Spherical Goblins: Inference of Destruction

One of the first columns I wrote under the heading of Spherical Goblins was on the subject of how to read the rules. This one is, I suppose, a bit of a sequel to that. I’m writing it to address one specific source of confusion for many players, and to use that example to discuss a general case that causes confusion.

This is fitting, as “specific” and “general” are watchwords for rule interpretation in 5th Edition, as in “specific beats general”. These three little rules are meant to be a key that unlocks the rules, or a razor that cuts through confusion, but the problem that some people have is figuring out what is specific and what is general.

There is a whole series of Frequently Asked Questions in 5th Edition that are all founded on the same misconception about a general rule. It’s really just one question, phrased in different ways. Two of the most common:

  • “When I make a bonus attack using Crossbow Expert, do I get to add my ability modifier to the damage?”
  • “If my monk uses Martial Arts or Flurry of Blows, can I still add my ability modifier to the damage of the extra attacks?”

Now, in that column I linked to in my opening, I spoke about how the rules for a special ability like Martial Arts or Crossbow Expert are largely self-contained, how there aren’t more detailed rules governing them somewhere else, only the general rules with which they interact. An attack you make using one of your special abilities follows the general rules for attacking, except where th ability itself lists an exception.

In the case of these abilities, the exception is that you can use a bonus action to do it. None of these abilities say anything about adding an ability modifier to your damage or not, which means you follow the general case. The general case is yes.

Yet enough people are unsure about this that they feel the need to ask, and whenever there’s that much uncertainty about a thing, there are most assuredly going to be people who are certainly wrong about it. There are tables out there where players are shorting themselves damage in every round of every combat. Why? Because instead of following the general rules for attacks, they are inferring the existence of a general rule that governs bonus action attacks in particular, and they are applying it to any ability that grants one.

The existence of this rule is intuited from the specific rules governing what 5E calls “two-weapon fighting”. When you are wielding a light melee weapon in each hand and you use one of them to attack using the Attack action on your turn, you can spend a bonus action to attack with the other one, at the cost of not adding your ability modifier to the damage. So if you’re a rogue and you’ve got Dexterity of 16 and you’re wielding two short swords, you’ll do 1d6+3 with one and 1d6 with the other.

This specific rule is meant to be a quick-and-dirty representation of the ability to make a slightly wild swing or stab with your off hand, though in keeping with 5E’s general economy and simplicity, the words “off hand” and “main hand” never appear in the text. If the weapons you’re using are identical, then it doesn’t matter at all which one is in which hand; if they’re not, then it hardly matters, to whatever extent that it does, you’re subtly encouraged to envision the hand holding the bigger or better weapon as your main one. Any actual mechanical difference comes out in the wash.

The key thing here is that the fact that you don’t do as much damage with a bonus action attack when it’s delivered in this fashion is not a general rule, though, but a specific one that represents a specific thing; the off-hand nature of the strike.

How is this specific rule mistaken for a general one?

Because it seems to be the closest thing to a general case for bonus action attacks. It doesn’t require a feat. You don’t have to be a particular class or choose a particular option. Anyone who can hold a couple of small axes, swords, or daggers can do it.

And because to someone who has not yet grasped the underlying logic and structure of 5th Edition’s rules it feels like there must be a general case to defer to, it feels like this must be it.

As I said at the outset, this is a specific example of a general problem. This tendency to turn towards specific rules for clues to some hidden general one permeates the way new people tend to approach the game.

For instance, another common question is if poison resistance confers advantage on saving throws against poison. Dwarves have both, and it’s hard to get poison resistance as a PC without being a dwarf, which leads many people to conclude that yes, they are inextricably linked.

But the key here is that dwarves specifically get both. As in, the rules specify that they do. If poison resistance—resistance to poison damage, following the general rules for damage resistance—was the same thing as advantage on saves against poison, the description of the dwarves’ particular hardiness against poison wouldn’t need to mention both.

This is counterintuitive to most people, even veterans of most roleplaying games. A lot of rules texts include more clarifying elaborations than 5E does, reminders of general rules that are useful in context but ultimately redundant within the larger system the text represents. In such a rule book, the entry for dwarves would mention advantage on saving throws alongside the damage resistance as a reminder, a reference to a general rule officially spelled out elsewhere. The rule about two-weapon fighting would likewise be repeating a general case, if not establishing it.

5th Edition eschews these redundancies almost to a fault. None of the spells that inflict the charmed condition spell out the effects of that condition, for instance. They merely reference it and add their own specific effects. In another game system, the text for charm person might include language like, “The creature is charmed by you, becoming unable to attack you and granting you advantage on Charisma checks while the condition lasts.”

But that would be repeating information that is general to a condition shared by multiple spells and other effects, so they don’t. Instead they only mention things that are peculiar to the spell charm person in particular, like the fact that the charmed creature regards you the caster as a friendly acquaintance, or becomes hostile to you when the condition ends.

And so, almost predictably, this creates a new source of confusion among people who feel the charmed condition itself must be more than the two bullet points listed under the heading of “charmed” in the appendix on conditions. They look to charm person as being the nearest thing to a general case, in the same way that dwarves are seen as the general case for poison resistance and two-weapon fighting is seen as the general case for bonus action attacks, and so they conclude that the charmed condition intrinsically includes the notion that it changes how the target sees you, or induces hostility afterwards.

In other words, they operate under the assumption that being under the charmed condition basically means “being under the effects of charm person“.

Categorizing these errors and picking apart the thinking behind them might seem mildly interesting to someone who knows better, and pointing out that they are errors might be useful to someone who’s confused, but my purpose here is to do more than that. My purpose in writing this blog post is to help everyone who plays 5th Edition Dungeons & Dragons to better understand how the rules work and why they work, so that specific corrections or clarifications about the rules become less necessary.

In my post about procedural logic, I talked about how to read a rule as a process. Now I’m going to zoom out one level and talk about the same thing on a more macro level, how to fit rules into a hierarchy.

Applying procedural logic to a rule is a bit easier if you have any kind of a background in computer programming. The same is true of this skill, particularly if you’ve ever studied any kind of object-oriented programming language or made use of typed data structures.

Basically, it comes down to inheritance.

Imagine the rules of the game as a chart that looks like a family tree, or rather, a forest of related trees. At the top of one tree, we have top-level rules, like “Roll a d20, add modifiers, compare to a Difficulty Class”. That one would have three branches, for attack rolls, saving throws, and checks, and each of those would have their own branches, which would branch further.

Any rule that occurs at any level of any tree is going to be passed down to the branches below it. This includes rules that modify or even negate higher-level rules. For instance, the rules for attacking include rules for damage rolls, which specify that you add an ability modifier to your damage roll. The rule for two-weapon fighting branches off that, and it includes a rule that removes that ability modifier. One branch below that you can find the Two-Weapon Fighting style available to warriors, which puts the ability modifier back on.

The Dual-Wielding Feat, which (among other things) removes the limitation that the weapons be light could be found either branching from the general case of two-weapon fighting or the style of that name, and would inherit whichever damage modifier rule was being followed by its parent.

So when you have a case and you’re not sure what rules it’s inheriting, you just need to figure out what it’s branching off from. This sounds more complicated than it is. In practical terms, it’s just a matter of following references.

The geas spell inflicts the charmed condition. Its text references the charmed condition. So if you’re not sure if you should be looking at the charm person spell or the charmed condition, you just follow the reference to the condition. The charmed condition does not reference any spell; spells reference the condition. So geas and charm person are both just branches hanging below the condition in the inheritance tree. No matter how much it intuitively feels like a spell that charms a person should refer to the spell called charm person, no such reference exists.

Similarly, the rules for Crossbow Expert might read a bit like the rules for two-weapon fighting, but at no point does it say “you can make an attack as a bonus action, as you would if you were using two-weapon fighting”. The reference is simply to making an attack, so the next thing up the tree would be the general rules for attacking.

It’s just an extension of applying procedural thinking: follow the reference, follow the rules that cover that reference, and you’re done. There’s nothing else to do.

Would it be tedious to run through this process each and every time someone wants to cast a spell or use a special ability? Obviously. But it’s also not necessary. For most people, there will only be a small handful of areas where this kind of confusion arises. Once you’ve used this technique to acquire a basic understanding of where a rule fits into the structure of the rules, you don’t really need to do it again for that rule. And the more you do it, the better an understanding of the overall shape of the system you’ll have, which means you’ll have a better ability to intuitively understand how it hangs together when you encounter a new ability, a new rule.

Perhaps the best skill for avoiding this confusion, though, is simply the ability to recognize when you’re making an undue inference in the first place. Once you understand how self-contained the rules that govern a spell or feature are, and you understand how rare redundant elaborations or reminders are within the text, it gets a bit easier.

Then you can read a spell like shocking grasp and realize that the fact that metal armor gives you advantage on the attack roll is not stating a general rule for all spells that do lightning damage, or it wouldn’t be listed in this specific spell.

You can see that wizards and some druids have an ability that lets them regain a few spell slots in the middle of the day and realize that if this were something that all casters could do, it wouldn’t be listed for these two specific cases.

And while sometimes this means you’ll wind up nerfing some great idea you had, in general, this kind of clarity benefits players, both by allowing you to have a better idea of what exactly your character is capable of while also preventing you from applying limitations from multiple unrelated sources.


Alexandra Erin has been playing roleplaying games since 1989, and has been experimenting with game design and related theory since about one week after that. If you enjoy her writing, you can find her original D&D content on the DMs Guild, or support her directly.

Spherical Goblins: Conceptual Problems

In past columns, I’ve spoken about the issue of Dungeon Masters nerfing player characters on the fly because they can’t make sense of what the game allows the PC to do. The flipside of the DM nerfing a PC “because it doesn’t make sense” is a player thinking it’s only fair that they be allowed to do something with their character “because it makes sense”.

Now, generally, if we’re talking about something that comes up in play and isn’t covered by the rules, I’ll side with the player who reasons that they should maybe be able to at least try this cool thing they thought of. Back in 4th Edition, I would let players cash in an encounter power or daily power use to do something that did not meet the strict mechanical definition of what that power did, but which was clearly related. Example? The ranger power “Hunter’s Bear Trap” lets you pin someone with an arrow to the leg. If someone saw an ally or civilian topple off a wall and wanted to save them, I’d let them make an attack roll and use that power to try to save them by pinning their cloak.

That kind of improvisation was necessitated by the rigid powers structure of 4E, but it comes up in every edition. The rules can’t cover everything. If you feel like your Acrobatics proficiency should let you do something that’s not explicitly covered by it and which isn’t covered by another skill, I’m game. If you’ve got a slightly tweaked use for a spell that doesn’t replicate another spell, I’ll listen.

Where this gets to be a problem, though, is when we’re talking about adding base level capabilities to a character that exceed what is otherwise available. Often these capabilities are explained as being logical outgrowths or even requirements of the character’s backstory or nature, which both can make the player feel like they’re not being unreasonable (they don’t want it to be more powerful, they want it because it makes sense) and like the DM who says no is being unreasonable (they’re ruining your character!)

As an example, imagine a player whose character concept is this: blessed by a goddess of nature, this character has a profound connection with the natural world and an almost empathic ability to relate to animals.

If I hear a concept like that in D&D, I’ll think, great! That one’s easy. Be a ranger, druid, nature cleric, or totem warrior barbarian, or just take a background that gives you animal handling.

But a player who comes up with a concept like that might have a dot dot dot rather than a period at the end of that, and the next part is “so my character should have double proficiency bonus on Animal Handling and Survival, and be able to cast speak with animals at-will.”

Because of the connection with nature and the empathic ability to relate to animals.

The first time I ran up against this kind of thinking, I wasn’t playing D&D. I was trying to put together a superhero themed campaign in GURPS, and I had one player who kept arguing about the point buy system on principle. She wanted her character to have the ability to step out of phase with reality, becoming not just intangible but also invisible every time she did so.

That was fine. GURPS can handle that kind of thing really well. Basically, you pay the cost for both abilities and give yourself a discount for invisibility only working while you are intangible.

Her issue was, “But the invisibility is a natural side effect of her phasing, so it should be automatic. Why do I have to pay for something that happens naturally?”

The answer I gave her is the same one I give to players who feel that the mechanical options in D&D should be supplemented by benefits from their backstory: game balance doesn’t care where your capabilities come from, only what they are.

Being able to turn invisible and intangible at the same time is just as valuable a pair of abilities no matter what explanation you give for them, so it costs the same amount of points.

Dungeons & Dragons does not use such an intricate and comprehensive point system for anything, though it does offer point buy as an option for setting your ability scores (i.e., basic attributes), so let’s take this as an example. For simplicity, I’m going to be dealing with scores before racial modifiers throughout this next bit, just so I don’t have to keep clarifying that the 15 I’m referring to might be a 16 or 17.

Say there’s a player whose character, as a lad, would eat four dozen eggs every day to help him get large. Now that he’s grown, he eats five dozen eggs, so he’s roughly the size of a barge.

Now, the player has done the point buy and put 15 points into Strength, the maximum the system will allow him to buy. But he comes to his DM and says, “I think my Strength should be 18. My character has performed an exhaustive strength-gaining regimen ever since he was a child, and that should be reflected in his stats. It’s just not realistic that I have the same strength score as anyone else would have, given how completely my character’s life has been dedicated to gaining muscle.”

A lot of DMs, in this situation, will be bothered by this argument but not be able to put their finger on what’s wrong with it, so if they wind up saying no out of principle they feel like they shot down part of the character’s concept, or they’ll not be able to defend the rules or their decision any better than by shrugging and saying, “Well, the game doesn’t always make sense.”

But this is really the same situation as in that GURPS game. Having a “mere” Strength of 15 doesn’t require you to ignore your character’s backstory where they have distant giant blood, or had a massive protein intake and bodybuilding regimen, or were blessed with phenomenal strength by a god, or whatever. Rather, it reflects it. Even though as a player you are absolutely entitled to take your attribute points and spend them in order to get a 15 on whatever attribute you see fit, the character you create was shaped and molded by the circumstances of their life in order to wind up with such an extraordinary (and it really is extraordinary!) capability that a 15 represents.

I know, it’s hard to remember that an unmodified ability score of 15 is extraordinary when you’re traveling with 3 to 5 other people who pretty much all have one, too. But this is where it’s important to remember two things: one, you’re not the star of the show here, but two, you are one of the stars.

Back in the day, when we all rolled our stats in order (and then fudged the heck out of them), player characters were theoretically average individuals who might, if they lived long enough, become great heroes. From 3rd edition onward, the trend has been against that, with 3rd edition formally adopting as default weighted die rolls and a generous point buy system. 4th edition really solidified what I call the Lake Wobegon school of character design, “where the halflings are strong, the half-orcs are good looking, and all the adventurers are above average”. 5E more or less continues that trend, not strengthening it but not shying away from it.

But even if a 15 doesn’t make you anything special among the sort of heroic adventurers you find yourself banding together with, it is still special. A score like that really cries out for explanation, honestly, and a backstory that explains how you came to be so blessed is not just acceptable, it’s great.

But while your backstory can explain how you came to have the abilities you do, it can’t be used to justify a score that’s even higher.

That’s talking strictly about ability scores, which have a literal point buy scale associated with them. What about other capabilities, though, like our nature lover?

Even without an explicit “point cost”, the answer is the same. When you choose a background, class, race, skill proficiencies, feats, etc., you are essentially spending a single point to buy a thing or package of things. You can (and by all means, should, if you have the inclination) explain the results of that purchase in terms of your character’s backstory, but you don’t get more points to spend because of the backstory.

If your character is a druid or ranger, then your nature skills and natural magic represent your character’s profound spiritual connection to nature. If you’re making a human with the variant rules and take the magic initiate feat and use animal friendship or speak with animals as your one level one spell, that’s likewise a result of that connection.

“But that’s the feat I get just for being human!” the player might say. “This comes from something unique to my character.”

This is confusing choices you make as a player with things about your character. The ability to choose a feat just for being human (or rather, just for making a level 1 human adventurer who is intended to stand alongside humanoid dragons and quasi-immortal forest jerks) is a player resource; but the feat you choose is (or represents) something unique about your character. You can do as much or as little as you want to integrate that unique thing into your character’s concept and backstory (or ongoing story).

Now, there is a difference between wanting something extra and wanting something different. If a player wants to use a custom feat in place of any of the existing examples, or wishes to have a custom sub-class to take instead of one of the existing ones, that’s theoretically no different than spending their points differently in order to achieve a desired result.

For instance, it’s not uncommon to encounter a player who wants to play as a heroic werewolf or vampire. The Monster Manual already provides rules for converting a bitten character into monsters of those type, but they add an immense amount of power in the process and balance-wise, they really work better for “your character died and now there’s a new villain” than “your character was bitten by the fantasy equivalent of a radioactive spider and now you have all of the superpowers”.

This isn’t to say that you have to say no. The trick is to figure out what abilities a vampire or werewolf character class would have at level 1 that would be comparable to what any other character class gets at level 1, and give them that as an option for their next character level. If they want to keep developing their monstrous new abilities, then at level 2, you do the same thing, and so on.

Obviously this is a lot of work compared to the alternative, which is saying, “No, lycnathropy/vampirism is not a cool origin story but a horrible, horrible curse which basically means you have no control over your actions/you are dead and a twisted reflection of you is occupying your body,” and it may or may not be worth putting in that much effort to keep things balanced. And if it’s not worth it to you, then I urge you to just say no, because having one player who is disappointed is better than wrecking the game for everyone else.

Also? It’s not like there’s a shortage of cool character concepts that can be modeled with just what’s in the Player’s Handbook. Players who are that blasé about the possibilities embodied by high level characters of any class probably don’t understand the potential of them, and would benefit by spending more time looking at them.

But at the end of the day, there’s nothing wrong with coming up with alternatives for players whose character concepts are on the same general level as what’s available, but not otherwise mechanically represented. The trick is to make sure that whatever they get from the custom feat or sub-class or whatever is not any better than what they give up in not taking an existing feat or sub-class.

This can be trickier than it sounds, which is why I recommend that inexperienced players work with what’s already available and, even more so, that inexperienced DMs practice saying no to any custom options pitched by players. You have to know how the system works and why it works the way it does before you can understand if and how a new option is going to work with the system.

One area that can be really tricky is the idea of using a custom “race” build to implement a character concept. There’s a certain school of thought that already views character class and race as nothing but abstract bags of mechanical pieces. If you have a character concept for an elf that would benefit more from a high Constitution and more hit points, they’ll tell you “Take dwarf and say you’re an elf. That’s just roleplay.” Why anyone would say something is “just roleplay” in reference to a roleplaying game is beyond me, but we’ll let that go. The point is that if you see the races in the PHB as nothing but means to achieving an end for your character, then the next step is to treat them as a sort of smorgasbord.

This might be a pet peeve of mine, but I’m always immediately suspicious of a player who wants to play mix-and-match with racial abilities. It might seem like this is a decent approach to keeping things balanced, but they rarely weigh what a given feature is worth in any real way. It’s more a sense of “Well, there are races in the book that have this, so it must be balanced if I have it.”

The most egregious example I saw was a player who wanted a character who was human who had been blessed by angels but who also gained power by drinking demon blood. He took pieces from the tiefling, aasimar, and variant human race, and added another feat (on top of the one you get “just for being human”) to represent demonic strength/power. Another, less egregious recent example was a player who wanted his character to have both the halfling luck trait and the Lucky feat at level 1, in order to make a character whose concept is the luckiest person in the world.

Neither player was at my table, but I would have given both of them the same answer: make your character normally and then use this backstory to explain what you’ve got.

If your demonic-angelic avenger character is a paladin with reasonably high Strength and Constitution scores, you’ve got your blessings (the paladin abilities) and you’ve got your demonic power (the physical attributes). Take the tiefling or aasimar race to mechanically represent more demonic or more angelic flavor, or just take variant human and use your feat to get a couple of cantrips that represent both.

If your character is supposed to have luck as an overwhelming defining trait, play them as lucky. Take the halfling or variant human race to get some mechanical luck, sure, but don’t stop there because those two abilities put together won’t come up often enough to really define you. Take high dexterity but describe your character as no more than usually graceful, just prone to getting off lucky shots and being missed due to happenstance.

A rogue with this kind of character concept might have their Sneak Attack damage defined not as a matter of skill but everything lining up perfectly. A monk might have their martial arts described as coincidental slapstick. And so on.

Now, the player with the angel/demon knight concept was already planning on being a paladin. He wanted more supernatural blessings than you get “just for being a paladin”, though. The real problem with the character was that he, as my boyfriend Jack puts it, “had a level 20 backstory for a level 1 character”.

I.e., he’d already met angelic beings who singled him out for special powers and a great destiny, already learned the secret of gaining power from drinking demon blood. D&D player characters are by definition potentially epic heroes, but you have to go on the epic quests to get there.

And that, more so than any point about game balance or creativity, is my final answer to any player who feels strongly that they should be allowed to have a bunch of extra powers not because they want them for any advantage but because they think it would be fun to play a vampire or a god in human form or an indestructible construct or a celestial half-fey half-dragon. The game you’re playing is designed to model certain things, and those things are mostly “the increasingly epic adventures of increasingly epic mortal heroes.”

You start out special, you start out larger-than-life, but there’s a ceiling on those things, and if you can’t work within those limits, you need to dial your vision back in one way or another, if not by throwing it out than by deciding to grow into it.

4th Edition had epic destinies (endgame prestige classes, basically) that represented a character who gradually realized over the course of their career that they were the living embodiments of a god or primordial, or for characters who become royalty in hell, or powerful members of a fey court. In each case, though, you got no mechanical in-character benefit from these things before level 21, and then you spent 10 levels growing into it.

5th Edition does not have anything like that, but by the time you reach level 20, you are metaphorically or in some cases literally a force of nature. If you can’t raise the dead or reshape reality with a word, you can stand shoulder-to-shoulder or toe-to-toe with gods and titans.

But you’ve got to get there.

Going strictly by encounter XP budget, by the time you’re ready to tangle with a vampire one-on-one you’re going to be level 19. This is not a perfect measurement, but it’s close enough to make the point that a single vampire is comparable in awesomeness to a very high level character. You can’t expect to get that at level 1.

As a final note: DMs who wish to head off a certain amount of more reasonable “concept fishing” and allow more customization at level 1 might consider a house rule that allows each player character to choose a single feat at level 1. You might also consider taking the feat-based human variant off the table if you do this, though in my experience, the appeal of that option falls off sharply when it’s not the only path to getting a feat at level 1.

Alexandra Erin has been playing roleplaying games since 1989, and has been experimenting with game design since about one week after that. She is the author of the character option manuals Clerics of Lesser DomainsWarlocks of Other Patrons, and Feats of Heroism: 36 New Feats for 5th Edition, currently in the top 5 titles on the DMs Guild.

Spherical Goblins: The Healing Power of DON’T


Some of the advice I give here is for DMs, some is for players. Today’s column is a little bit of both.

It’s about four magic words that are the answers to a lot of the questions, problems, and dilemmas facing people who play D&D.

It won’t help you with a rule question or a disputed ruling. It’s more a matter of implementation.

Let me give you an example of some of the problems that this miracle answer can help you completely circumvent.

“I want to try this whole group variant initiative rule from the DMG, but I’m worried that if the monsters win initiative they’ll all just focus fire on one player character and kill them in round one before anyone else can do anything.”

“My character concept is a zany chaotic neutral character who is basically a living cartoon character, but I’m worried that I’ll steal focus and annoy the other characters.”

“I really want my PCs to encounter a dragon for story reasons, but they’re not powerful enough to take it on and live, and I don’t want it to be a TPK.”

“I’m playing a traditional lawful good character and it keeps causing problems because my character won’t go along with the rest of the group unless they do things my character’s way.”

These kinds of questions are among the most common non-technical questions you see being asked. A DM has concerns about a scenario or a rule variation that might lead to a less-than-fun curb stomp of doom against the players, or a player has a character concept they’re really attached to but they can’t figure out how to make it work without driving everyone else away.

However, as I said, there are four little words that can solve any one of these situations and hundreds more. Those words are:

Maybe don’t do that?

Understand, this isn’t an injunction to abandon the plans that are giving you pause. See, in each of those cases, it’s not the scenario that the person is worried about, but what they see as the inevitable outcome.

If every single monster goes before every single PC, the monsters will focus on one target to slaughter unopposed.

If the low level PCs fight a dragon, the dragon will kill them.

If you have an honor-bound or zany character, that character will annoy your teammates or derail the story or otherwise ruin everybody’s enjoyment, including yours.

But none of these things are actually inevitable consequences. They’re all choices that, in these cases, are invisible to the people who are making them.

Hence the solution: maybe don’t do that?


Stick-in-the-mud paladin, you seriously don’t have to change your alignment or personality at all. But in the moment when you the player, the person sitting there playing a fun game for fun with your friends who are also looking for fun, are making the decision whether to drag your feet or allow the game to progress, instead of uttering the dreaded “My character wouldn’t do that.” when something obviously needs to be done, figure out how your character would.

In real life, a personality isn’t an iron bound code of parameters that cannot be breached. People do things against their conscience and against their better judgment all the time. Even more to the point, people rationalize the things that seem necessary for them to do.

You might say “My character would never compromise on this.” You know who else says they’ll never compromise on the things that matter to them? Literally everybody on the planet. You know who does? Literally everybody. We’ve just got a reflex in our brain that recasts things before we even realize we’ve decided to do them. When you’re dealing with a fictional character whose adventures are completely imaginary and you can imagine their actions at your leisure, it’s easy to imagine that they’d actually manage to hold on to some kind of moral, ethical, or personal absolute, but actual people are more complex than that.

You can stay in character just fine by voicing an objection as the plot progresses anyway, I promise. If you’re actually into roleplaying, there’s actually more dramatic potential in roleplaying reluctant action than in roleplaying obstinate inaction.

This isn’t just for stick-in-the-mud paladins, either. Your greedy, take-everything-that-isn’t-nailed-down rogue type? You don’t have to steal from your teammates. Roleplay the temptation. Roleplay the frustration. Roleplay the character growth. You don’t have to forget that your character is a thief, you just don’t let this fact steal the fun from everyone else.

Your fussy, squeamish, homebody type who hates fighting monsters in dark places and would rather curl up at home with a good book? That’s a fine character concept, but if you attach it to a D&D character in a standard campaign, you’re tacitly agreeing that said character will go into dark places and fight monsters. Bilbo Baggins slept late, but he still got out the door and ran after the dwarves.

Now, you can create a character who absolutely wouldn’t go on the adventure, who wouldn’t work with the others in the party, who wouldn’t work towards the same goal or any goal at all… but, this is an ensemble game about working together to achieve goals on an adventure, so, y’know, maybe don’t do that. Don’t take it that far.

When I run a campaign, I tell my players that no character concept–and no alignment–is off-limits, but the presence of their character in the game means they’re part of a story, and they need to come up with a reason for their character to be there.

If your character is so fussy, so evil, so wacky, so honorbound, so possessed by demons, so a plural entity in one body, or so whatever that you honestly can’t see your character doing the adventure thing with the others, then that character might be perfectly valid in every other way, but you are playing as Ser Not Appearing In This Story.

And Ser Not Appearing In This Story does not appear in this story, which means you need to make another character. This other character might be exactly like Ser Not Appearing in every other regard, but the new character has a reason to be there, and knows how to rope it in.

Remember: there’s no audience watching you at home. Rather, you and the friends with whom you play are the audience. The other players are your audience. If they’re not amused by your antics, then why are you doing them? If your character’s drama and pathos do not help create a richer experience for them, then why is it there?


On the subject of pleasing audiences, when I talk to individual players about not going overboard with their character concepts, I like to talk about Ferengi. When the Ferengi first appeared in the Star Trek universe, early on in The Next Generation, they were these comically exaggerated villains.

Seriously, everything about them was overblown. Their reputation (initially they sounded a bit like the Reavers in Firefly, if the Reavers were trying to sell you something) was overblown. Their mannerism were overblown. Their speech was overblown. Their greed and above all the criminal credulousness caused by their cupidity were overblown.

A thing happened, though. The more that Ferengi appeared on screen, though, the less ridiculous they became. Their first on-screen appearance was the worst, and then it was uphill from there. We could think that maybe it just took the creators a while to get a handle on them, but I don’t think that’s the whole story.

See, the biggest leap forward that the Ferengi got in terms of characterization happened when Deep Space Nine started. Armin Shimmerman, who had helped originate the portrayal of the Ferengi on The Next Generation, was cast as a series regular named Quark, the first major Ferengi character and one that viewers would have in their living room week after week.

If Quark had been one of the shrieking, hissing Ferengi of the “HOOO-MAHNS GIVE GOLD”-style (okay, that’s a slight exaggeration, but only very slight), viewers would have been turned off by Quark’s presence, and either he would have gone, or the show would have. Even if the show had survived, it would have been very different without the character we got, and his relationships with the rest of the cast of characters.

Actually, let’s talk about Quark and the show’s lawman, Odo. Quark is chaotic neutral with some evil tendencies, Odo is lawful neutral with some good tendencies. But their obvious mutual antagonism is consistently played in a way that is conducive both to the running of the space station and the running of the show. Neither character runs wild in a way that derails the whole thing, and yet if you watch it, it doesn’t feel like they’re holding anything back.

In the first season, Odo drops by about once an episode to say “I’m watching you, Quark”… but by the very virtue of that, they have more scenes together and lines with each other than either of them do with many other characters.

So there’s your more complete answer for how to play a stick-in-the-mud paladin or a career criminal in an adventuring party: just look to the stars.

It might sound a bit glib, but really, most TV shows give you examples of this principle. All of the characters in an ensemble will have flaws and quirks and codes of honor that would ruin the show nigh-unwatchable if they were treated as simple ironbound rules or always played to the hilt, to say nothing of how it would sabotage whatever the characters are trying to do.

But what happens instead is that the characters all work together anyway. They have to, or else it would all fall apart. This isn’t to say that it’s not “realistic” to have people who just don’t get along or to have a situation go completely pear-shaped because of the foibles of the people involved. Sure it is. But those situations are by definition short-lived compared to ones where people work things out, or just click.


There’s a thing I tell DMs as a general rule of DMing, and that is player agency is all-important. That is, DMs should never blithely narrate that a player’s character does a thing the player did not choose to do.

But this is advice some players need to remember, too. Your character doesn’t decide what happens. You do. Even if it seems so obvious to you that they must do this or they can’t do that, it’s a decision you’re making.

And listen, I’m not saying there’s nothing wrong with making a bold stand for what your character believes in, or for letting your character live and breathe as a living, breathing, flawed individual.

But this post is premised on the existence of players who see “their characters” doing things they don’t want them to do, who feel trapped by their character concept into doing things that they’re sure will alienate the rest of the group and ruin the game.

The solution for that really is as simple as four little words:

Maybe don’t do that?

That’s all it takes. People often hype up the roleplaying hobby by saying that you have the power to do whatever you want, the endless power of your imagination. That’s maybe a bit overstating the case, as you don’t need rules or dice for that. What the roleplaying game brings to the table is some kind of limits, definitions, and delineations to give structure and meaning to an imaginary adventure.

But there’s a basic core of truth to the idea that you can do whatever you want in a roleplaying game. The problem some players have is they forget or don’t realize this means they don’t have to do what they don’t want.


The situations I’m talking about for DMs are completely different from the ones that afflict players, even though the answer is the same.

This is because beneath the surface, they all share a common cause: the denial of that agency, the abdication of that responsibility. I spoke about this in a previous Spherical Goblins post, regarding the idea that you just have to form the so-called Conga Line of Death in editions where there’s an advantage to flanking.

I’ve seen DMs say they can’t adjust the difficulty of an encounter if, say, the players encounter the Big Bad a few levels earlier than they planned, “Because my players will know I’m pulling punches.” But this is a roleplaying game. You the DM aren’t the one facing the too-weak opponents, the character is. And would your villains really be pulling punches if they just don’t fight like cornered rabbits?

That’s really what it represents, in-game, when you play every combat out to maximum tactical advantage, [matching the players metagame for metagame], or worse, making them match you. The conventional wisdom that says that tactical metagaming is expected or required for DMs is that someone in a life or death fight would do anything that increases their chance of winning, and thus surviving… but does this logic really apply equally to every conflict?

If you’ve got a situation where you know the creature or creatures facing the PCs could wipe them out in one round if you played them to maximum ruthless efficiency, then how is it the same life-or-death fight you’d expect if both sides were evenly matched?

And if it’s not a life-or-death fight for one side because that side is overpowered, then there’s nothing wrong with playing it as something else, is there? Nothing wrong with the villain being a little cocky, a little complacent.

A dragon that would be an intense challenge for a group of 5 or 6 high level adventurers doesn’t have to go all-out to kill a group of 3 level 1 adventurers.

Listen, I get a lot of pushback when I talk about this in open forums. People say I’m suggesting they “go easy” on their players or “let them win”, but the thing is, I only make these suggestions in contexts where people are talking about how a situation would definitely kill all the player characters and how this is bad.

Letting the players win is not part of the game, even though honestly, 5E is very much built around the idea that they’ll win most of the time, with the challenge coming in how small fights wear them down before the big ones. (People who’ve followed my own game design efforts will understand immediately why I’m such a fan of it.)

But neither is stacking fights to kill them. Unless you’re playing seriously old school with truly random encounters, you don’t give them an inappropriate level of challenge in either direction. You give them a fight they could lose, but one they can win, too.

And no one thinks this is “going easy” on anyone. No one derides it.

Even more so than the players staring at the front of the DM screen, the DM really does have the power to do anything. Yet even more so than among players, there’s a severe tendency among DMs to completely overlook all the choices that go into what they do, until the game in general and combat in particular basically plays out like a script being rigidly executed.

If you’ve fallen into this trap, it’s time to free yourself from it with the magical, life-affirming power of maybe don’t do that.

If you feel like you have to focus fire on a single player character because it’s the tactic that “makes sense” to you as a wargame but you don’t really want to do that because it makes the game less sense, maybe don’t do that.

If you want to have a fight with a small army of some ridiculous number of goblins but you know it will be a crushing TPK if they all attack at once and they all use their Nimble Escape every turn to dance away, maybe don’t do that.

If you want to have (or circumstances create) an encounter with a powerful dragon or other giant creature before the party is ready to actually fight it, it doesn’t have to be a fight. Even if the encounter leads to violence, the monster doesn’t have to “fight” the way it would against credible threats. If there’s something you know the PCs absolutely could not survive, maybe don’t do that.


The bottom line is that whether you’re a player or a DM, if there’s ever a point where it seems inevitable that you’re going to do a thing and this thing will ruin the game, stop and think: is it really inevitable? What or what is the active agent that would make it happen?

Chances are the answer is you. Chances are you’ve already realized this, on some level. Chances are that once you’ve admitted it, it’s just a short walk from there to embracing the power of don’t.

Just don’t.


Alexandra Erin is a 26-year veteran player of Dungeons & Dragons. You can find her original gaming content on the DMs Guild.

Spherical Goblins: Start Making Sense

In my last Spherical Goblins post, I talked about how to read and interpret the rules governing a special ability on a mechanical level. In this post, I’m going to talk about how to do so thematically.

An awful lot of DMing advice centers around the idea that you as DM have complete control over the world and over interpreting the rules of the game. “The rules are just guidelines!”, after all. And “if something doesn’t make sense to you, you can just change it!”

At the risk of going against the conventional thinking, I’m going to say that the latter suggestion is actually terrible advice, particularly to new DMs. A lot of things that “don’t make sense” about a roleplaying only don’t make sense because you don’t yet have the experience to see how it all hangs together, or you’re hung up on a preconceived notion about how you think things should work.

For instance, a lot of DMs will disregard the cap on falling damage (it stops accumulating at 20d6) or simply rule that certain falls are insta-kills, because “obviously there’s no way you could survive that.”

But the thing is, if we’re talking “realism”? There is no actual upper limit on how far a human being can fall and survive. The world record for surviving free fall is something like 10 miles. Terminal velocity means that past a certain point, a body in motion stops accruing kinetic energy, which means that just like in the game, there is a cap on the damage that can be done from falling.

Of course, it’s not likely you’ll survive a fall of multiple miles, or even multiple hundred feet. People die falling fifteen feet sometimes. A lot of it is down to what part of you lands on what kind of surface, and you don’t have a lot of control over that in the moments you spend falling, which is why the game uses dice for the falling damage.

So this is an example of a rule that is if not a perfect model of reality, at least based in reality, but which many people throw out for being counterintuitive.

Other rules that commonly get ruled against in the name of reality include:

  • The ability to sneak attack when you’re not literally sneaking.
  • The ability to regain HP by resting.
  • All manner of things having to do with the monk’s Martial Arts and related abilities, like doing more damage with weapons and using Stunning Strike on various non-humanoid creatures.

One thing all of these things have in common is that they’re all changes that take something away from a player character. I mean, sure, the rules for natural HP recovery could be used by a DM to figure out how much antagonists bounce back between encounters, and an NPC could have levels in rogue or monk, or the equivalent abilities. But by and large, DMs who engage in these kinds of off-the-cuff reality checks are making calls that “nerf” the player characters.

Less formally, it’s not uncommon to encounter DMs who do things like have all the players’ enemies immediately stick their hands through any illusion cast by a PC on the grounds that “obviously they’re not going to believe it’s real when it just appeared out of nowhere.” Notably, those same DMs never have a character stick their hand into a wall of fire or poke at a summoned bear, even though these things also “just appear out of nowhere”.

It’s the illusion one more so than anything that reveals one of the subconscious motivations of these kinds of objections and the rulings that they spawn: an inability to let go of an adversarial approach to the game. If you think of the NPCs under your control as “your side”, playing pieces to fight against the PCs—even if you’re just trying to give them a run for their money, not outright crush them—then you’ll be concerned, consciously or not, with making sure “your side” gets a fair shake.

For things like illusions, disguises, mystical compulsions, and other sundry ruses, there’s the added dimension that you as DM are expected to play along with them even though you know better. If you have a hard time separating your ego from the persona of the characters who are being fooled, this can be a tad insulting.

So you rule that “your side” doesn’t fall for the sneak attack, or illusion, or mind control spell, or whatever… not because you’re on an overt power trip, but because it’s in your perceived interest to find objections, and you also happen to be the one who rules on such objections.  If you were a player and your character was a gelatinous blob with no nervous system and the DM said you were paralyzed by a fancy martial arts punch, you’d be like, “Wait, how…?”, and so you do the same as a DM, only you’re the one you’re asking, “Wait, how…?”, and so without an obvious answer, you say those four little words that should strike more fear into the hearts of a player than any other, up to and including “The dragon looks hungry”.

Those words are, “That doesn’t make sense.”

The thing is that when you’re DMing and you ask, “Wait, how…?”, not only are you implicitly asking yourself, but you have no one to blame but yourself if you don’t bother coming up with an answer. This is a game powered by the human imagination. Nothing in it makes sense in any inherent way; it’s all abstract rules and arbitrary numbers until someone comes along and makes sense of it, and when you are the Dungeon Master, that someone is you.

As soon as a DM declares that some intersection of the rules, the dice, and a player’s decisions don’t make sense… well, that’s like a pilot wandering back into the cabin and announcing that no one is flying the plane. It’s technically true, and yes, it is a pressing problem… but one easily solved by the person pointing it out turning around and doing their job.

As a DM, it is your responsibility to make sense of things. The rules of D&D are not a detailed model of reality. They are (particular in 5th Edition) a very simplistic way of calculating success or failure for broadly defined tasks, with a bunch of exceptions stapled on to represent Cool Things Some People Can Do. There is a high level of abstraction and simplification

Ideally, players are making their decisions (both when creating characters and while playing the game) based on an understanding of what things the rules allow them to do, or at least try to do, and what the general odds of success will be.

If a player chooses to play a rogue based on the understanding that they have the ability to Sneak Attack for extra damage in certain circumstances and that same player chooses to engage an enemy based on the understanding that they will be able to use this ability, a DM who declares, “Wait, that makes no sense. The guard knows you’re standing right there. How could you Sneak Attack?” is invalidating that player’s choices and more, destroying the ability of the player to make informed choices.

Now, you can accept that the rule is what it is and just chuckle about how ridiculous and unrealistic the world of D&D is, that you can Sneak Attack someone who knows you’re there. Many people do, concluding that the game makes no sense and just rolling with it.

But there’s no reason to choose between accepting a lack of plausibility or nerfing the everloving heck out of certain classes. You can make sense of these things. The most important step in doing so is realizing how simplified the rules are. They act as though combatants are standing still 5 feet away from each other while fighting. They act as though battle is a matter of taking turns. If you realize that the actual battle that the rules and die rolls and turns would be describing would not unfold as neatly or nicely as the one the turn system and movement system approximate, it’s a lot easier to understand how a moment’s distraction on the target’s part is enough to score a “Sneak Attack”.

It isn’t that the victim didn’t expect to be attacked; it’s enough that they didn’t see a particular attack coming.

Now, the text of the Player’s Handbook never actually explains this. In 5E, they have a huge tendency to stick to the nuts and bolts of mechanics rather than explaining the intent. There is, however, a single line of descriptive text at the start of the Sneak Attack ability:

“Beginning at 1st level, you know how to strike subtly and exploit a foe’s distraction.”

A lot of the people who are hung up on how Sneak Attack can work when you’re not sneaking miss that (just as they miss the fact that the rules for Sneak Attack don’t mention stealth once). Last time, I mentioned how remembering the shorthand version of a rule can trip you up. This is the thematic counterpart of that: only remembering a single shorthand version of what it is an ability is supposed to represent.

Sneak Attack works (among other times) whenever you have advantage, and you can get advantage by attacking from hiding… that coupled with the name makes it easy to think of it as the “attack-from-hiding” ability. But that one line of descriptive text should be enough to prove that wrong; Sneak Attack is the ability to “exploit a foe’s distraction”.

But say that text wasn’t there. Say you had nothing but the bare mechanics to read:

“Once per turn, you can deal an extra 1d6 damage to one creature you hit with an attack if you have advantage on the attack roll. The attack must use a finesse or a ranged weapon.

You don’t need advantage on the attack roll if another enemy of the target is within 5 feet of it, that enemy isn’t incapacitated, and you don’t have disadvantage on
the attack roll.”

Can we make sense of this? Of course we can. First, the rogue can do more damage when using weapons that may use dexterity to attack, so we can infer that the added damage is due to what we might term precision. The extra damage only happens if the attack has advantage—meaning the target is in a vulnerable position—or the target has another enemy within 5 feet.

Why would having another enemy nearby make the target vulnerable like this? The most obvious answer is: they are distracted.

Why aren’t they so distracted that anyone can do this extra damage?

Because no one else has this feature, is the obvious answer. And maybe it sounds glib, but from that we can make the next inference: only the rogue has the expertise necessary to do this.

How do we know it’s this and not something else?

Because this is the explanation that makes sense.

Many other common complaints about a lack of sense in the rules similarly comes down to failing to understand or accept the amount of fuzziness in the model. The rules for HP recovery are an example of this. I’ll admit, this one can be a matter of taste, which is why the DMG includes official variant rules for making healing either more or less forgiving than the default. But I have seen a lot of people running for the “gritty” variant as their new default because “it doesn’t make sense that someone could be almost dead and then rest a bit and get all their HP back.”

But this is a fundamental misunderstanding of what hit points are. The truth is, the number of dagger wounds it takes to kill a person is one, as long as it’s the right one. No amount of fighting experience changes this; instead, it makes one better at avoiding that last hit that will put you down. This is what Hit Points are: not the size of the dog in the fight, but the amount of fight left in the dog. The hit points you have don’t measure health or vitality or life force or blood left in your body, and the amount of hit points you lose don’t measure your woundedness in any absolute sense.

The hit points are your grit, your determination, your pain resistance, your ability to keep your guard up and take hits in a way that won”t take you down. The damage is pain, fatigue, eroding morale, minor injuries that slow you down. The last hit that takes you down is the only one likely to be “bad”, and how bad it turns out to be is going to depend on how you roll afterwards… after all, if you were bleeding to death from severe internal or external injuries, you couldn’t make three “death saves” and spontaneously stabilize within 30 seconds.

It wouldn’t make sense.

So we make sense of it: being reduced to 0 HP is being taken out of the fight, with a hit that is certainly worrying to look at but may or may not be serious.

Now, it is a serious simplification that the game by default does not allow for even the possibility of a long term injury from such a hit, and given that the player characters are the only combatants we follow for the long term, this simplification is definitely in their favor.

But that is all the more reason not to change the rules on the fly. If you want to inject a little gritty realism with things like lasting injuries, have at it. Just make sure everyone knows the score before they wind up in a situation where they’re going to be saddled with one.

If a player makes the choice to rush into the dragon’s den or jump off a cliff, they are already accepting the possibility that their character might die… but that is balanced against the possibility that they might not. They make the decision based on what they know of their character’s capabilities, as measured by their abilities (including hit points) and the rules for interpreting those abilities.

Changing the rules on the fly makes the whole system make less sense, because every rule in the book now has an implicit asterisk after it that says, “Unless something bugs me about a particular example, in which case we’ll just make something up.”

At the end of the day, it’s true that the DM is the final arbiter of the rules, and there’s nothing wrong with adding a house rule or two or seven.

Just follow these simple guidelines:

  1. Make sure you understand the rules as written before you start trying to fix them.
  2. Only make mid-game rule adjustments or exceptions to the rules when they benefit the players.
  3. Do not introduce changes that hurt the player characters without discussing it outside of gameplay first. Ideally players should know any such house rules before the campaign begins.
  4. Never introduce changes that remove or massively deflate a class feature.

And above all, when you find that something doesn’t make sense… make sense of it. It’s what you’re there for. The dice and the rules, they tell you what’s possible, what succeeds and fails. They don’t tell you the story. You, the DM, do that.

Spherical Goblins: Rules, Exceptions, and Procedural Thinking

A lot of the pitfalls I talk about regarding gaming come down to the attitude that players or DMs take towards it, but even the best attitude can lead to confusion and frustration and arguments when you don’t grasp how the rules of the game actually work together. This is how you wind up with situations where the people on the opposite sides of the table have very different ideas of what the PCs’ actual capabilities are, or where neither side actually understands what the characters are capable of doing.

It usually manifests something like this:

A character has a certain feature or skill or spell. This ability is described in the rulebook as a set of rules, procedures, really, that govern how it works. The rules are designed to convey how to apply it to the game, so they are not easy to hold in your head unless you’ve got a particular turn of mind. So you don’t remember the rules, you remember what the rules do, what they mean for you.

Really, it’s a failure to understand how the rules are written and why they’re written that way, which leads to the failure to understand how to read them in the first place, which leads to you asking (yourself or the DM), “So, what’s this do in simple terms?” and only ever internalizing the plain speech explanation.

One root of the problem is not understanding that the description of an ability or spell is its rules, to a greater degree in 5th Edition than any edition before. “I can’t find the rules for Sneak Attack.” or “I can’t find the rules for Stunning Strike.” or “I can’t find the rules for [a particular spell].” are all things you hear pretty frequently, sometimes with the explanation of everywhere they looked: in the combat rules, in the rules for spellcasting, in the Monster Manual, the DMG, et cetera.

Sometimes it’s apparent their approach was pretty exhaustive. They might have literally looked through all three core books trying to find the rules for this one ability.

The one place they don’t look is under Sneak Attack, or Stunning Strike, or under the heading of the spell they’re confused about. I mean, they read it, or at least looked at it. They figured out what it’s saying, generally. And then they went looking for the actual rules, and failed to find them.

Usually this question will be accompanied by a situation that arose at their table, where a player tried to do something (like, let’s say, stun a Fire Elemental with Stunning Strike) and the DM wasn’t sure it would work. So they went looking for the rules governing it, and when they couldn’t find it, they just ruled on the fly based on their understanding of how it should work, then went looking for a definitive answer for next time.

If you’re thinking about the text below the heading of a special ability or spell as nothing more than a sort of general description in game terms of what happens, not only are you going to be looking in some nebulous elsewhere for clarity about its exact limits and implications, but you’re not going to have any reason to prioritize the text over your own, simpler version of the general description.

Stunning Strike and Sneak Attack are frequent sources of confusion, because they’re two abilities that people tend to think of in very precise “in-game” terms that aren’t really portable to different situations: you’re hitting someone in a vulnerable spot, like a nerve cluster (Stunning Strike) or organ (Sneak Attack). That’s a great capsule description of what the ability might represent happening in-game in a particular (fairly frequent) case, but it’s not even a general description of what the abilities do. 

Stunning Strike lets you stun someone. That’s what it does, in shorthand terms. Sneak Attack lets you hit for more damage. That’s what it does, in shorthand terms. And the rules that govern each are neatly self-contained within the short paragraph or two of text that describes each ability. And if you don’t know this, if you don’t get this, then you’re likely to look at those short paragraphs and say, “There has to be more to this, but where?”

A buzzword people use a lot to describe 4th and 5th edition is “exception-based design”. The actual rules describing what characters can do in general are relatively few and fairly simple. An individual character, be it a heroic adventurer or monster, is a set of exceptions to those rules. Fighters replace some of the normal rules of combat with rules that represent better fighting. Rogues replace the normal skill rules with rules that represent better skill. Wizards replace all sorts of rules, temporarily, using spells.

A buzzword that should be used more often than it is for describing how 5E works is “procedural logic”. What people frequently write off as a general description or just overly wordy stuff designed to sound formal and technical is not just actual rules, but in fact a set of instructions that tell you how and when something applies, and what it does. The vast majority of apparent ambiguities in how things work and disputes over what abilities can do could be settled simply by looking at the abilities in question and following them, word by word and line by line, like you’re a computer executing a compiled program.

For example, let’s take what is both a common special ability and a common cause of confusion. About half or so of the character classes in the Player’s Handbook gain an ability at level 5 or 6 that’s called Extra Attack.

Now, you probably don’t even need to read the rules for Extra Attack to understand what it does, especially when you notice that the classes that get it are the “warrior” ones like Fighter and Barbarian. It lets you attack more often. It gives you an extra attack. The general rule is that you can attack once per turn; this lets you attack again.

And if that’s your guess, you’ve nailed it. Here’s the actual text for the feature:

“Beginning at 5th level, you can attack twice, instead of once, whenever you take the Attack action on your turn.”

Oh, boy. They crammed what looks like some extra verbiage in there, but even if you don’t understand the formal meaning of “take the Attack action”, you can tell that Extra Attack does what it says: it gives you an extra attack. That’s all you know and it’s all you need to know.

Except if that’s all you know, you’ll run into situations where you’re not sure how it applies. If you’re just thinking, “I can attack twice”, what happens if you cast a spell that is kind of an attack, or even definitely an attack? Do you get to attack again? Could you cast two such spells? What happens if you take an opportunity attack, do you get another one? If you have two copies of this ability (after all, multiple classes have it), do you get an extra attack from each one?

Now, if you have any familiarity with the rules of the game and maybe even if you don’t, sitting there with the actual text staring at you, the answer to all those questions might be pretty obvious. No, you can’t more attacks by taking the ability more than once because it doesn’t give you an additional attack in the first place, it lets you attack “twice instead of once”. No, you can’t mix it with spells or use it when taking an opportunity attack, because it only applies “when you take the Attack action on your turn.”

The text I excerpted above for the Extra Attack ability is not just a clunky way of describing the bare fact that you get to attack more. They are exact instructions for how the ability is applied. IF you are 5th level, and IF you take the Attack action on your turn, THEN in place of the one attack you normally get, you can make two instead.

Procedural logic. We have conditions that must be met, and if those conditions are met, we process the ability, which provides us with an exception to the normal rule that you can attack once by taking the Attack action.

I’ve seen a load of questions about how this one ability interacts with this, that, or the other thing… but they can all be answered just by parsing the one line of text I excerpted above.

Let’s take the Monk’s Stunning Strike. The rules for that read, in their entirety:

Starting at 5th level, you can interfere with the flow of ki in an opponent’s body. When you hit another creature with a melee weapon attack, you can spend 1 ki point to attempt a stunning strike. The target must succeed on a Constitution saving throw or be stunned until the end of your next turn.

That’s all the text under the heading of “Stunning Strike”. It is, in fact, the complete rules for how Stunning Strike works. There’s a bit more descriptive text. I mean, “interfere with the flow of ki in an opponent’s body” is not a precise game term, but a flavorful description of what’s happening. It does at least do the job of telling us that no, this ability does not depend in any way on the target having identifiable nerve clusters, which should stop a lot of arguments that it shouldn’t work on creatures without nerves (though, it doesn’t).

Still, it’s pretty straightforward: if you are 5th level and if you hit another creature with a melee weapon attack, then you can use the ability, the effects of which are—again—contained within this short block of text.

Now, you have to know what a melee weapon attack is and what constitutes a creature and how a Constitution saving throw works and what it means to be stunned, and those things are defined elsewhere (poorly, in the case of “melee weapon attack”, though I’ll get to that later). But they’re all general rules. You don’t have to look them up in relation to the Stunning Strike ability. You don’t have to know some secret intersection of those rules and this rule.

Got a question about how and when you can use Stunning Strike? You can answer it with this text.

“Can I use Stunning Strike when I hit with an opportunity attack, or only on my turn?” It doesn’t say anything about your turn, so it doesn’t have to be your turn. It doesn’t say you have to use the Attack action, so literally anytime you hit with a melee weapon attack it’s good.

“Do I have to declare I’m using it in advance?” It doesn’t say that, so you don’t have to.

“If I have a spell that lets me make a melee weapon attack, can I stack its effects with this?”

It says “when you hit with a melee weapon attack.” If you hit with a melee weapon attack, then you can use Stunning Strike. Nothing else matters, not how the attack happened or what other awesome things the attack does.

“Can I use this on a Fire Elemental?” It doesn’t say particular types of creatures, it says “creature”, which all monsters and characters recognized by the rules are considered to be. Fire Elementals are creatures; ergo, you can use Stunning Strike on them.

Now, I have to point out that “It doesn’t say you can’t, so you can.” is not a great principle to follow, generally. But when you’re parsing out the rules of an ability, any requirement it doesn’t explicitly lay out does not exist. Any limitation it doesn’t explicitly lay out does not exist. If you have a question about what does or does not work, just follow it through procedurally.

Let me lay out an example that’s going to get a little complicated and nerdy, but it just shows how to apply procedural logic to determine what is and isn’t possible, what does and doesn’t work together.

Say you are a 5th level Monk with Stunning Strike and the Extra Attack feature, and you have learned through some means a spell called Greenflame Blade. This spell’s description reads, in part:

“As part of the action used to cast this spell, you must make a melee attack with a weapon against one creature within the spell’s range, otherwise the spell fails. On a hit, the target suffers the attack’s normal effects, and green fire leaps from the target to a different creature of your choice that you can see within 5 feet of it. The second creature takes fire damage equal to your spellcasting ability modifier.”

You also have the Monk Martial Arts ability, which includes this line:

“When you use the Attack action with an unarmed strike or a monk weapon on your turn, you can make one unarmed strike as a bonus action.”

Wow, that’s a lot of stuff going on. You’ve got a spell that lets you make an attack and then zap someone else with green fire, you have the ability to attack twice every time from Extra Attack, and you have the ability to make a bonus unarmed strike! That’s a whole lot of moving pieces to sort out, but it sounds like there’s going to be a whole lot of attacking going on, right?

So it’s our turn. We cast Greenflame Blade, and “as part of the action used to cast this spell”, we “must make a melee attack with a weapon”.

Let’s say that we’re using a short sword.

A short sword has a blade, and you might think this is important because the name of the spell is “Greenflame Blade”. You might even spend some time looking all over the place to find where it says that Greenflame Blade does or doesn’t allow non-bladed weapons, but the text is right there: it says “with a weapon”. Type unspecified, so any weapon will do.

So we use our action to cast the spell, and then we immediately make a melee attack using the short sword. We hit, let’s say, an Ogre Zombie. At this moment in time, we have hit with a melee weapon attack.

That means we can use Stunning Strike.

But wait, can we stun zombies? Nothing in the rules for Stunning Strike says it only works on certain creatures. We can check the “stunned” condition, but spoiler warning: it doesn’t list exceptions, either. None of the conditions do. That’s because conditions are a general rule, and in exception based design, the exceptions are self-contained. This means if the combination of ogre and zombie is immune to being stunned, it will be found in the listing for Ogre Zombie. We check, and it lists only one condition it’s immune to: poisoned. They can be stunned.

So we use Stunning Strike. Depending on how well the Ogre Zombie rolls, it may or may not be stunned, but we have used Stunning Strike. This brings me to an important point: imagine that the Ogre Zombie was immune to being stunned. This would not mean you can’t use Stunning Strike on the Ogre Zombie; it would mean the effect would be negated. This might seem like splitting hairs, but it does help you understand the flow of things better. The “program” we’re following continues until we hit an exception.

But, another thing is happening. Greenflame Blade provides another thing that happens when this attack hits: fire leaps to another target (let’s say a Zombie Donkey). And with that, we are done processing Greenflame Blade.

Can we use Extra Attack? No. Even though we attacked, we took the “Cast A Spell” Action, not the “Attack” Action.

Can we use the bonus attack from Martial Arts? No. Again, we did not take the “Attack” Action.

So the next turn comes and this time, we decide we want to use our attacks. We take the Attack Action, though we probably just say, “I attack the Ogre Zombie,” just we didn’t actually say, “I take the Cast A Spell Action” last turn, we just went straight to “I cast Greenflame Blade.” We are taking the Attack Action, which lets us use Extra Attack, allows us to attack twice.

We attack the Ogre Zombie twice, and let’s say we missed both times. Can we still use Martial Arts for a bonus attack? Yes. Greenflame Blade and Stunning Strike “proc”, as they say, on a hit, but Martial Arts doesn’t mention having to hit, only taking the Attack action and a requirement that we be using an unarmed strike or a monk weapon.

“Monk weapons” are defined elsewhere in the text of the Martial Arts ability, and they do include short swords, so we’re good. We make our bonus attack.

Can we apply Extra Attack to this to make two attacks? No. A bonus action granted by a special ability is not the same thing as the Attack action.

Of course, if you’re familiar with the Monk class, you know they’ve got the ability to make two bonus attacks if they want to, resources allowing, but there’s enough going on here for you to get the idea.

Any place where people are discussing the game, you’ll see questions about these kinds of interactions repeated again and again. Can a Paladin combine a spell like Searing Smite with their Divine Smite ability? Does a Paladin have to declare the use of Divine Smite in advance? Can a Rogue Sneak Attack even if the target is aware of them? Can a Rogue Sneak Attack if they hit on someone else’s turn? Can you use an unarmed strike with a smite spell? Can you use it with Greenflame Blade? Can you use Greenflame Blade as an opportunity attack?

In every case, if you understand the basic game terms, then the answer to the question is contained within the text of the abilities or spells under discussion.

(For reference, the answers are yes, no, yes, yes, yes, no, and no.)

Of course, some of the game terms are better defined than others, and I really can’t claim to be telling you how to read a character ability if I don’t explain the taxonomy of attacks.

You will never see this spelled out in a rulebook, but almost all attacks in the game fall into one of four categories: melee weapon attacks, melee spell attacks, ranged weapon attacks, and ranged spell attacks.

Both melee weapon attacks and melee spell attacks fall into the supercategory of “melee attacks”, just as melee weapon attacks and ranged weapon attacks fall into the supercategory of “weapon attacks”, but you will never see something that is both a spell attack and a weapon attack, or both a ranged attack and a melee attack. These categories are explicit and exclusive of each other.

Because of this, we end up with some oddities, like an unarmed strike being considered a “melee weapon attack” even though it is not an “attack with a melee weapon”. This means you can use the Monk’s Stunning Fist when you hit with an unarmed strike, but you cannot use an unarmed strike with the Greenflame Blade spell. The spell specifies “with a weapon”.

If all of this procedural stuff and precise taxonomies doesn’t sound very magical or adventurous to you, don’t fear. You very rarely have to sit down and explicitly proc a thing, like, “I take the Cast a Spell Action. I use it to cast Greenflame Blade.” and so on. One of the reasons that people stick with the shorthand version so often is that so often, it works. It’s only really when there’s a question about whether or not a combination of things work together that you have to sit down and process it step by step, and once you’ve done that, you know.

Now, this post has been talking about how to read and understand an ability like Stunning Strike or Extra Attack, and I’ve mostly been framing it in terms of not understanding the logic of the rules because it’s never been explained. There is another thing that is sometimes at play, sometimes, in the nagging insistence that surely there must be additional rules for Sneak Attack, and that’s the belief that whatever the player is proposing is too ridiculous to possibly be allowed.

I’ve got a whole post about that topic that I’ve actually been prepping for a while. I realized as I was writing it, though, that this one really needed to go up first for it to make any sense. For now, I’ll give you the shorthand version of the next post’s message, which is: never change the rules mid-game to make them make more sense. 

I’ll tell you why next week.


Spherical Goblins: Never Metagame I Didn’t Like

Tanks For The Memories

I saw a thread on a D&D forum recently where someone was talking about making a defensive Fighter build, but said they worry about what they called “the standard tank problem of ‘if I can’t harm it, then why focus on it?'”

A little bit of background here to get everyone on the same page: the concept of a “tank” comes from computer games based (through winding paths) on Dungeons & Dragons. In Dungeons & Dragons, the Fighter’s high AC and Hit Points has always meant that other characters would tend to hide behind the Fighter, especially during lower levels.

In a computer game where there’s no human mind directing all the monsters and minions you fight, this is represented by giving Fighter-type characters abilities that allow them to attract a monster’s hostile attention—“aggro” in the parlance—and hold it better than more vulnerable characters. In simple terms, a tank is a higher priority target for the computer than other characters are. The tank’s two defining abilities are to be attacked more often than their teammates and to survive more attacks than their teammates.

In D&D, a human mind is directly responsible for directing the “aggro” of any NPC enemies the party fights, so all you really need to have in order to tank is the second clause: the ability to survive more attacks. Or that’s the way it’s supposed to work.

The “standard tank problem” described by this prospective player is the notion that in the absence of any mechanical incentive for monsters to attack the Fighter, game theory demands that they don’t. The odds of hitting the well-defended Fighter are lower than the odds of hitting another target. It will take more hits to drop the “Fighter than anyone else on the field, a side’s ability to inflict damage does not decrease until a member of that side is taken out of the fight, and it decreases fastest when characters optimized for doing damage rather than taking it are knocked out first.

This is an example of the law of unintended consequences in action: in making it so that your character is the best at taking hits, you become the least attractive target on the battlefield.

This was considered a big enough mechanical flaw in the era of 4th Edition that they did attempt to import some computer-game style “aggro grabbing” mechanics that complicated battle but made it easier for tanky warrior characters to do their job of standing between the oncoming horde and their squishier friends.

Some of those mechanics survive in 5E as an optional rule in the DMG and an equally optional feat in the PHB, but I’m not going to talk about those rules or their effectiveness. Instead I’m going to talk about the mindset that leads to the situation where they seem necessary, and how to avoid it.

What’s A Meta For?

The “metagame” is essentially the game of playing a game. A boxer who notices that their opponent has trouble keeping their left glove up because of an old injury that’s bothering them is metagaming the sport of boxing, as is a boxer’s manager who cherrypicks opponents to build their fighter’s confidence and public profile.

Though you can game a system for many purposes, metagamers typically analyze the rules of a game, the social practices of playing it, and even the context in which it’s played in order to figure out how to achieve their goals within the game.

When people talk about metagaming a roleplaying game, they usually mean taking advantage of their mastery of the system in order to first build the most mathematically optimal character and second to use that character to overcome all challenges they encounter while playing it. My previous column touched on the problem of thinking there’s a single path to “optimizing” a character, alluding to metagaming in the process.

A lot of players will metagame combat in particular, out of some combination of desire to crush their opponents decisively because it’s awesome and fear that if they don’t leverage their characters’ abilities to the hilt, those characters will die and they will lose the fight and fail in their quest. They see metagaming and playing to win as being interchangeable concepts, and don’t understand why anyone would not play to win.

I will argue until I’m blue in the face against the idea that such metagaming is the only way to play the game, but I’m not overly concerned with players who do it. The problem comes when Dungeon Masters match a player’s metagaming with metagaming of their own. This is what creates the supposedly standard problem for tanks. DMs who metgame in the same way that players do and for the same reasons will inevitably deform the game system, breaking down the tacit assumptions on which the combat system is based and robbing it of any hint of realism.

The Conga Line of Death

A textbook example of how two-way metagaming deforms things is a phenomenon of the 3E/d20 system known as the Conga Line of Death. Internet personality Spoony explains the concept in a video.

The basic idea of this: since you get a mechanical advantage to hit enemies when you have an ally on the other side of them (called flanking, in the system), the “correct” strategy is to always flank.

And since you can still form a flank against an enemy that is flanking you, once someone has you flanked you’re halfway to flanking them back; it just takes one ally moving into position opposite you on the other side of a flanking enemy and now the two of you are flanking that enemy.

And now that enemy is halfway towards flanking your ally, so another enemy moves into position, and so on.

If this continues, then before too long the entire battle will consist of people standing in a line, alternating between your side and their side. A battle that is 5 people against 5 people somehow becomes a series of fights that are all two-on-one fights.

The first time I saw someone watching this video, I was fascinated by the fact that Spoony is describing the concept with a bit of frustration, but he treats the Conga Line of Death as an inevitable outgrowth of the rules, because “you always want to flank.” You never don’t want that advantage. There’s a sense of, “This is terrible, but what are you going to do about it?”

Despite Spoony’s air of resignation, the Conga Line of Death is not inevitable. Probably more groups have played 3E-style games without it happening than have played it with it happening. Or maybe it happens once, everyone sees how ridiculous it is, and never again. It’s just that in the circles in which it does happen, it will keep happening. Because the logic that drives it will lead to the same conclusion in every situation where there are enough combatants, because everybody involved wants that mechanical advantage of flanking.

But all it takes to break the deadlock that creates the Conga Line is for one person in particular to abstain: the DM. If the DM controls one entire side of the battle. If the DM isn’t in a dancing mood, the conga line can’t happen. In a very real sense, players can’t create a Conga Line situation. It’s what happens when the DM matches them flank for flank, meta-tactic for meta-tactic.

People who defend this type of metagaming by the DM often make the argument that it would be unrealistic to not metagame, because everybody in the fight knows that it’s likely to the death and the best way to survive is to win, and the best way to win is to use the best tactics. It’s not really metagaming to take advantage of rules like flanking because the characters in the game know how flanking works, too.

But I think it’s obvious to everyone that the Conga Line of Death is not a realistic depiction of how a skirmish among two groups of about half a dozen or so combatants would break out and then break down. People would never line up like that for a fight.

As a DM, you can’t prevent this by telling the players, “That wouldn’t happen.” when they try to grab the first or second flank in a sequence. The rules say they can, so they can. And so can you. But you don’t have to. If you know that it wouldn’t happen, your job is to make sure it doesn’t. As a DM, your number one power for preventing the game from devolving into a ridiculous farce is to not do things that wouldn’t happen.

So I Just Have To Let Them Flank Me?

Yes, you do have to let the players grab mechanical advantages without concern for verisimilitude, the suggestion of reality. It’s not their job to provide that kind of feeling. It’s your job to do so. The rules are there to provide them with knowable, predictable limits of what they can and can’t do. The fact that they’re knowable and predictable means they’re game-able. The fact that they can be gamed means that some people will game them. But you don’t have to game them back.

A lot of times, when someone goes on a D&D forum asking for DM advice about what to do with metagaming players, particularly ones who only seem to be metagaming, not roleplaying or engaging with the game on any level except “How do I win?”, the advice is that you have to match them move for move or they’ll walk all over you.

Unfortunately, this tends to make the game less fun for everybody, and it also reinforces the mistaken notion that this is just how the game is played. It not only reinforces the behavior in the first player, but it teaches everyone at your table that they have to play the metagame to the hilt, even if it comes at expense of playing the actual game.

The truth is that as a DM, you are really not playing the same game as the players are, which means you shouldn’t be playing the same metagame. There are some goals that everybody at the table shares. Having fun is the obvious one. Telling a story is another possible one.

But the players are also trying to complete a quest, and/or advance their characters’ goals, and/or amass riches and power through daring exploits and/or living out their heroic (or villainous) fantasies. You as DM are not trying to do any that. You’re trying to provide an environment where they can do all that while being entertained and challenged. The balance of challenge might vary from group to group, and there’s no right level except the level that provides the most fun for the group (yourself included). But even acknowledging that challenge is one of your goals, you’re still not playing the same game as they are.

Unlike a player who is playing to win, a DM’s metagame doesn’t—or shouldn’t, at least—consist of things like knowing that if you combine this ability with that rule and take advantage of an exception provided by interference from a conflicting interpretation of an optional rule then you can do an extra 37.5 damage per round. A DM’s metagame consists of taking the knowledge that Player A is motivated by the thrill of combat, but Player B likes a good mystery and Player C just wants to goof off and coming up with an encounter that has something for everyone.

A player who builds a “tank” character by making a Hill Dwarf Fighter with the defensive fighting style and maximum Constitution, AC, and HP for level one is using metagame knowledge to create a character who can fulfill a certain role on the battle field: being a living meat shield for the rest of the party to hide behind.

The DM’s metagame should take into account that this player created this character to achieve a particular experience in combat; i.e., the thrill of being in the thick of things, laughing as sword and axe blows slide off their armor with nothing more than a bruise underneath. This knowledge helps the DM “win” by allowing them to run the game in a way the player will enjoy more.

A DM who choose to ignore the Dwarf in order to pile on the weaker party members and obliterate them so they can then dogpile on the defense-optimized Dwarf safely is playing the player’s metagame.

Now, is it realistic for the party’s foes to focus their “aggro” on the one target they’re least likely to meaningfully affect, in favor of immediately changing the course of the battle by taking out the unarmored wizard in a single round?

Of course it is.

In terms of in-game “reality”, none of the little people on the battle grid are actually little or on a battle grid. None of them are watching the battle with a bird’s eye view. None of them are seeing things in terms of polite, orderly turns during which they can assess, over the course of a minute or so, the possible targets and note their odds of a successful attack. They don’t know the D&D rules of combat, only the reality that they are intended to—with some help from the DM—loosely emulate, and the reality is that there’s big armored dwarf right in front of them who is the most obvious, most immediate threat, too dangerous to be ignored.

The weakling in the robe, the skinny person with the knives? The DM knows that they are a Wizard and a Rogue, two of the most dangerous people to ignore on a battlefield. But most people in the world who don’t wear heavy armor or carry big weapons aren’t particularly dangerous in a fight, and the average NPCs won’t know on sight that they’re looking at exceptional individuals who are the protagonists of a heroic adventure story.

The trick is actually to let it go. Let them have their metagame and focus on your own. Once you do that, all the things that seem inevitable—the “standard” problem for tanks, the “unavoidable” Conga Lines—just sort of fall away.

Death of the Ego

The hardest part about adopting a better attitude towards metagaming as a DM is the feeling that when the players are winning, you’re losing. Very few DMs consciously believe this, but when characters under your control are dying and the people across the table are crowing about how easily they dispatched them, it can be hard to let go of the idea that you yourself are suffering a defeat. It doesn’t help if the player is obnoxious about it and talking about “owning” you, of course.

The trick is to remind yourself as well as them that you are not the bad guy. The best way to do this is to celebrate with them. Say “Nice one!” instead of “Ouch!” or “You all are really kick butt.” instead of “I’m getting my butt kicked here.” Involve yourself in what the players are doing.

Ironically, the more you separate yourself from the NPCs in an encounter, the easier it becomes to consider how the battle looks from their actual point of view instead of what the winning move in the metagame would be. This also discourages even more objectively deplorable types of meta-DMing, such as shutting down a player’s illusion or charm spell or clever non-magical trick because “Sorry, I just can’t see this person falling for that. It’s not realistic that they’d be caught off-guard.”; i.e., you feel like a chump when you have to go along with something that you know is a trick.

I’ll likely talk more about that—particularly as it applies to illusion and trickery—in a future column.

Don’t Just Let Them Win…

…but remember it’s not your job to make them lose, either. Give the metagaming players challenges that give them the benefits of being awesome, rather than challenges that take away their awesomeness. If someone made a character who really is the best at killing spherical goblins in a vacuum, throw a lot of things at them that they can kill. If someone is enjoying always having the answer to everything (because they know the books inside and out), make that knowledge relevant some of the time instead of trying to put a mechanical stranglehold on it (“You don’t know that. Your character wouldn’t know that. Make a knowledge check.”) or constantly pulling gotchas and switcheroos like trolls that are healed by fire instead of being vulnerable to it.

Now, you can’t let the game become about celebrating the one player who has mastered the metagame the most to the exclusion of everyone else, of course. As a DM, your metagame is figuring out what every player is looking to get out of the game and, within reason, helping them find it.

Related: sometimes people will lament that you can’t balance combat when one character is heavily optimized for fighting and the others aren’t, because anything that challenges the overpowered character will wipe the floor with everyone else, while anything the others can handle will be wiped out by the overpowered one.

This is actually less of a problem in 5E than many people think, because of the care that went into the system’s design. But even to the extent that it’s true, it’s actually pretty easy to balance encounters for characters of disparate power levels, whether because one’s super optimized, or you rolled for stats and someone got lucky, or because they’re all different actual levels.

The key is to remember your metagame as DM: you’re not playing the combat to win, you’re playing it to provide an experience. And no one said you have to give everybody the same experience.

Spherical goblin-type game theory says that the PCs’ enemies should always focus their fire on the same target, preferably the one at the greatest intersection of “easy to kill” and “does the most damage per round”. But that’s not a realistic model of how the fight would actually go down.

Again, in real life, the most obvious, immediate threats grab the most attention first. And in stories, what usually happens when one person in the group is the best warrior is that someone on the other side recognizes this and chooses to engage them, either for the protection of their friends or because “At last! A challenge worthy of my skills!”

One of your goals is to challenge the players, yes. But it helps to think about designing adventures and encounters the way you would a similar thing in a video game; i.e., remember that the point of the challenge is that the players can overcome it.

Don’t Go Soft, Either

Finally, remember that this not about being a soft touch as a DM. Allowing players to metagame doesn’t mean you have to put up with argumentative rules lawyering or let them tell you how to run the game or dictate the terms of combat.

When it comes to interpreting the rules as written, I can be pretty hard line. I have the exceptions and house rules I allow at my table, but I know they’re exceptions and house rules. When it comes to what you can do and can’t do in combat, I’m very much a stickler.

I mean, if you’ve got a rad stunt that just doesn’t follow any particular rules, I’ll let you try it… if it fits the situation, doesn’t exactly emulate an existing ability that you don’t have, and it’s clear you don’t think you’re inventing a new rule that lets you do this cool thing at will. But don’t try to tell me that Quicken Spell should let you ignore the per-turn limits attached to spells with a casting time of bonus action because you think it stands to reason that it does. I’ll ask you to read what it actually says and then hold you to that.

If that example in particular means nothing to you, I’ll put it like this: allowing players to have their own metagame you don’t need to let players build “game breaking” characters or tactics that rely on questionable rules interpretations. If you don’t think the thing a player is trying to string together should work, you don’t have to give in. Deciding and enforcing the limits of what’s possible is part of your job as DM.

Also, just because you don’t automatically follow the same metagame tactics as the players doesn’t mean the rules that enable those tactics are off-limits to you. You can flank, if your edition includes flanking rules. You can decide that some of the enemies will gang up on a particularly troublesome hero (once the hero has had a chance to prove troublesome).

In particular, it’s worth remembering that you can always increase the ruthlessness of your tactics if a fight is going a lot easier than you expected… but it’s also worth remembering that there should be some easy fights. The fact that the dice are kind to the players isn’t something you have to actively counter. The law of averages will take care of that down the road for you.

But I don’t set limits within the rules on what players can “get away with”, nor do I try to match them meta for meta.

Remember: you’re the DM. If you won the game by killing the player characters, you could win it at any time. If you don’t just want to declare “Rocks fall, everyone dies,” you could have them accidentally wander into the dragon version of a chamber of commerce meeting. There’s no rule that says you can’t, only guidelines that say you shouldn’t.

You don’t do that kind of thing because it’s not fun. (Maybe funny, maybe once, but not fun the way the game is supposed to be.) And while working out the limits of what’s allowed and making use of that knowledge is a metagame for players, working out the limits of what’s fun and making use of that knowledge is the metagame for DMs.

Spherical Goblins in a Vacuum

There is a known bug in the human psyche, called the “quantifiability fallacy” or sometimes the “metrical fallacy”: we overestimate the importance of things that can be measured. The easier they are to measure, the more important we assume they are. And because the easily-measured metrics are the easiest things to test for and to brag about, their perceived importance just reinforces itself over time.

This problem crops up in how we handle just about everything: health care, corporate finance, sports, governance, even our military strategies and priorities. Everywhere you go and everything you do, people want to see numbers, numbers, numbers. Hard numbers, big numbers. This kind of thinking pervades pretty much every important aspect of modern life, but perhaps none as important as the fine art of pretending to be an Elf Bard rifling through the pockets of dead kobolds for spare change.

Yes, friends. For today’s Thing of the Day, you get a blog post about the dungeons and the dragons. This will actually be the first installment in a semi-regular feature, an ad hoc column I am calling “Spherical Goblins”, short for “Spherical Goblins in a Vacuum”. Some people who stand at the right nerdy intersection might understand that title immediately, as will people who pay attention to my twitter ramblings.

For everyone else, this post will explain it.

There is a joke—a whole species of them—about physicists solving some problem like doubling a dairy farm’s milk production, but then revealing that their solution only works “for spherical cows in a vacuum”. This references the fact that physicists often simplify the problems they’re dealing with by making assumptions that eliminate complications.

This brings me to the nature of what is often called “theorycrafting” in D&D, as it currently exists, as it applies to min-maxing or character optimization.

The concept of min-maxing is not a new one. You make a character who is only deficient in ways you can ignore and work around or just don’t care about, and strongest where it will have the most impact, or where it’s most important to you.

Min-maxing works best in games that let you gain more power in one area by taking a weakness in another, which are mostly freeform point-based games. It’s a form of metagaming (playing the game of gaming the game system itself, essentially) and it’s not terrible, in moderation and in and of itself.

Even at its worst—especially at its worst—it relies on assumptions like “I can ignore this area and focus on that one because the first area won’t come up.”

If you mess with those assumptions, the whole thing falls apart. A good DM can work with min-maxing to keep its effectiveness to a reasonable level. Of course, a good DM will also let players who just want to be awesome be awesome, also at a reasonable level. You very rarely want to actually match the players meta for meta, though that’s a subject for another column.

So min-maxing, in and of itself, is fine.

The current breed of “theorycrafting” that often surrounds it, though, is another story.

I talked at the top about the metrical fallacy: if we can measure it, it matters. Characters in D&D, particularly in any 21st century edition, can do all kinds of fabulous and fascinating things: talk to animals, change their appearance, create illusions, communicate telepathically, and… of course… they can kill things like goblins.

So, if you have a Gnome Ranger who can talk to small animals, produce minor illusions, and has a variety of survival skills and nature spells, and who can kill goblins with a magic mark and bow attack and a Human Warlock who can communicate telepathically and bring people to their knees with a word and change their appearance at will and who can kill goblins with a curse and a blast of eldritch power, and you have to compare which one is better, how do you do it?

It’s hard to put a value on most of their respective abilities, much less compare them to each other. The one thing they can both do is kill goblins, and as luck would have it, their ability to do so is already reduced to raw numbers! We can figure out their damage output adjusted for hit rate to get their damage per round, and settle the question once and for all of who is better at killing goblins.

At that point, you might feel like we’re one step closer to being able to compare them, or that we know who is better at one thing in one situation… but for those who are invested in creating a character who is objectively the best, that’s the whole comparison. When you know who does the most damage, you have your answer.

I mean, you can get more complicated, and many do. You can drag in how much damage each character can themselves avoid, mitigate, or heal in order to figure out who will stay alive the longest while killing goblins, but that’s still just a facet of how good they are at goblin-killing.

So that’s the goblins. Why are they spherical and in a vacuum?

Our theoretical goblins are spherical in the sense that we assume everything about them is simple. None of them are using unusual tactics or equipment, or exhibiting unusual behavior. Every turn they behave in a straightforward fashion that conforms exactly to whatever our game theorists think they should do.

And they are in a vacuum in the sense that we assume there is nothing interesting about the environment or situation in which they are fought. There is no terrain or ambient condition or external event that has any impact on anything.

The reason we keep to these assumptions is that if we don’t, it becomes harder to make the comparison between characters. If the battlefield is hard to navigate, the Warlock’s ability to teleport might give them an advantage over characters who can’t. If the battle is happening at a long distance, the range of the bow vs. the blast matters. If the goblins are riding mounts and using hit and run tactics, the question might become who is better at controlling them and pinning them down.

Given a certain set of circumstances, we can decide which set of abilities is more valuable in that circumstances. But we can’t compare their overall objective value without knowing not only which one is more valuable in each and every possible circumstance, but how much more valuable, and how likely that scenario is.

It is impossible to do so, which means it is infinitely hard to measure the objective value of anything other than direct killing (or not-being-killed) power.

And, that stubborn, blinkered thinking fallaciously insist that things that are hard to measure don’t matter as much as things that are easy to measure.

So the only thing that really matters is how good a character is at killing spherical goblins in a vacuum.

It doesn’t matter how clever a character is, unless that cleverness comes with damage dice attached. It doesn’t matter how charming they are, or how much many fantastical magical things they can do… except for the ones that do damage. Anything about a character that isn’t the thing that kills the most goblins the fastest in ideal circumstances is just stuff, just fluff… nothing that counts, nothing to concern yourself with, nothing to worry about.

This is the kind of thinking that I abhor, and that I think is toxic and corrosive to the hobby when it’s treated not as fun thought experiments but is handed down to the new and unsure as the way the game is supposed to be played, the way it must be played. In this semi-recurring blog feature, I’m going to be directly countering this kind of thinking with advice to both DMs and players about other ways to approach character creation, the game rules, and running games.

Sometimes I’ll be taking on the fallacious credos and sacred cows that are promulgated by the spherical goblin theorists. Other times I’ll take a more positive approach, offering good advice without any particular point to refute. In either case I’ll be sharing the wisdom of someone who has been playing D&D since the 1980s across multiple editions and through multiple media.