You read the title right.

If you have early access to my work — thank you for the generous support, by the by — and you’re reading this the moment it goes live, it might seem like I’m changing my calendar page twelve hours early or whatever.

Look, I’ve been stuck in Extreme Catch Up for two weeks thanks to my overambitious recovery and publication plan and the last three things I’ve published have been huge and heavy. I’ve done self-management, social management, psychology, commentary on the state of the entire roleplaying game hobby and my community’s place therein… that’s weighty shit. It’s exhausting to write and to read. Important, but exhausting.

Except that Untitled Bullshit thing. That was not important. Holy crap, what was I thinking?

Anyway…

We’ve got to remember that we’re Game Masters. In the end, that means our job is to write and run games. We’re supposed to be rolling dice, writing challenges, making calls, and building games. That’s what we’re focusing on today. I purposely picked out some stuff for this mailbag along exactly those lines.

Oh, right…

Hey kids, it’s time for my crappy monthly advice column, Ask Angry. Each month, I pull a few questions from my overstuffed mailbag and provide objectively, axiomatically, inarguably correct answers and slather them in sardonic… ism?

Hey, editor, look up the noun form of ‘sardonic’ before publishing this. There’s got to be one, right?

If you want the chance to get yourself some abuse-laden correctness to help you with a gaming dilemma, send it to ask.angry@angry.games. Just keep it brief, cut right to the point, give me explicit permission to call me whatever you want me to call you, and, if you really want to make sure I pick your question from the dozens I receive every month, know that I can take credit and debit card payments through Square and I accept PayPal, so, you know, just let me know what you think it’s worth to have me tell you whether dwarven women have beards or give you permission to backhand your players for trying to drink potions underwater.

Nico asks…

How can I use the Investigation skill in D&D 5E when “making deductions and figuring things out” is a player’s job and deductions based on specialized information the player doesn’t know like medicine or magic are already covered by other skills?

Nico, you scared the piss out of me with that subject line. On seeing 5E Investigation Skill, I was terrified this was going to be yet another e-mail asking to clarify when to use Perception and when to use Investigation or what Investigation is even for. I hate that question. It’s not a stupid question, mind you, and I don’t hate people for asking it. I hate the question because it’s the question that Jeremy Effing Crawford and his typewriter-banging monkeys should have asked themselves before they wrote the damned rules.

Seriously, 5E’s Investigation skill is like the infield fly rule and anti-racketeering laws: no one knows what the hell they are or how the hell they’re supposed to work. No one.

I hear your fingers going, commenter. Skip it. It was a joke. Except about Investigation and if you think you know how Investigation should work, you’re wrong and I can prove it, but I don’t care to, so don’t comment.

I don’t actually hate D&D 5E as much as everyone thinks. It’s a fun system; I get fun games from it. It’s actually pretty well-designed. But every system has its little problems. 3E has grappling, 4E has skill challenges, Pathfinder has all of Pathfinder, and 5E has advantage and disadvantage. But 5E also has Investigation. It’s just a really shitty, really badly-conceived skill that was carelessly tossed in there without any thought.

If you know your 4E lore, consider that Investigation is to Intelligence what Endurance was to Constitution. Go ahead and noodle that.

So I’m glad you’re not asking me to give you a good, logical framework for how and when to use Investigation or why you’d choose it over Perception or whatever.

But now, I gotta ask you, Nico, what’s the deal with the sarcasm quotes around making deductions and figuring things out? Do you not believe those are real things? Do you disagree that figuring shit out is a job for players and not dice? What do your implied finger quotes mean, dude?

Forget it. I’m just going to take this one at face value. Otherwise, I’ll end up yelling at you for projection and give you a two-thousand-word summary of cognitive behavior therapy best practices or some shit like that. My doctor keeps telling me I gotta calm down and my editor is noting that I’ve now written 350 words that don’t answer your question.

So, moving on then…

What I’m getting at is that this ain’t really about the Investigation skill. It kinda is, but it also kinda really isn’t. As you’ve pointed out, there are other skills that do the same thing as Investigation, but deal with more specialized situations. You’ve got Medicine, for example, and Arcana and Religion and History and, hell, even Perception actually fits into this same category. All of these are skills that let a player get information with a die roll. Usually, but not always, from the environment.

Personally, I think that from the environment should be an always thing and not a usually thing, but no one likes me when I’m saying correct things, which is why I have no friends. Because I’m never not saying correct things. And also because I’m really an asshole.

Perception is how you find information that’s concealed, somehow, from notice. Investigation is how you get information that’s totally not hidden, but which needs to be assembled from several different details. Medicine and Arcana and all the rest are how you get specific kinds of information using expert knowledge and techniques in the right situations. I could probably do a whole thing about how to do information gathering right using all these skills.

In the past, I’ve said that a die roll should always precede gameplay, never end it. I’ve also reminded people that gameplay ain’t rolling dice, but rather it’s making decisions and drawing conclusions. It’s figuring shit out. Let me give you a clear — if very exaggerated — example of the difference.

If a player rolls a Perception check to discover the murder victim wrote “It was Eustace of Murderville,” that ends gameplay. The question is answered, the mystery is solved, the puzzle portion of the adventure is a done deal. If a player rolls a Perception check to discover a few strands of red hair clutched in the murder victim’s fist, that leads to more gameplay. Assuming, of course, that Eustace isn’t the only ginger suspect and especially if Eustace claims to have a solid alibi.

See, a good mystery isn’t solved by finding a smoking gun, it’s solved by finding some wisps of smoke and a bunch of gun pieces and assembling a smoking gun in your own brain. It recognizing all the elements add up to Eustace of Murderville. An even better mystery is also recognizing some of the details don’t fit the others. Before you can be sure whether it was or wasn’t Eustace, maybe you need to disprove his alibi or prove the red hair was planted to frame Eustace.

You see what I mean?

You can’t let players use the Investigation skill to make the final conclusion. That’s why the description of Investigation as making deductions and drawing conclusions — oh, wait, were the quotation marks because you were literally quoting; that did not occur to me. I am not used to people using quotation marks to quote things on the Internet — that’s why Investigation’s description is kind of misleading.

In my free introductory module, The Fall of Silverpine Watch, I used Investigation — and Medicine and Religion and History — to let the players glean information by interacting with the environment. But, in each case, all they got was raw information. Just facts. They were useless until the players put them all together.

For example — and this is also a good example of what I mean by using Investigation to glean information that’s totally visible but has to be assembled with reason — for example, the players find a desk in a military officer’s bedroom. On it, there’s an unfinished letter, an unlit but half-burned candle, and an unstoppered and dried-out bottle of ink.

Intelligence (Investigation) DC 10. The ink was left open to dry and the strongbox was left unlocked. Whoever was working here was interrupted. The chair is not toppled, and the candle was not left burning, so they did not jump up and rush out. They probably intended to return later, but it looks like they never did.

In the same space, I also let a player Investigate a suit of armor to get information about its owner and to note the conspicuous absence of something they’d expect to be with the armor.

Intelligence (Investigation) DC 15. The armor is well-taken care of, but it is also well-worn. It is not ceremonial. Strangely, there is no weapon with the armor. Nor is there one hanging nearby.

The information provided is just facts. Someone was working here. Someone got interrupted. Someone left in a hurry, but not a panic. Someone may have taken their sword with them. Why did the person leave? What was the emergency? Why didn’t they come back? That’s for the players to discover, deduce, guess, or piece together. The information doesn’t solve the mystery, it just gives clues that could add up to a solution.

Since we’re talking about information presentation and mysteries and shit now, let me give you a little bonus advice. What you said about specialized knowledge is significant. Game Masters struggle with using specialized knowledge — and the skills to glean it — in mysteries and puzzles.

As a Game Master — or adventure writer or whatever — you have to provide enough context around esoteric knowledge that any layperson could use it to make deductions. That means, first, taking the time to make yourself a bit of a subject-matter expert — within reason, of course; you’re writing a murder mystery pretend elf game, not getting a degree in criminal investigation — and, second, it means you’ve got to translate that information so laypeople can draw conclusions from it. You also have to pinpoint the precise line between providing factual information and giving the player an answer.

Let’s say I want the players to conclude that a body was moved after the victim was killed based on the pattern of postmortem bruising. Obviously, typing that sentence means I already did the subject-matter homework. As I was building a murder mystery and a key part of the mystery is that the victim died somewhere other than where they were found, I did some research about how medical examiners might conclude that a body was moved.

All you dumbasses, by the way, who think the greatest thing AI can do is write flavor text for you or generate shit plots and adventures? You’re wrong. The best thing Grok and ChatGPT can do is answer a question like, “How can a medical examiner tell if a body has been moved after death?”

And look at that! The first of the seven things ChatGPT said was livor mortis. Also called postmortem lividity. Also called postmortem bruising. Holy crap, I rock.

Anyway…

So when someone with Medicine examines the body, I’ll describe the lividity so the player knows what the character sees, then I’ll explain its significance to a Medicine expert, and then I’ll stop before giving the player an actual conclusion so they can figure out the answer. Or fail to.

You note long, wide purple-black bruises all along the victim’s left shoulder, arm, calf, and foot like someone dragged a paint roller down their side. Livor mortis, you note, bruising that comes when a dead body has been left for a while. Without the heart pumping, blood pools wherever gravity pulls it. Strangely, though, the victim is lying on her right side, with her left side in the air. Turning the victim over, you find no bruising on the right side where the body’s been lying on the floor.

That’s a simple example and I’m sure you’re all thinking that the conclusion is so obvious that no player would ever miss it. Maybe so, but even letting the player be the one to say, “That must mean the body was moved after the killing,” is still more gameplay than feeding them the conclusion. Of course, the real gameplay comes in fitting that information in with all the rest to solve the bigger mystery. That’s where the gameplay is. That’s where the challenge lies. That’s where the players’ skills matter.

By the way, that is also the difference between helping the player tell a story about a forensic investigator and actually letting the player be a forensic investigator.

Now I’ve given you a bunch of really useful information, it’s on you to use your skills as a Game Master to make a fun, interesting mystery adventure or whatever from them.

Speaking of interesting…

Desperandos asks…

What makes for a mechanically interesting magical spell or magical item? How would you design one?

Wow, thanks for this easy question Desperandos. I said my brain needed a break and you’ve certainly given it one.

What makes for a mechanically interesting spell or item? Nothing. How would I design one? I wouldn’t.

Nailed that one in under fifty words! How about that, editor? What? Holy mother of crap would you pick a lane? Write less, write more, drop fewer f-bombs, that sentence is going to us sued; you’re totally fucking impossible to fucking please you fucking fuckwit. I swear to fuck.

Well… apparently, Desperandos, I need to expand on my answer a bit…

Seriously… mechanics don’t make things interesting. They’re the opposite of interesting. I’m not being funny there. See, I’m a big fan of the MDA Approach to Game Design. It explains the relationship between what game designers make to what players enjoy, but let’s skip the academic horseshit because I can do this easier.

Mechanics don’t come first. First, you need an idea for a spell or magical item. One that’s interesting to use or lets the players do interesting things or whatever. Look, I really hate the word interesting — it’s useless; don’t use it — but I don’t want to get any more sidetracked.

The point is, I can’t tell you how to design something interesting because you don’t design anything without a reason or a goal and it’s that reason that makes gameplay interesting. The mechanics are just how you pull off the goal in the game.

Someone came to me a few months ago because he wanted to give his players a ship and actually let them experience what it’s like to have a ship and sail it around instead of just using it as a kind of fast-travel thing.

I said…

So, what kinds of things do you want them to do with the ship? How much do you want them to interact with the ship? Do you want them to be traders? Do you want naval battles? Do you want them to risk being lost or stranded or marooned? Do you want them to hire and interact with a crew? Do you want them to customize the ship with skins they can buy from you with pizza or by washing your car?

That’s the shit that makes the ship interesting. What can you do with it? How do you interact with it? What problems can it solve? What problems does it cause? What are the costs of owning a ship and what benefits does it provide?

Nail that ship down — those interactions are part of gameplay Dynamics which are the D in MDA that comes between the Mechanics M and the A…enjoyment A — nail that ship down, then design Mechanics to make that ship happen. If you just start building Mechanics and trying to make them interesting to tinker with — maybe with tokens and cards and shit — you’re just making minigame gimmicks. You’re like Nintendo forcing you to blow into the microphone in every DS game.

Come up with the actual concept you want in the game. Then build the mechanics. Then pare them down to their simplest, most elegant, most practical, and least complicated form possible using as many existing mechanics as you possibly can.

You tell me you’ve got a great idea for a spell or magical item — tell me what they do in the world of the game — and I will help you get there mechanically, but ask me to come up with something mechanically interesting, I can’t help you.

Mechanics aren’t interesting.

Last question…

Calliope asks…

What are your thoughts about cards instead of dice in systems? Also, what are your thoughts on separating attack rolls from damage rolls?

Okay… joke aside…

Even though I said that mechanics aren’t interesting and you shouldn’t start with mechanics, it is possible to productively analyze mechanics. You have to do it right though. That means understanding that every game mechanic has both costs and benefits and it means being intellectually honest and rigorously dispassionate.

What do I mean? Well, I recently got myself into a discussion about how attack rolls were resolved in D&D prior to the turn of the millennium. Someone asked why so many gamers say that the THAC0 System is kinda clunky and unintuitive.

If you ain’t in the know, let me explain.

Back in the day, characters and creatures had a THAC0 score that measured their ability to fight. THAC0 was literally just the number you needed to roll on a twenty-sider to hit a creature whose Armor Class was zero. To Hit Armor Class 0; T-H-A-C-0. Get it?

I guess I should also explain that, back in the day, Armor Classes went down as they improved, not up, so a naked, screaming wizard had an Armor Class 10 and a steel-clad fighting man had an Armor Class 0. There’s a reason why it was done that way, but explaining it would be outside the scope of this out-of-scope digression from answering the question, and I’m just spitballing because I have no idea how to make my editor happy anymore.

Anyway…

When you Rolled To Hit which is what we called making an attack roll back then, you’d first subtract your target’s Armor Class from your THAC0 to determine the number you needed to see on your die to land a blow. Then you’d roll. Of course, if you had bonuses or penalties, you’d apply them to the die roll before comparing it to the score you needed.

Love old school games or hate them, whether your years of practice have left you able to do this shit in your sleep or not — I mean, I can do it easily because I spent fifteen years doing it — like it or not, you have to admit it’s a clumsy process. You also can’t deny that people’s brains hate to subtract. I mean, there’s literally a body of scientific research here, so denying it is just you being provably wrong. By kindergarten, children show a measurable aversion to subtraction over addition. The human brain subtracts slower than it adds. The average adult human is significantly less likely to use subtraction to solve a problem unprompted even when it’s the most efficient way to solve it.

When I pointed all of that out, several people countered with the point that I was an idiot that I should shut up that I had a skill issue, and that they had never had any trouble with the THAC0 System and had literally emerged from their mother’s womb rolling attack rolls for their thief-acrobat.

You gotta be able to shut your preferences off and look dispassionately at processes and how humans interact with them if you want to play the mechanical analysis game. Cost-wise and benefit-wise.

See, there’s actually a benefit to the THAC0 System. Generally speaking, as long as you don’t have any attack modifiers — which were rarer in the old days than they are now but which still did come often, especially once AD&D 2E hit the scene — as long as you didn’t have any attack modifiers, you knew as soon as the die hit the table whether you’d hit or not. That’s a pretty big payoff due to something called the Tension Curve and now my editor is tapping his watch so…

Let’s consider separating attack and damage rolls like this. What does that get you? Well, it gets you variable damage but it gets you variable damage in two, distinct ways. The first way is obvious: if you generate a random damage number using a plastic random number generator, you get a variable result. The second way only becomes obvious when you consider alternatives to having a separate damage roll.

Imagine you tried to calculate damage off the attack roll, for example. Say, each point by which you exceed the target’s Armor Class is a point of damage. Easy, right? But now all damage for every attack is locked into the same range and that range is always defined by the relative difference between the attacker’s Attack Modifier and the defender’s Armor Class. Having a separate damage roll means you can differentiate attack types by potential — and highly variable — damage.

Now, variable damage itself has its pros and cons. In early computer roleplaying games — and in some old wargames — attacks did fixed amounts of damage. That lets players count precisely how many hits they — and their targets — were from death which turned combat decisions into simple optimization calculations. In fact, it kind of negated the point of even having Hit Points and damage at all. Just count hits at that point.

This is actually why critical hits were invented. They added variability to otherwise fixed damage systems to throw some unpredictability into the mix.

Of course, variability has its downsides. All else being equal, variability favors the underdog. If you’re building a system and you want to balance it so the players are likely to win most fights provided they use some amount of basic strategy, every element of randomness you add messes with that. In modern roleplaying games, all else being equal, separating attack and damage rolls actually works against the players’ odds of success. So do critical hits and fumbles. Fun fact.

Having two separate rolls to resolve every attack also means that it takes two rolls to resolve every attack. That’s obviously inefficient, but it’s easy to overlook how significant that efficiency loss can be without drilling down into the entire process.

On the face of it, it doesn’t seem too bad, does it? First roll to hit, then check the result, then roll and calculate the damage if it’s a hit, then apply it. But now recognize all of the little invisible substeps in each of those steps. For example, before a player rolls an attack roll, they check — or confirm — their attack modifier on their character sheet. Then they check for — or confirm or recall — any situational modifiers from any feats, active spells, abilities, weapons, whatever. The Game Master might also rattle off a couple of modifiers due to the circumstances surrounding the attack. Once all that’s done, the player adds everything up and holds that number in their brain. Then they roll the die and then they add the number in their brain to the number on the die — however long that takes — and then they announce the result to the Game Master who checks or confirms the Armor Class on the creature’s stat block and so on and cetera and nauseum.

All that shit, by the way, is why rolling attack and damage dice in one toss doesn’t actually speed play up by much. The time sink here ain’t tossing the dice, it’s all the checking numbers, doing pre-math, doing post-math, checking the damage code, physically gathering the proper dice, checking the date to see if Carlo Acutis has been canonized yet so you can toss off a quick prayer because you really need this hit, and so on.

Then there’s the frigging psychology. Games are unfortunately made for humans and humans have brains and brains are stupid. I mean, look at that crap above about how subtraction shorts them out. See? Stupid.

You can’t analyze game elements without considering the human factor. I know many gamers disagree with me here because most gamers don’t think psychology applies to them. After all, they’ve never noticed it happening, but it is a big deal. After all, human brains are the things that decide if they’re having fun playing a game so they’re the target audience for every game ever.

Look at it this way: hitting by a lot feels different from hitting by a little even though, mechanically, it’s exactly the same thing. Especially when the stakes are high. Imagine the party’s one sneeze from a TPK but the last enemy on the field is also one solid hit from death. Now imagine a player chooses to attack and rolls… exactly the precise number on the d20 that they needed to hit. How is that player going to react? What about everyone else?

If you don’t concede that that’s extremely different from a player rolling a 15 to hit an AC 12 in the first round of the first random encounter of the day when everyone’s fresh, you’re just being a willful fuckwit.

Psychology always matters.

So imagine what happens when that player who rolled just the right number to hit the last foe on the field and save their party… rolls a single point of damage on the follow-up. How’s that going down do you think? What’s the psychology there, do you think? Because I think the psychology is about to cost your nice dining room table and maybe a window too.

When you separate attack and damage rolls, you’re letting the damage roll re-contextualize the attack roll. Now, no one rolls damage rolls when they miss, so the only time anyone tosses a damage die is when they’re already sitting on a victory. A good damage roll is gravy then. “Not only did you hit,” it says, “You did good damage. Yay for you.” A bad damage roll, on the other hand, says, “That hit? Screw you. The monster barely felt it. Why? Because I’m the frigging dice. Suck it loser.”

A good damage roll makes a success better; a bad damage roll renders your success meaningless. This is down to a thing called Loss Aversion. People feel losses more than they feel gains. Someone who hits and deals max damage might gain +10 Emotions Points, but someone who hits and rolls the worst damage suffers -20 Emotion Points. That’s what Loss Aversion means and everyone has it.

Let’s look again at why rolling attack and damage rolls together just to drive this psychology shit home. When you roll both the attack roll and the damage roll in the same toss, now you not only lose your hits to crappy damage results, but you also get to see how much damage you would have dealt if you hadn’t missed. You get to feel cheated out of your hits when you hit and get cheated out of your damage when you miss.

Good times, right?

Of course, looking at this shit analytically can also make it seem like a bigger deal than it is. In reality, Loss Aversion is only a big deal once in the rare while. Most players are going to roll average damage most of the time and most of the time the stakes won’t be that high. But that doesn’t mean the issue doesn’t exist. Just that the cost isn’t as high as I may have made it seem.

So what’s all this shit mean? Is it good to separate attack rolls and damage rolls? Is it bad? Unfortunately, it means neither one. I just laid out the major benefits and the major costs. It’s down to some game designer to figure out whether those benefits are worth paying those costs for and, if the game designer knows what they’re doing, a lot of that depends on the game they’re designing.

Yes, this does come back to what I said above about figuring out what you’re designing first and then making the mechanics to get you there. Are you surprised? You shouldn’t be. I pull this full circle shit all the time.

If we’re talking about Dungeons & Dragons here — or any similar fantasy adventure roleplaying game engine — I think splitting attacks and damage rolls is a good mechanical choice. It’s a very easy, very efficient way to add damage variation and to differentiate attack types by damage. The efficiency loss ain’t great, but I can’t think of an obvious way to get the same kind of variability that would be more efficient. I don’t think the extra randomness hurts anything, especially in modern incarnations of D&D, and I don’t think the loss aversion is a big enough issue to counteract the benefits. In fact, I actually think players occasionally having luck snatching away a victory can be a good thing for a game like D&D.

That said, if I were designing a game that was less action-focused and involved higher stakes and higher risks, or if the action resolution system didn’t lend itself to having damage rolls at all, I’d feel differently.

See? It’s all about what you’re designing.

Now, I could do a similar analysis of that whole cards instead of dice thing, but, honestly…

Like I told Desperandos above, nothing is interesting about naked game mechanics. Analyzing mechanics is a stimulating intellectual exercise once in a while, but I don’t think it’s the best way to design shit. Really, if you want to know what swapping one mechanic out for another will do, take it to the table and see how it plays out. That’s a million times more useful than two thousand words of over-intellectual analysis.

To make this long story ever so slightly less long — yes, thank you, editor, I see the word count; I’m wrapping up — to make this long story ever so slightly less long…

If you’re asking me what I think about swapping a bunch of D&D players’ plastic math rocks for a deck of numbered cards to shuffle and draw through, I’d say that probably sounds like a gimmicky waste of time. That’s my gut talking; forgive me for not showing my work here.

Honestly, I’d much rather have a conversation about the gameplay experience you want to create at the table and how you reached the conclusion that a deck of cards is the best way to introduce the randomness your game needs.

Now, that would be interesting.