Roleplaying games use a lot of different initiative systems: individual, side-based, hot potato, fixed, freeform, and shot clocks, to name just a few. And these initiative systems can, in turn, interact with other combat mechanics in myriad ways — readying, delaying, interrupting, reacting, and such-like mechanics being common examples.

This means your best practices for handling initiative at the table will vary quite a bit depending on what system you’re running, but this particular tip is going to assume that you’re using a system in which a separate initiative check is made for each character and their actions taken in order from best to worst result. This is, of course, a fairly common arrangement, and most or all of the techniques I’ll be sharing here should be useful across common variations (like reverse declaration or rerolling initiative each round).

BATCHING INITIATIVE

Having individual initiative checks that set a specific turn sequence has a lot of advantages:

• Separate turns for each character create clarity in resolution and declaration.
• It removes the analysis paralysis of players trying to figure out who should go next.
• It allows the GM to use advanced techniques like on-decking players to speed up play.

On the GM’s side of the screen, however, it can begin creating big headaches as the number of NPCs in a fight grows. If twenty goblins led by an ogre and accompanied by a couple of earth elementals show up, for example, the GM not only needs to roll twenty-three initiative checks, they also have to keep track of them.

The solution is to batch initiative: Instead of rolling a separate initiative for every cyber-ninja in the fight, for example, you can instead roll a single initiative check for all of the cyber-ninjas and have them all take action at the same time.

Not only is this easier when it comes to rolling the initiative check, it also tends to speed up resolution. First, it lets you consult the cyber-ninja stat block once per round instead of needing to flip back to it multiple times per round. Second, it makes it a lot easier to use techniques like rolling fistfuls of dice that also speed up combat.

This technique is so useful, in fact, that it’s not unusual for RPG rulebooks featuring individual initiative checks to nevertheless recommend that the GM batch NPC initiatives. What’s oddly less common, however, is explaining exactly how you should batch the checks, particularly as enemy groups become larger and more diverse.

Here’s how I’ve learned to break it down:

First, start with the bosses/leaders. If a group has a leader, the leader gets a unique initiative, even if they’re using the same stat block as their minions. (This might also be a narrative lead rather than a diegetic one.)

Second, take all the enemies with the same stat block and group them together. This gives you the simple efficiency of the cyber-ninja.

Third, split large groups into smaller groups. This is really where the art comes into it. What constitutes a group that’s “too large” depends on you, the system you’re running, and the specific circumstances of the encounter. Generally speaking, if a group has more than five to eight NPCs in it, I’ll probably start thinking about splitting it up. I’ll sometimes make exceptions for large groups of mooks, but there are a few factors to consider:

• Action economy. From a balance standpoint, you don’t want too many bad guys ganging up on a single PC and beating them half to death before they have a chance to do anything about it. Sometimes that will happen naturally because the PCs rolled poorly on their initiative checks, but the risk increases substantially if you batch a dozen bad guys together. So split them up.
• Tactical interest. Flipping that around, fights also tend to be more interesting when your bad guys can respond to what the PCs are doing. If all the bad guys are clumped up, then you can also run into a situation where all the PCs are taking their actions without interruption. Since everyone at the table (including you!) is a fan of the PCs, this is generally less of problem. But breaking things up a bit can also make for a more interesting encounter.
• Complexity. Batching bad guys together will generally make you more efficient in running them, but — depending on the system you’re running and your comfort level with it — there will be a point where trying to deal with a large number of bad guys all at the same time becomes a liability. Pay attention to your experience here and learn where, in the current system, you need to break them up.
• Pacing. Along similar lines, no matter how quickly you’re resolving things, if you can feel your players checking out in the middle of resolving your thirteenth goblin attack, that’s a sign that you should be using smaller batches in future encounters.

VARIANT: SQUAD LEADERS

As a variant to this approach, instead of batching NPCs by stat block, there are times when it makes more sense to batch them according to tactical groups. (For example, you might have a mixed group of goblins and hobgoblins on one side of the PCs and a different mixed group on the other side.) You might also find this approach useful if you’re using adversary rosters and multiple action groups converge into a single encounter: There was a reason you grouped those baddies together into an action group in the first place, and that reason can often persist into combat. (Particularly if you have a group joining the fight after combat has begun.)

But if a mixed group has different initiative modifiers, how can you roll a collective initiative for them?

If you think of the mixed group as a squad, who’s the leader of that squad? This might be a formal or informal designation, but in either case you can simply roll initiative for the squad leader(s) and have their squad(s) take action with them.

TRACKING SPLIT BATCHES

If you have identical creatures in multiple batches, how can you keep track of which batch each creature belongs to? You don’t want Goblin A taking action as part of Batch 1 on initiative count 15 and then accidentally taking another full turn as part of Batch 2 on initiative count 8!

The first thing is to keep the batches clearly delineated in your notes. Don’t, for example, mix all the goblin hit point totals together into one big group. Keep the hit point totals for Batch 1 separate from the hit point totals for Batch 2.

If you’re running combat in the theatre of the mind, check out Random GM Tips: Visualizing Combat.

If you’re using a battlemap, there are several ways to can keep things straight in the chaos of the battlefield:

Different miniatures. Just because they’re all goblins doesn’t mean you couldn’t have one type of goblin miniature for the goblins in Batch 1 and a different goblin miniature for the goblins in Batch 2. This also works with other tokens, of course. For example, I’m a fan of using glass beads, of which I have several different colors. Each color can be used to indicate a separate batch.

If you’re painting your miniatures, it can be useful to keep this utility in mind. If you’re painting a dozen goblin miniatures, for example, you might make sure that six of them have red shirts and six of them have blue ones.

Tag tokens. If you only have one type of goblin miniature, you can still distinguish them by adding an additional tag. For example, I can also use my glass beads for this purpose, although it has the disadvantage of needing to move both miniature and bead at the same time.

You can also make discs of colored paper and attach them to the miniature’s base using blu-tack. Colored Post-It flags can also be effective.

Geographic separation. If you can keep the batches physically separate from each other (e.g., Batch 1 goblins are attacking the paladin; Batch 2 goblins are attacking the barbarian), then you may not need any additional indicator to keep them straight in your mind.

POST-INITIATIVE BATCHING

Whether you’re batching initiative checks or not, what you’ll discover is that initiative orders will often create de facto batches. For example, here’s the rolled initiative roster from one of my recent games, with PC names in italics:

18 Spider-Eaters
17 Chieftain
16 Tee
15 Harpoon Spider
14 Spider Goblins
14 Agnarr
12 Elestra
12 Nasira
8 Ranthir
6 Giant Spiders

In this case I’ve pre-batched the Spider-Eaters, Spider Goblins, and Giant Spiders by their stat blocks, rolling a single initiative for each group in addition to the initiative rolls for the Chieftain and single Harpoon Spider.

Now that I’ve rolled initiative, however, other batches naturally appear: The Spider-Eaters and Chieftain will both be taking action together before the PCs, and the Harpoon Spider and Spider Goblins will, similarly, be taking their actions together between PC turns. Notice, too, that the end of the round is also connected to the beginning of the round: At the end of the first round, the Giant Spiders, Spider-Eaters, and Chieftain will all become batched together after Ranthir takes his turn and before Tee takes her turn.

In some cases, for the sake of simplicity, you’ll still be better off ignoring this post-initiative batching and just resolving each group one at a time. But there are often times when I find that recognizing and using these de facto groups can be very useful. For example, if I’m looking ahead while Tee is making her attack rolls, it would make sense to not only grab two red d20’s for the harpoon spiders, but also five blue d20’s for the spider-goblins. (Then I can roll all seven dice at the same time.) This can also open up more interesting tactical options, such as having the harpoon spiders and spider-goblins working together.

Flipping things around a bit, this post-initiative batching also applies to the PCs! For example, here’s another initiative roster from the same adventure:

19 Elestra
18 Harpoon Spiders
15 Spider Goblins
12 Ranthir
8 Nasira
6 Tee
5 Agnarr

Here you can see that Elestra will go first, but after the harpoon spiders and spider goblins have taken their turns, the PCs will effectively all be batched together. (Remember that Elestra will get grouped in as the round loops.) What this means, in practice, is that the PCs are now free to take their actions in any order. (And the same thing is true for your NPCs.) This fight is now effectively using side-based initiative, and if you – and your players – recognize this, then you can take advantage of that to gain all the benefits of side-based initiative (e.g., making it easier for PCs to coordinate their actions on the battlefield).

There are a couple caveats with this, though. First, some systems – including, notably, D&D 5^th Edition – will resist this batching by locking down the initiative order. Most RPGs, however, will include some method for characters to change their position in the initiative order (e.g., the Delay action in D&D 3^rd Edition), which will allow this batching to just naturally flow from the mechanics.

Even if your RPG locks down the initiative order, though, I encourage you to embrace this post-initiative batching: It unlocks creative collaboration and enables multitasking that can greatly enhance your combats.

Your players, of course, may not recognize this opportunity. You can help them learn this technique, and also help prompt them by saying things like:

GM: Okay, as Tee ducks under the last of the swooping griffons, Agnarr, Ranthir, and Elestra are the next group. What are y’all doing?

GM: Agnarr’s up next, with Ranthir and Elestra on deck. Who wants to go first?

Usually, once they get the feel for this, the players will start naturally taking advantage of the batching. You won’t need to prompt them as much, particularly if the initiative order is visible to the whole group.

The other nice thing is that, while this post-initiative batching lets you gain many of the benefits of side-based initiative, it can also avoid some of the disadvantages because there’s still a default order you and the players can fall back on to keep things moving. It can often be the best of both worlds!

Sometimes, though, this won’t work out. Whether your system locks down the initiative order or not, you’ll find that some players adapt poorly to this. They’ll either resist embracing the freedom of the batch initiative, or they may even push back and say things like, “But Teresa has to go first!” There can be a lot of different reasons for this, but it’s okay. If you’ve got a group that’s more disrupted than helped by batching initiative, then this is also a situation where you can just fall back on the default order.

PATTERNS OF ACTION

The last thing to point out is that you can use post-initiative batching for the PCs even if they aren’t all bunched together. Looking at the first initiative roster above, for example, we can see the PC vs. NPC groupings:

• NPC: Giant Spiders / Spider-Eaters / Chieftain
• PC: Tee
• NPC Harpoon Spider / Spider-Goblins
• PC: Agnarr/ Elestra / Nasira / Ranthir

Looking at your initiative rosters through this lens of batching will make clear the pattern of action, and you’ll quickly learn that different fights will naturally have different patterns, creating different pacing. Side-based initiative has one feel to it. Staccato back-and-forths have another. One-half of the PCs being separated from the other by a single creature will feel differently from when there’s a large group or multiple groups separating them.

And if your RPG allows for initiative orders to shift, then these patterns of action can, of course, also change over the course of a fight. Even barring that, as various groups are eliminated or added to the fight the pattern will change.

Here we begin to delve pretty deep into the true art of running a roleplaying game: As you get a feel for these patterns of action and the pacing that naturally flows from them, you’ll also learn how to take advantage of them, leveraging them and enhancing them and building on them.

The batching of dice rolls we discussed earlier is a simple example of this, but another advanced technique is to wait until you’ve mechanically resolved multiple actions and then batch them into a single description. Instead of resolving the Andorian’s attack roll, describing how it misses Kirk, and then resolving Kirk’s phaser fire and describing how it hits, you can instead resolve both attacks mechanically and then:

GM: The Andorian makes a break for the door, whipping out his pistol and firing wildly in your direction. Ducking under the glittering beam, you return fire, striking him in the back. He collapses in a heap, the automatic doors hissing open to let him fall into the doorway.

This technique lets you weave the action together and opens up new descriptive palettes to keep your fights fresh and interesting.

Batched initiative also creates natural groupings for this batched description – e.g., Boromir being struck by a flurry of orcish arrows; a barbarian holding a door against the battery of a dozen spider goblins; a tactical team sweeping a room. (This is particularly true if you’re rolling all those arrow attacks and door-smashing checks at the same time in one big fistful of dice.)

Recognizing the mechanical beats of the fight can help you form and pace these batched action descriptions ­– either by taking advantage of them or playing against them. They’re a key that will let you begin exploring the art of running combat at a deeper level, beginning a whole new journey for you as a game master.

Mouser: Okay. Let’s get out of here. I open the door.

GM: When the Gray Mouser opens the door, where’s everyone standing?

Pippin: Ahhhhhhh!!!!!!

Merlin: As far away as possible!

Conan: Rogue go boom!

On the Alexandrian Discord, CorpCord asked, “How do you avoid revealing that a decision is important by framing it, when the characters wouldn’t know?” In other words, there are lots and lots of little, incidental things that our characters do in the game world — e.g., choosing exactly where to stand in a room; picking which tie to wear; deciding whether to drink the white or red wine — that we don’t typically call out or interrogate at the gaming table.

But if a situation arises where that information suddenly becomes vital — e.g., the trapped door is about to explode— the fact that the GM is suddenly asking for it makes it difficult or impossible for the players to make the choice as if they were their characters: Their characters don’t know that their choice of wine is vital (because it’s poisoned), but the player does. And that metagame knowledge will either influence their decision, affect their immersion in the scene, or both.

Let’s consider a few other examples:

• They’re walking down a hall and a trap goes off… what was the marching order?
• The pressure plate for the trap is in the middle of the hall… so are the PCs walking there or are they hugging the walls?
• They’re all back at Suzie’s “safe” house when someone throws a pipe bomb through the window… who was in the living room at that moment?
• The businessman they’re meeting is a close friend of Adlai Stevenson… so is the PC wearing their I Like Ike button?

When dealing with a situation like this we can start by asking ourselves how much benefit of the doubt we should give the players. (For example, is it reasonable to assume that the group is always cautiously backing off whenever the rogue is opening a suspicious door?) This is actually a subset of figuring out the threshold for player expertise triggering character expertise, as described in The Art of Rulings.

The answer here is situational. (Do we assume PCs are as cautious opening a door at the local tavern as they are opening a door in the dungeon?) But it’s also about achieving desired gameplay. I generally recommend leaning towards giving the players the benefit of the doubt (because capricious or unfair “Gotchas!” usually aren’t fun), but if we just assume that the PCs are always doing the optimal action — they always check the chests for traps; they would have obviously hired an NPC security team to keep watch outside their safe house; their character would clearly know to make polite inquiries about the political allegiance of the businessman — then the game starts playing itself, and that isn’t fun, either.

In other words, you need player expertise to activate character expertise. Which means, sometimes, you have to ask the question and tip your hand.

So what can we do about this? Let’s break it down.

First, specific declarations by the players should, obviously, always be respected. For example, if someone said they’re standing in the hall keeping a lookout for goblins, then they’re in the hall. None of the techniques below will cause them to suddenly be standing by the door when the explosion goes off (unless, of course, they make another declaration changing their location).

Next, if there’s a particular type of information that you’re constantly finding yourself wanting, you should set up a standard operating procedure with the players to provide it. Marching order in a dungeon is a good example of this, whether it’s determining who gets hit by a trap or where people are standing when the group is ambushed.

In some cases, you might have multiple SOPs and the players can indicate which one they’re currently using. For example, in an urban campaign what you’re wearing might be frequently important, so the group might be in Adventuring Gear or High Social — and, in each case, everyone at the table will know what outfit each PC is wearing.

If you don’t have an SOP, then one technique is to lay down a false trail. In other words, don’t ask the players where their characters are standing ONLY when the rogue is about to open a trapped door. Instead, occasionally ask them for “incidental” details like that when there’s no risk. This obfuscates the metagame knowledge being imparted by the question.

You can also strategically use this technique to build up a bunch of false tension… and then releasing it with a feeling of relief when nothing happens. (If you’re wondering why you’d want to do this, check out how horror movies us this technique.) Even better, this can also result in the players getting lulled into a false sense of security. (“Justin’s just trying to scare us again. You’re not fooling me this—” BOOM!)

You can further throw them off the trail by using these declarations to, instead of assessing danger, paint the scene. For example, you can ask everyone where they’re at or what they’re doing in the safe house and then use that information to give a little description of what everyone is doing. This can help to draw the players into the reality of the game world (by getting them to actually form a specific picture of what their character is doing)… while also setting them up for those moments when the same or similar question is determining where they’re at when the pipe bomb goes off.

Another effective technique is to vary the right answer when asking these hand-tipping questions. Instead of a trapped door exploding and hurting everyone close by, for example, it might be a situation where everyone taking shelter suddenly finds themselves stuck outside the room by a force field. Or maybe when the rogue touches the weird, glowing blue sphere, everyone within ten feet gets blessed.

Building this type of variation into your scenario design means that, even if the players are triggered by the metagame knowledge that you’re asking the question, they’ll still need to think about what their answer will be. (Instead of just automatically running as far away as they can.)

Basically, all of this is aimed at allowing their character to be competently aware of the heightened stakes of a situation, while not necessarily giving them the equivalent of a spidey-sense that unerringly warns them of incoming danger. False alarms, mixed signals, reversed expectations, and the like can all help.

Along similar lines, you can disguise the question by getting the incidental information you need as part of a different question. For example:

GM: The door is locked.

Conan: I’m going to kick it open!

GM: Is anyone helping Conan force open the door?

This is a mechanically significant question (since it will give Conan a bonus on their Open Door check), but you’re also sneakily establishing who’s at the door. (And, using benefit of the doubt, you could then infer that anyone not helping is standing far enough back not to get hit by the trap.)

Another effective disguise can be hiding the question in a little throwaway add-on to a different question. For example, as you’re getting ready to transition to the PCs’ meeting with Paul Dubois:

GM: Okay, you finish feeding the corpse into the wood chipper and then dump the chunks into the tank. You’re covered in blood, but the sharks will take care of the body.

Suzie: Well… I probably shouldn’t walk around town looking like this. I’ll head back to the safe house and get changed.

GM: Give me a Stealth check to see if you can cross town without anyone noticing your appearance.

Suzie: 12.

GM: Great. You make it back to the safe house without incident. After stripping off your blood-soaked track suit, what’s your new outfit?

Here the question of, “What are you wearing?” (i.e., will you mention your favorite I Like Ike pin?) seems to just naturally emerge from the chain of events. You could even reinforce this by waiting for Suzie to describe her outfit, and then saying something like, “Okay, it takes about an hour to get cleaned up, but then Suzie walks out the front door in her beautiful blue dress…” (By using the information to paint the scene, you’ve also provided an explanation for why you collected that information.)

Finally, you can enhance all of these techniques by anticipating the decision. For example, if you know that the door out of the room is trapped, don’t wait until someone goes over to the door to figure out where everyone is standing. Instead, ask the question early. The players might get suspicious… but then nothing immediately happens, so they’ll assume you were just trying to spook them. Or you take that information and use it to re-establish the scene, so they just dismiss it as part of the natural flow of the scene. But then, having established the scene (“so while Conan is poking at the pile of rags, Merlin is studying the orrery”), you now know exactly where everyone is and will continue to be unless they explicitly declare that they’re moving.

Now, when the rogue over and starts checking out the door, you don’t need to ask the question again, and the players are never tipped off.

Thanks to CorpCord for asking the question that inspired this tip.

Should the GM hide their rolls behind a GM screen or should they roll openly where the players can see the results?

A lot of people actually think that hiding their dice rolls is the primary or even ONLY reason for a GM to use a screen, and this can even mire discussions about using GM screens in a debate about whether or not the GM should be hiding their rolls. And the debate about whether or not a GM should be hiding their rolls can often be entirely swallowed up in an argument about whether or not a GM should be fudging their rolls. (Which is, according to these debates, the only possible reason a GM would have for hiding their rolls.)

At this point, as you can see, the argument is already several layers deep in largely unexamined premises.

Let’s see if we can unpack things a bit.

First, I discuss a bunch of great reasons for using a GM screen in On the Use of GM Screens, and hiding your dice rolls doesn’t even make the list. In fact, it’s fully possibly to use a GM screen and NOT hide your dice rolls. So let’s lay aside the idea that these are intrinsically linked.

Second, for the purposes of this post, let’s take it as a given that the GM should never fudge their rolls.

Having discarded fudging as a motivation, why would a GM want to hide their rolls? In my experience, there are three factors:

Convenience. As I mentioned, there are a lot of great reasons for using a GM screen. Therefore, although I don’t always us a screen, I do often use a screen. And while it’s possible to use a screen without hiding your dice rolls, it’s frequently inconvenient.

So when I’m using a screen, I mostly roll behind the screen because it’s easier. In most systems, it would be a huge pain in the ass to stand up and roll the dice on the far side of the screen every time I needed t roll.

Secrecy. There’s a wide variety of situations in which a dice roll is generating information which the players’ characters don’t have access to. (Or, at least, not immediately.) Therefore, it often makes sense also hide that information form the players.

Examples of this includes Stealth checks, random encounter checks, saving throws against illusions, and any number of other possibilities.

Dramatic Effect. When properly framed so that everyone at the table knows what number needs to be rolled on the dice — without doing any additional math; just “I need a 17 or better” — there can be an immense amount of suspense placed on the die roll and a hugely effective and emotional moment that happens when the dice are rolled and the result is immediately seen!

When a dramatic moment like this is happening, you certainly don’t want to under cut it by rolling the dice in secret! And you may even want to make a special effort to make sure the dramatic moment can happen (e.g., precalculating the die result needed even in a system where you typically don’t do that)!

D&D generally doesn’t frame rolls like this, but critical hits are an exception — everyone knows immediately what a natural 20 means! — and can give a little taste of what it can be like. On the other hand, Monte Cook’s Cypher System, if you run it properly, is set up so that almost every die roll works like this unless there’s a reason for secrecy (which, of course, provides its own dramatic impetus).

IN CONCLUSION

On that note, we can see how these three factors can be weighed for each roll to determine how we want to handle it.

So, for example, if I’m not using a GM screen, then I generally don’t care and just roll the dice, unless there’s a specific reason why secrecy is significant for a particular roll.

On the other hand, when I am using a screen, then I’ll generally roll behind the screen for convenience, unless the stakes are high enough that dramatic effect makes it worth the bother of standing up and rolling on the far side of the screen.

Other GMs, groups, or even game systems can easily have different opinions on the relative importance of these factors.

For example, maybe you’re playing a game with very few rolls and, therefore, every roll is a big, dramatic moment:

On the other hand, a GM might feel strongly about not giving their players the metagame knowledge that “there’s a reason this roll should be hidden, and therefore I’m hiding it,” and therefore they’d prefer to hide as many of their rolls as possible. (And this might be something that the GM only cares about because this particular group is prone to metagaming that knowledge. Or they may have had one of the players ask them to mask the metagame information because that will help them enjoy the game more.)

The point is that there’s not really a One True Way™ here. But hopefully a clear understanding of these factors will help you think clearly about when and why you’re hiding your dice rolls, and find the right solution for you, your group, and your game!

BONUS PLAYER TIP: GET DRAMATIC!

If you’re a player, you can set up your own dramatic dice rolls!

Remember that the basic concept is that (a) the stakes of the dice roll are clear, (b) everyone at the table knows what number you need to roll on the dice (with no additional modifiers); and (c) the roll is made in the open so that everyone can immediately see the result!

The stakes of the check put pressure on the roll; and the result of the roll being immediately known provides an instantaneous release of that pressure, regardless of whether the result is jubilant or catastrophic!

It is not, of course, unusual for the stakes of a roll to be known before the roll is made. Assuming you have access to all the other numbers involved, all you need to do to create your own dramatic dice roll is precalculate the result, announce it to the table, and then roll!

In some systems, as we’ve discussed, this will basically be done for you automatically. But in others, including D&D, you’ll need to jump through a couple extra hoops. (You might also need to ask the GM to give you an additional piece of information, like the DC of the check in D&D.)

The other thing to note, of course, is that if you try to make every single roll ultra-dramatic, the net effect will often be to make nothing dramatic. Excitement and emphasis can all too easily turn into tedium.

But if you choose your moments well, you can enhance the game for everyone at the table!

Player: So in Dweredell the Guild is like a local trade organization, right?

GM: That’s right. It’s ruled by a large number of powerful merchant families and its official function is to maintain commercial standards and regulate all matters of craft or trade. But in practice it’s more like a protection racket.

Player: Great! Can you tell me every single member of the Guild and also their immediate heirs?

GM: Uh…

As a GM, it’s actually kind of surprising how often you’ll run into questions like this. In the real world this is the sort of encyclopedic data that you could pull up with a five-second search on the internet. But the game world, of course, does not actually exist, and there’s no quicker way to strip back the veneer and reveal that harsh reality than saying something like, “Tell me the names of everyone who lives on Albert Street!”

On the one hand, it’s a lovely vote of confidence: The players are so impressed by the verisimilitude and depth of your game, that they just naturally assume that the answers to these sorts of questions actually exist!

But however flattering that may be, it doesn’t really change the fact that it leaves you staring out into the vast void of the unknown that they’ve invoked, wondering how it could possibly be filled.

Other examples I’ve encountered include stuff like:

• Can you name every Imperial church and chapel in the city?
• Can we get a list of every front page headline from the Gazette for the month of March 1929?
• Before we question [fictional author], can I get a list of every single book she’s written?
• I’m going to go through the warehouse and check the label on every crate. What do they say?

Of course, sometimes you actually will have a list of every Imperial church and chapel in the city. Those moments — as you reach out, grab the information the players are asking for, and present it with a flourish — are, of course, delightful.

But it’s far more typical, of course, for you to have NOT prepped a label for every crate in the warehouse.

And staring into that void, it’s easy to become trapped in it: Maybe you try to improvise your way through it. Maybe you burble some inanities and then stammer to a halt. Maybe you bring the session to a slamming stop as you spend five or ten minutes brainstorming a bibliography for the fictional author.

Sometimes you’ll want to slam the door shut on the void, even if it doesn’t make any sense: “The identities of the Guild families are a secret!” you’ll cry. Or perhaps, “The crates are all labeled in an unbreakable code!”

Stalling for time is another option, particularly if Google, Bing, and their equivalents don’t exist in your campaign setting: “How are you going to find that information?” (The only drawback here is that the stalling tactic is often limited in its effectiveness and frequently rather boring to actually play through.)

What I usually find effective in resolving this kind of research fishing expedition, however, is a much simpler technique:

“What are you looking for?”

The players have made a very large ask and you’ve become fixated on the impossible scope of it. In actual practice, though, the players are actually interested in some very specific thing related to an unspoken plan they haven’t shared with you yet. (For example, they want to know some details about the author’s work so that they can pose as fans when they talk to her. Or they’re searching the Gazette specifically for any reports of odd occurrences in the Ravenswood neighborhood.)

If you can get them to tell you what they’re really looking for and/or what they’re hoping to do with it, then getting the information they want or creating it or giving them an alternative option or whatever else makes the most sense is often A LOT easier than improvising entire history textbooks or Yellow Pages listings for a fictional setting.

In many ways, this is another invocation of a general principle we first explored in Random GM Tips: Are You Sure You Want To Do That?:

If you don’t understand what the players are trying to achieve with a given action, find out before adjudicating the action.

I can’t do a murder mystery because the PCs will just cast zone of truth!

Awhile back I shared Random GM Tip: Speak With Dead Mysteries, which looks at the particular challenges of designing a murder mystery scenario when the PCs have access to the speak with dead spell (or some similar magical or technological effect). A common follow-up question from people reading this article is, “But what about zone of truth?”

As I mentioned in the original article, you have to start by embracing the fundamental dynamic of a mystery: It’s not to withhold information from the detective(s); it’s about the detective(s) acquiring information.

Insofar as the zone of truth serves as one method that the PCs can use to acquire information, therefore, it won’t be a problem. It will only become a problem if (a) it trivializes all other methods of acquiring information and/or (b) short circuits a specific investigation.

Keeping that in mind, let’s take a closer look.

WHICH ZONE OF TRUTH?

There can be considerable differences in how zone of truth and similar effects work, and this will obviously have an impact on how it affects investigations.

First: Do spellcasters know when the targets of their spells — particularly enchantment spells — make their saving throws?

Personally, I prefer No. (And will often apply this as a house rule even in systems where the answer would be Yes under the rules-as-written or rules-as-intended.) I think pretending to go along with a caster’s domination spell, for example, is a classic genre trope.

This is how zone of truth worked in D&D 2^nd Edition and 3^rd Edition, and it obviously adds a layer of ambiguity to the spell. (If two people contradict each other, is that just a difference of opinion/belief? Or did one of them make their saving throw? Or both of them?) D&D 5^th Edition, on the other hand, adds specific language allowing the caster to know when the spell is in effect, stripping ambiguity.

Second: How often can a character resist the effects of the zone of truth? Do they make a single saving throw when entering the zone? Make a new saving throw every round? Make a saving throw each time they need to make a declarative statement or answer an explicit question?

Once again, D&D 5^th Edition’s version of the effect is the most troublesome for mysteries.

Third: Does the target know they’re in a zone of truth before they speak? If so, it allows them to account for the zone in what they say and how they choose to answer questions.

This has been true for every version of the D&D spell. Wonder Woman’s golden lasso, on the other hand, is usually depicted as surprising those bound by it.

Fourth: Does the effect compel the target to answer direct questions? This would obviously also make the spell a much more powerful tool in the detective’s arsenal, but it’s also not how the D&D spell works.

IDENTIFY THE QUESTION

To paraphrase something Margaret Frazer’s Dame Frevisse once said, the secret to solving a mystery is less about the answers you get to your questions than it is knowing what questions to ask and who to ask them of in the first place.

Often when talking about zone of truth mysteries in a hypothetical sense rather than a practical one, it seems people often default to thinking about an Agatha Christie-style murder mystery where there are eight specific suspects isolated on a country estate. In this scenario, with access to a zone of truth, it’s quite easy to identify both the question (Did you kill Bob?) and who you need to ask (the eight suspects).

But this type of scenario isn’t really common in RPGs to begin with: The plotting of these stories depends on the author having tight control over which questions are asked and when they’re asked in order to create a clever logic puzzle, the solution of which can only occur to the detective at the moment of the author’s choosing. Even if you want to prep a plot in an RPG — and you shouldn’t — it’s still basically impossible to force this kind of sequencing at the table.

As a result, RPG mystery scenarios tend to be built around other structures. For example, it’s not unusual for a mystery scenario to begin at a crime scene filled with physical evidence — e.g., the bloody carnage of a worg attack or the pale gray corpse of a vampire killing — and nary a suspect in sight. Before the PCs can start asking questions, they’ll first need to figure out — as Dame Frevisse said — what the questions are.

To put it another way: Zone of truth can’t short circuit the investigation if the investigation is about figuring who you need to question in the first place.

Next time you’re reading a well-made RPG mystery scenario — like Eternal Lies or Quantronic Heat — take note of how rarely NPCs actually lie to the PCs. Even NPCs who are just withholding information aren’t terribly common. In other words, even without a zone of truth, it’s not unusual for every NPC in these scenarios to say nothing but true things, and the scenario still works just fine.

There is one exploit, however, that clever players can use a zone of truth to unlock:

Can you think of anything that would help our investigation if we knew about it?

Given any kind of limited suspect pool, this question can be used as a quick shortcut for identifying the question(s) they need to ask to solve the mystery, so if you don’t want to get caught flat-footed at the table, it can be useful to prep the clever answer your bad guy(s) will use to evade it.

A good, one-size-fits-all solution here is the incomplete answer: They have to speak the truth, so they do, in fact, have to give the PCs something helpful. But it doesn’t have to be everything that they know would be helpful, nor does it need to be the most helpful thing. Stuff that wastes a ton of time while, technically, being useful is a great fit here.

Can you think of anything else that would be helpful?

Clever player.

The NPC can’t say, “No,” of course, but deflecting back to the first answer is usually possible.

Player: Can you think of anything that would help our investigation if we knew about it?

Bad Guy: I’d check the security cameras.

Player: Can you think of anything else that would be helpful?

Bad Guy: Hmm… I really think the security cameras are what you should check next.

GAIN ACCESS

The other crucial thing about deploying a zone of truth is that you need to (a) get the suspect in the zone and (b) get them to answer your questions.

Returning to our Christie-style manor mystery, if seven of the suspects all readily hop into your zone of truth and the eighth suspect refuses, then the jig is probably up. To avoid this problem, you need to have multiple suspects refuse. The PCs may still be able to use the spell (or the threat of the spell) to narrow the scope of their investigation, but not close it.

The reason for refusal might be specific to each character. The most likely explanation would be some dark secret unrelated (or at least not directly related) to the crime that they nevertheless do not want discovered.

Alternatively, it might be a society-wide condition that can be broadly applied. For example, it might be a cultural more that casting an enchantment spell on someone is unacceptably rude or unethical, such that even suggesting it will likely provoke an outraged response. It could even be a matter of explicit law, with the use of such spells being tightly regulated, limited, and controlled.

This touches on another issue, which is whether or not the PCs have the authority to question people. If they’re police detectives, then it will give them an advantage. (Although they’ll also probably have to abide more closely to the aforementioned laws, in which case they may need to do a bunch of legwork before they can deploy their spell.) If they’re just a bunch of random bozos who have wandered into town, it may be a lot easier for the suspect to dodge their calls.

Which leads us to another obstacle to gaining access, which is just literally finding them. Sure, you want to question Bob. But the door to his apartment has been kicked in, the place has been ransacked, and he’s either been kidnapped or is in the wind. In other words, the mystery — or a significant part of the mystery — can just literally be trying to track down the person you want to use your zone of truth spell on.

Of course, in order to find the person you want to question, you first have to identify them. The manor mystery, of course, still assumes that the list of suspects is immediately obvious, but in a lot of mysteries it’s anything but.

For example, you’ve found Bob: He’s been brutally murdered and his corpse dumped in an alley. You can’t question every single person in the city, so you’re going to have to figure out how to narrow down your suspect list first.

Even once you’ve gotten a suspect into the zone of truth, though, access can continue to be a problem if you enforce the time limit. The D&D 5^th Edition spell, for example, only lasts for ten minutes. That time can vanish surprisingly quickly, so set a timer.

This can be even more of a limitation in a manor house mystery: How many suspects can you rush through the circle before you run out of spells for the day?

If the PCs are trying to rush bunches of NPCs through their zones and you don’t want to play through every encounter, might rule that each witness requires 2d6 minutes or questioning; or perhaps 1d4+1 minutes if the PCs take disadvantage on their interrogation checks.

DESIGN THE CRIME

Imagine that it’s 1850 and you’re anachronistically GMing a roleplaying game. The science fiction game you’re running describes strange devices known as “security cameras” which record everything that takes place in a room.

“How am I supposed to design a mystery scenario when there are security cameras everywhere?!” you cry. “They can just see who did the crime!”

Modern criminals, of course, simply know that security cameras exist and they plan their crimes accordingly. In many cases, the evidence they leave behind while countering the security cameras will be the very clues detectives use to track them down!

An easy solution, of course, is to say something like, “I drank a potion of anti-enchantment that lasts for forty-eight hours,” thus negating the zone of truth entirely. (A clever criminal, of course, will make sure they have some perfectly reasonable pretext for having done so. An even cleverer criminal will have slipped it into the drinks last night and everyone at the manor house has immunity.)

This sort of stonewalling — where the PCs’ abilities are simply negated — is mostly just frustrating, however. It’s usually more fun to find ways that don’t just completely shut off the spell.

To put our suspect at their maximum disadvantage, let’s once again return to the manor house scenario that takes so many other options (access, identifications, etc.) off the table. This, however, is precisely the situation in which the murderer would anticipate a zone of truth. So how would they plan for that?

• They would anticipate the question, “Did you kill so-and-so?” and therefore have planned their crime so that they can honestly say, “No,” (e.g., they tricked them into suicide, arranged for a convenient accident, or hired someone else to kill them).
• They sent a dominated doppelganger disguised as themselves to enter the zone of truth. If the doppelganger’s identity is discovered, the PCs will find the NPC “knocked out” and tied up in their room. Who did this to them?! The mystery deepens. (Meanwhile, the zone of truth has expired.)
• They covertly trigger a dispel magic effect that destroys the zone of truth, possibly while someone else is being questioned. How many of those spells can you cast today, exactly?
• They arrange for a distraction. “Can you confirm that your name is Miguel Cavaste?” “Yes.” “Can you—” EXPLOSION. By the time of the chaos of the explosion is dealt with, the zone of truth has once again expired.

And so forth.

CONVERSATIONAL GAMBITS

Taking all of the above into consideration, there will nonetheless come the time when the suspect is in the zone of truth and faced with the PCs’ questions.

Maybe that’s all she wrote: Either they confess dramatically, surrender meekly, or initiate the final action scene by attacking the PCs, summoning reinforcements, or attempting to flee.

On the other hand, maybe not.

Instead, the zone of truth questioning can become a cat-and-mouse game: a final riddle for the players to unravel.

Let’s take a look at the conversational gambits an NPC (or PC!) might use when trapped, literally or metaphorically, in a zone of truth.

We’ve already mentioned giving incomplete answers. Just because you have to say true things, doesn’t mean you need to say EVERY true thing you know. Questions like, “Did you see anything suspicious last night?” gives the suspect a huge latitude in directing the interrogators’ attention towards any number of useful distractions.

They can also answer a question with a question. A question, technically, cannot be a lie. So if they can slip this past the interrogators, it can often create the illusion of a false answer: “Do you think Robert could have done this?” or “I had drinks with Marcia last night… what time was it? Around six?”

Another option is to simply refuse to answer. This, of course, looks suspicious. However, we’ve previously talked about cultural mores (“I plead the fifth!”) and simply feigning outrage (“How dare you ask me that?!”) can deflect or, at the very least, buy time.

On that note, simply rambling, delaying, feigning confusion, or otherwise running out the clock is also a perfectly viable strategy when you know the zone of truth has an unforgiving time limit.

SOLUTION IS A KEY, NOT A CONCLUSION

If you think in terms of node-based scenario design, the goal of any interrogation is to gain a lead that will point you to another node where you can continue your investigation.

As long as this remains true, the zone of truth only provides a key that you wanted the PCs to get in the first place. And therefore, of course, the zone of truth isn’t a problem at all.

So this can also bring us back to some basic first principles in RPG scenario design: Don’t prep the specific things that the PCs will do. Instead, prep a robust, dynamic situation for the PCs to interact with.

If you try to prep a specific thing for the PCs to do, the zone of truth can thwart you by giving them an alternative method of achieving their goal. But if you’ve prepped an interesting situation, then the zone of truth will often just be one more prompt for you to respond to with all the cool toys you’ve made for yourself.

FURTHER READING
Speak with Dead Mysteries
Three Clue Rule