shift, teleport, etc.) or scroll use/magic items to do the same. I ran Ghosts of Dragonspear Castle which features a rakshasa who coordinates minions, deceives the party, uses the party to retrieve magical items/relics. The villain stole the first relic from right under the PCs noses and then used plane shift to get away. It helps that the rakshasa is immune to non-magical weapons, and spells 6th level and under. That's some awesome defense.

With spellcaster PCs I've run over the many years I've played, I always develop escape plans for when the going gets tough. It only serves that powerful villains do the same.

As a side note, I've actually found that my preferred level of difficulty/style of play as a DM has become the 3 PC party. For me, 3 is the magic number for danger, speed and fluidity.