Good gameplay starts when the players want more they can have. Whether you’re designing a game system, an adventure, an encounter, running a combat, or adjudicating a trap, you’ve got to understand why you can’t let the players’ goals line up cleanly.

---

Don’t Let the Players’ Goals Align

So, I did this thing about Pass-Fail Dice Mechanics, right? About how they’re fine, but occasionally need some spicing up with risks, costs, and consequences, right? I ain’t gonna rehash the article. It’s right there. Go read it if you haven’t already.

I bring it up because a comment reminded me of this other game design principle that’s super important. It’s not just vital when you build a roleplaying game system, or hack one, but also when you build adventures and encounters. Really, every homebrewer Game Master needs to grok it.

In the Pass-Fail thing, I was talking about sweeping for traps, right? And I mentioned how there should be a risk that, if you fail to find a trap, you might trigger it. A rogue should risk taking a trap to the face every time he fondles an amazing chest or diddles a knob. Of course, the rogue ain’t great at taking anything to the face. That’s really more of a meatshield thing. Except that, D&D v. 3, the designers hardened the rogue against traps with abilities like trap sense and evasion.

Then came Paul. He’s a commenter who has been generously supporting my work on Patreon for a while, starting on my birthday some years ago, and who, therefore, is one of the people keeping this site alive. Thanks for that, Paul. Truly.

Paul observed that there’s another way to look at the whole trap-hardened rogue thing. Maybe, said Paul, the idea is that, when an adventuring party is faced with a seemingly empty but potentially dangerous situation, they have to decide whether to lead with the rogue or the tank, depending on whether they suspect they’re walking into a trap or an ambush. If the room is trapped and the party sends a tank, the tank gets spanked. If the room is ambushed and the party sends the rogue instead of the tank, he probably gets ganked.

Now, that ain’t a bad observation, and Paul’s no idiot. He’s smart enough to read my work, after all, and he’s clearly putting more thought into this crap than many roleplaying game designers do these days. But he’s also wrong and, and Paul, I’m as grateful to you for your wrong comment as I am for your support because it lets me talk about a mistake lots of designers and homerbrewers make. It’s also part of why lots of combat engines, not just the one in Dungeons & Dragons, tend to be more shallow than they appear. In fact, it’s why D&D and Pathfinder combats tend to devolve into either damage races or games of rocket-launcher tag.

A Right Answer Is the Wrong Answer

Let me start with this whole trap thing…

When the party’s standing on a threshold, considering what might be contained in the gorgeous chest they see across the room and what hidden dangers might be protecting said chest, what’s their goal? What is the party trying to accomplish?

I know lots of you are shouting lots of different answers, and I’m sure they’re all fine. Most of you probably said something like, “Find out what’s in the chest,” or “Explore the room,” or “Get the treasure.” Well, those are basically half right. But there’s something most gamers forget. Adventuring parties always have two goals. The other one is, “Don’t get dead.”

So, Paul says the party has to guess whether they’re facing a hidden ambush from camouflaged throat spiders lurking on the ceiling or else a spring-loaded scythe of disembowelment rigged to a pressure plate in the floor. If it’s the former, they’ll want to yell, “Close your mouth,” and shove Tanky McIronpants through the door. If it’s the latter, they’ll toss Sneakfoot Stickyfingers in the room to see what he can see. If they guess right, the person in the room is the best person to both survive the danger and clear the way to the amazing chest ahead.

That’s great from the players’ perspective, assuming they guess right, of course. But from a game design perspective, you’ve turned a potential tactical choice into a guessing game with a correct answer.

When the rogue can’t take a disemboweler to the gut like a champ, there isn’t a perfect solution to the problem. The rogue can disarm the trap, but if he screws up, he can’t deal with the attack and damage as effectively as the fighter. Meanwhile, the fighter can’t disarm the trap, but he can shrug off the damage or even avoid it altogether. That’s assuming that the fighter has better defenses across the board and more hit points.

This is actually another of those Double-or-Nothing type mechanics I recently discussed when I talked about haggling. If you send in the fighter, he’ll definitely get attacked, but the outcome won’t be severe. If you send in the rogue, he can make some die rolls. If he succeeds, no one gets hurt, but if he fails, he takes a bad hit because he’s not a tank and can’t take hits like one.

But this is also an example of a certain game design principle, which is what I really want to discuss.

Don’t Let the Player’s Goals Align Perfectly

Let me drop some of the comedy and clarify a bit lest some of you think I’m missing something. In the above situation, the goal isn’t really to get the treasure, while not getting dead. Modern roleplaying games have evolved past one-trap kills, and most roleplaying games aren’t trying to kill the characters at all. But there are still two goals operating in every adventure.

The first goal is whatever the adventure’s goal is. Rescue the princess, recover the Macguffin, escort the merchant, whatever. In most roleplaying games, that comes down to managing resources properly so you don’t get worn down before you accomplish the goal.

Imagine, for example, that somewhere in the dungeon is a guardian monster with the key to the princess’ cell secreted somewhere on its body. The party’s goal is to retain enough of their resources such that, when they finally find the boss monster, they can kill it before they themselves are killed. Or they’re forced to flee. Or they end up sharing the princess’ cell and waiting for another adventuring party to come along and pull off a mass rescue.

Looting an amazing chest in a side room, though? That satisfies a different goal. Probably. Maybe it’s about satisfying curiosity or gaining wealth. It doesn’t matter. The point is, though, that the chest might be trapped. Opening it, therefore, might risk the party’s ability to achieve the adventure goal. If the trap devastates the party, they might be screwed when they find the boss arena.

This is the sort of macro-level gameplay challenge most players and Game Masters rarely appreciate is even happening, even though it’s happening all the time. It’s hard to get people to grasp the concept of attrition as a macrochallenge because it’s broken up across a thousand tiny decisions, none of which, by themselves, fully determine the outcome. Which is why system designers need to build this crap in at the system level. Attrition is the invisible hand of the system designer, ensuring the players have to make some meaningful strategic and tactical decisions even when no one notices they’re happening and even when Game Masters and adventure designers aren’t smart enough to build anything on top of them.

This is also why it’s important that system designers actually understand this crap. Otherwise, they sabotage their own designs and rob all the meaning from them. It’s also why Game Masters and adventure writers should trust their systems instead of trying to rip out important, subtle elements they don’t understand. It’s why they should follow the system designer’s advice when building adventures and encounters instead of assuming they can do whatever they want and still get a good game.

But I digress…

The point is that, if the player’s goals are always in alignment, they’re not making strategic and tactical gameplay decisions; they’re just solving math problems. Or they’re trying to guess the right answer to a question with only one right answer.

Paul’s sneak-or-tank question is exactly that. It’s a guessing game with a single right answer. If you guess ambush, tank. If you guess trap, sneak. Honestly, I don’t think sneak-or-tank is actually a puzzle that comes up that often. Most of the time, when the party is worrying about traps, they’re facing doors and treasure chests and things. It’s pretty rare for an ambush to jump out of a treasure chest. Unless it’s a mimic. Which, by the way, is why mimics exist and why they look like treasure chests. Rogues are always the ones searching treasure chests and picking padlocks, and rogues can’t take mimics to the face. It’s good game design.

That’s why the core rules for traps need to create this tradeoff. There can’t be one right answer whenever you suspect a trap, even when the situation is as simple as an impressive chest that is very likely trapped. That tradeoff doesn’t just create gameplay, it provides the foundation on which more gameplay gets layered.

Imagine the party’s facing a door. They’re trying to decide whether to make the rogue search it for traps or just let the fighter open and take the potential trap to the face. Will they always make the same decision? What if the rogue’s already pretty hurt? What if the rogue’s out of Hit Dice? What if the fighter’s hurt and out of Hit Dice? What if it’s the beginning of the day and the party has all their healing resources available? What if it’s late in the day and the cleric’s tapped out? What if the rogue’s player is super risk-averse and the fighter is a gung-ho risk taker?

The secret sauce here is a thing called Internal Conflict.

What Do You Want When You Can’t Pick Everything?

Internal conflict arises when you have multiple priorities, risks, goals, desires, costs, or consequences, and there’s no way to get everything you want and avoid everything you don’t. Internal conflict is what makes choices choices instead of just solutions. When we’re talking about personal goals and desires, those choices reveal personality. When we’re talking about things like resource management and value propositions, those choices are strategic and tactical. In other words, they’re gameplay choices.

Are you willing to take a big risk for the chance to gain a big reward? Do you prefer to take the safer path? Are you willing to put yourself at a disadvantage in obtaining a long-term goal for the sake of accomplishing a short-term goal? How do you assign value to the unknown contents of a chest, the unknown amount of damage from a possible trap, and how much you’ll need every hit point in an unknown boss fight, some unknown number of encounters from now?

Game designers, module writers, and homebrewer Game Masters must avoid letting the player’s goals, desires, risks, costs, and consequences align too much so that the players are making strategic, tactical, and personal choices and not just solving puzzles.

The best way to illustrate this is with a little thought experiment about a kind of mini-hack that also reveals why so many roleplaying game combat engines, not just D&D’s, are a lot shallower and less engaging than they appear and why D&D, and Pathfinder have a tendency to turn into damage races or rocket tag. And also why spellcasting and ranged combat have totally ruined modern D&D combat.

Taking a Stance

Imagine with me that D&D’s Proficiency Bonus works a little differently. I’m actually going to offer a roughly balanced way you could experiment with this because I know lots of you like to take my half-assed, half-formed experimental hack proposals to the table. Honestly, this is one I’d take to my own table, and I’ve playtested a couple of similar mechanics in different systems. This ain’t a complete ass-pull.

Imagine that D&D’s Proficiency Bonus applies to Attack Rolls with proficient weapons, Damage with proficient weapons, Armor Class with proficient armors, and Saving Throws with proficient Saving Throws. To make this actually work, I’d reduce the base Armor Class from 10 to 8, and I would shrink every melee weapon’s damage die by one size. I would not change anything about Spell Attack Rolls, Spell Damage, Spell Save DCs, or Ranged Attack Rolls.

Trust me on all this crap. You do trust me, right? It doesn’t matter. The mechanics aren’t the point.

The rub, though, is that you don’t add the Proficiency Bonus to any of those statistics on your character sheet. Your Attack Rolls, Damage, Armor Class, and Saving Throws all get listed without any Proficiency Bonus at all.

Instead, at the start of each of your turns in combat, you pick a stance: Attack, Damage, Defense, or Saves. Until the start of your next turn, your Proficiency Bonus applies only to that one specific thing you chose and nothing else.

Noodle that, then read on, MacDuff.

Two Goals in Every Fight

In every combat, the players have two goals: they want to defeat their enemies, and they want to stay alive. Many have observed that, ever since AD&D 2E introduced an actual, good, balanced, tactical combat engine as an option in the 1995 Player’s Option: Combat and Tactics supplement, it has become increasingly true with every edition that the best way to win any fight is to kill the enemies before they kill you. In the 4E era, the running joke became, “The best condition to inflict on an enemy is Dead.”

In reality, the issue I’m about to describe plagues a lot of turn-based combat systems. Hell, most video games featuring turn-based combat struggle with the same problem that tactical roleplaying gaming combat engines have been struggling with since the designers realized that maybe combat should be fun instead of a punishment for not avoiding fights good enough. In his critique of Darkest Dungeon, Joseph Anderson said that you can sum up most turn-based battle systems as, “Make sure you spend a few turns healing.”

The problem is that most roleplaying games don’t offer players many ways to pursue their goal to not get dead. They have plenty of ways to defeat the enemy, but in-combat healing is really the only real defensive option outside of slapping on a few buffs. There’s nothing you can do, actively, to keep yourself alive, let alone prioritize it.

Well, there’s that Dodge thing, right? You can spend a turn with your thumb up your ass doing nothing in return for a +2 Armor Class bonus. It does technically prioritize survival for one round, but it also extends the combat by one round because you’re not dealing damage that round. So, the attack you might have avoided with that bonus? It’s just going to happen next round.

There are actually a scant few situations where Dodge is a useful action to take, but Dodge is such a crap option most of the time that players don’t even see it as a possibility most of the time. Worse, there are ways to make Dodge a viable, useful, smart option. You just have to make math work.

Consider, for example, that lots of monsters have an attack that spikes damage that they can only do once in a rare while or that has to recharge, and those attacks generally do about three times as much damage as their normal attacks, all else being equal. If you could predict which round out of every three the monster was going to throw that big attack, Dodging in that round would be smart. You’d avoid triple damage at the cost of extending the battle by one round. Since those attacks often carry nasty rider effects, Dodge looks pretty good if you know when to use it.

But this ain’t just about providing some defensive options, which you absolutely need to do. There’s just nothing defensive to do in D&D that’s worth a damn, so the only question you get to answer is how you deal damage every round. That’s why D&D’s combat engine can feel a mile wide and an inch deep.

But my Stance Hack above does something more. It doesn’t just give you a defensive option; it gives you a resource that you can either spend on offense or defense, but not both. Moreover, if you spend it on offense, you get to choose between accuracy and damage. If you spend it on defense, you have to predict whether you’re going to need more Armor Class or more Saving Throw, which means you have to pay attention to what the monsters have been doing. And maybe they switch tactics on you because they see your stance.

In other words, you get a bonus you can’t apply to everything and have to decide, from one round to another, what’s your highest priority.

That’s why, by the way, I nudged everything down a little to make room for the Proficiency Bonus instead of just making a bonus on top of everything. Without the Proficiency Bonus, you want everything very slightly underpowered so the Proficiency Bonus can bring one stat into its normal, balanced state, or even make it slightly overpowered as the Proficiency Bonus increases with level.

See, I want the players to think of it as a bonus, but mathematically, I want the players to actually choose the one thing to not risk or penalize. Again, don’t worry about that.

What you should worry about is why I told you not to futz with spellcasting or ranged attacks. Notice that I told you to leave the Proficiency Bonus in place on Spell Attacks, Spell Save DCs, and to not touch Ranged Weapon Damage and Spell Damage. That’s because the rule is that spellcasters and archers have to adopt special stances called Spellcasting and Archering. Or whatever. They get to keep their normal offensive output, but they can’t ever prioritize defense while attacking. By the way, their Saving Throws and Armor Class should be devoid of Proficiency Bonus.

Does that seem odd? Does it seem like a violation of the whole two goals thing? Well, it isn’t because spellcasters and archers already get to prioritize defense automatically when attacking. Melee attackers have to put themselves in danger to deal damage. You can’t make a melee attack without being in a position where you can be melee attacked back. Ranged combatants can attack from anywhere. They can attack from, say, behind cover. Or they can let their melee buddies screen them from danger.

Once upon a time, ranged combat and spellcasting had a lot of fiddly little restrictions. Targeting was a pain because creatures on the battlefield, friend and foe, used to provide cover. Attacks of opportunity made it easy to get locked down and hard to escape a lockdown. Spellcasters had limited access to armor and defensive options. Action economies worked differently. I can go on and on here. All of that offset the fact that any attack from range is better than any attack from melee from a survivability standpoint.

But, over the years, careless game designers who wanted to keep players happy and streamline the game mechanics kept softening or outright removing the restrictions. It got easier and easier for archers and spellflingers to fling spells and arches with impunity. If you’ve ever caught one of my swear-filled, spittle-flinging rants about kiting in roleplaying games, now you know why. Kiting shouldn’t be possible, not just because it’s utterly ridiculous, but also because it literally breaks the tactical archery game, as well as making everyone who brings a sword to a bowfight look like an idiot. That’s even though dude with sword is still the single most popular character option at the D&D table.

So, to prioritize defense, an archer or spellcaster has to use the battlefield or their allies to provide cover or protection. But even then, they run the risk of getting locked down in dangerous melee without a way to defend themselves. That’s the price they pay for being able to attack with impunity.

Basically, just as the rogue needs to risk getting gutted by a disembowler to clear the way for his party, the wizard needs to risk getting ganked in return for being able to explode any corner of the battlefield he wants.