Forward: Delvers of Dàrkmesa

The Delvers of Dàrkmesa campaign rules represent the culmination of my ten years exploring the theory and spirit of tabletop roleplaying games in the vein of Dungeons & Dragons.

This project began in 2008 in the wake of D&D 4th Edition. But that inflection point needs some context.

I had been playing D&D since 2003, when a new friend gave me a copy of the 3rd Edition D&D Player's Handbook, having recently upgraded to 3.5 himself. I read every word in that handbook one thousand times over. I was running games of 3.5 for my friends soon after, and quickly settled into a nice groove: each campaign I tried to run would last three to four sessions, only to spiral out of control due to my lack of experience.

In my heart, I usually blamed the players for how things would fall apart. I had played the epic campaigns through in my head, and tried to run them all to that end. Didn't they know that Tolkien-level excitement and wonder awaited them, if they could just go along with the story? They never seemed to march to the tune, despite their best efforts. Luckily, we were young, and all still in school in our home town. Starting a new game every month was never an issue. I had a million tired fantasy tropes that I yearned to emulate at the table, and a good handful of friends more than willing to indulge me.

I never admitted that I blamed them, and that was good, because it was my fault and I didn't know it. I didn't know that the stories I liked to read and write were not the ones I liked to play in or run. I couldn't see that my campaigns as conceived were little more than setting and stage direction. It was truly the bliss (or perhaps, frustration) of ignorance.

You see, the friend that introduced me to D&D was the first person I ever met who played. That made me the second person I ever knew who played. Each of my other friends who I invited to play then became the third, fourth, fifth, and so on. We were making it up as we went. Sure, we were using the rules. But the rules don't tell you how to play. Not really. And back then, it was faster to walk to the library and check-out a Dragonlance novel than to look up anything online. How other people played was a complete mystery to all of us.

Around 2007, the older cousin of a friend of mine came to town. His name was (is) Nick, and besides being a math genius, he was also an avid player of 2nd edition D&D. Once he found out his little cousin and co were all playing D&D on the weekends, he offered to run a (3.5 converted) game of Ravenloft for seven or eight of us.

That was good D&D. I played a monk, which was a terrible class in 3.5. But the D&D was good, so I had fun.

Then we went back to playing our own games (which I ran), and they sucked again. I would often pine away in my memory for the details of that game, searching for its secret sauce.

In 2008, 4th edition D&D is released, and for the first time, the veil of design is peeled back for me. The deliberate ways that the game was laid out and organized, the brutal crackdown on the numbers, the every inch regulated and standardized.

D&D it was not. My friends all lamented the something that was lost in translation. I maintained always that I loved 4th edition, but found it rather hard to express precisely why in the language of fun. In my eyes, 4th edition was designed to a very specific end: tactical combat play. In that, it succeeded. It was designed to be learned quickly. Again, it succeeded there. It was designed to look modern and aesthetically consistent. A+ again. It was the right product, but for the wrong market.

So my theory crafting and homebrewing began. "There must be a way," I thought, "to bring the principles of design from 4th edition into 3.5 such that we can play a superior hybrid game."

It should be said: 3.5 was universally regarded as broken by 2008. How broken exactly was up for debate, but the sheer lack of direction for the design of the game and its splat books was becoming a malignant issue. There was too much game, and no central idea to measure each component's worth and validity against. One of the primary draws to a new game from the player's perspective was the chance to try out new classes and options. But from the other side of the table, every new feature was a chance to break the game.

Notebook after notebook did I fill with my Frankenstein D&D hacks. If I recall correctly, the first things I ever scribbled house rules about were skills and feats. Did we really need over 36 skills? And who knows how many hundred feats? Do those options really enrich play, or do they just tickle the fancy of min/maxers?

By 2009, my family had DSL and I had my own laptop for school. So I decided to start Googling these questions and delving for answers in the dark halls of the interweb.

What I discovered was called the Old School Renaissance. It had begun around 2006 or 2007, as best I can tell, with some bloggers connecting on various forums and posts. They shared ideas from their history with the game, which stretched back to its very beginnings, and even before. They knew the old rulesets weren't perfect, and offered their solutions to the age-old problems. There was debate and proselytism among great grognard philosophers, and countless quiet members of the community, like me, just trying to absorb it all.

These champions of the old school tore my fragile conceptions about D&D apart, like the Twelve prophets of the Hebrew Bible chastising Israel. What was this game I had been playing all these years? Where had it come from, and what great pillars of its history had been lost to its many transmutations? Had we taken the wisdom of its forefathers for granted? Nowhere else have I seen such thoughtful effort by people to justify their ideas about a hobby and its function. While everyone else spent three years learning Pathfinder, I was mesmerized by the mysteries of OSR games.

If this era (c. 2006-2011) was indeed the Old School "Renaissance," then what happened in 2012 was the Old School "Industrial Revolution." In January, Wizards of the Coast announced that the short-lived 4th edition would soon be replaced by a new D&D, "D&D Next." In May, they began a massive open playtest of the rules they were working on.

The forums that existed on the old WotC website containing all of the playtest feedback and discussion were truly a treasure of knowledge and design theory. I was well into my obsession with writing a better ruleset by 2012, and I consider myself extremely lucky to have had the chance to explore those million musings before they were archived later that year.

For two years, there seemed to be no word of the next D&D, but my mind never stopped swirling around the problems the OSR had illuminated. As I got further into college, time for playing the game became scarce, but my interest in thinking about it only grew. The homebrews I made became smaller and simpler. There were fewer and fewer rules each time I would revise or start again. I was getting better at the design, but deep down I still didn't understand what I wanted. I was designing for better design's sake, but not for the sake of better play. Indeed, most of my homebrews never saw the light of day. The ones I did test largely felt like playing any other ruleset.

It doesn't matter much what font you write the stage directions in, turns out.

In 2014, D&D 5 came out, and my friends loved it. It was the greatest thing since sliced bread, fixing all the problems with 3.5, but still holding onto that something that 4 had abandoned.

But when I looked at it, I was disappointed. Sure, advantage/disadvantage is a great mechanic. And yeah, the monk is way better now. And oh hey, they are consciously keeping the official splat books and feats and skills and classes under control. All good things, no doubt, for those of us content with the quality of play we had achieved throughout the years. But D&D 5 didn't do what I desperately wanted it to do.

It didn't show me how to design toward something. D&D 5 is the everyman's game. It's a story game, it's a tactical combat game, it's epic high fantasy, it's gritty low fantasy. It's everything, and nothing. At the table, it feels just like 3.5 did when my friends and I couldn't yet afford the extra books.

Fast-forward another year, and Critical Role is D&D's technological revolution. D&D is sexy now, and it rivals the best TV shows for hours of content and number of dedicated fans. Like all instances of celebrity art, Critical Role is at first inspiring, and then soul crushing. Inspiring because it shows you what tabletop RPGs can deliver, and soul crushing because I am not friends with a cadre of professional voice actors who want nothing more than to practice their acting around my narrative, with some combat thrown in to spice things up. It isn't just an unattainable ideal, it may actually be an un-approachable ideal depending on the friends you have.

A cursory scan of D&D subreddits these days will show you the Critical Role heartbreak phenomenon writ large. DMs are afraid to let their players fail. They're afraid to change the plans they have for their narrative in response to what the players do. They hand their players stage directions, and instead of amazing voice actors and improv artists, it turns out their players are just regular people. The story-based D&D facade all comes crashing down when you try to fit it into the real context that the game has in our lives. A weekly hobby for a handful of non-writers and non-actors with shifting schedules and more important things to remember than the evil motivations of the Demon Vicar of Nowheresville in your tabletop soap opera guest-starring the player characters.

But even a few years into playing and DMing 5e, I still maintained that top-down DM fiat narrative was a viable play strategy. And it wasn't until a little later, when one of my close friends ran a short-lived game in which I played, that the final trigger for my OSR metamorphosis would occur.

The setup was this: the party (which had several new members) were asked to convene at an inn to receive a quest from an NPC. We did, and the NPC told us how to get to a mythic dungeon full of (literally) civilizations worth of treasure. If we do it, the NPC gets to keep whatever scrolls and spellbooks we find in the loot. My guy is a fighter, so that is fine with me.

The party leaves for the dungeon, which is in the center of a crazy necrotic swamp that has emerged around it because the dungeon is the fossilized remains of an ancient dragon god (cool) which is causing the land to transform as it decays (super cool). When we arrive, there is a crew of evil dudes camping around the dungeon entrance, which we battle and defeat in a fight (sweet). Then, upon entering the dungeon, we find out that all of the treasure has already been looted and thrown through a portal into the plane of Limbo (...what?).

Now, from my friend's perspective, he was offering the potential for a sweet adventure on a rarely-explored plane of existence. And had that been the hook for the adventure from the beginning, I would have been more than excited.

But from my perspective, my character's motivation to adventure had been flushed down the drain. That might not be so bad (hey, things don't always turn out the way your character wants), but my poor dwarf was also trapped by the swamp. I tried to make a choice about what adventure my character would pursue, and then when that was turned on its head, I was denied a second choice. Not only was the original intent of the adventure no longer an option, but if my character jumped through that portal, there was literally no way for me to control whether I could turn back. No way for me to make an actual choice.

I was outraged after the fact, and it surprised me! This was no different than a hundred other times that one of my friends or I had weaved a D&D web around our players to get them to do what we wanted. The difference this time was how obvious it was to me given the mindset that I was in. I didn't blame my friend, I blamed our collective understanding of the game.

The swamp was not an obstacle to the treasure, because the treasure wasn't there anymore. The swamp was an obstacle to beginning the adventure, or any other adventure for that matter, if the party had changed its mind. The dungeon was the single source of player-DM agreement ahead of time, and it was thrown out. The foundational structure of the game, that there was a treasure to be had if some obstacles were overcome, was abandoned for the sake of an unsure reward (the treasure was in Limbo, after all) after an indeterminate amount of time and obstacles.

The contract of the game was broken, I was mad, and that was the answer I had always been looking for. What was the contract that my games should make between the DM and the players? Whatever that contract, design the game toward it.

D&D 5 makes no contracts. All forms of play are equally valid, and therefore there can be no measure of what games succeed or fail except on the scale of fun. This makes DMing, and DM prep especially, a nightmare of guesswork no matter how many books or tips or tools you have. What type of game will be fulfilling? What type of game will last? How do I get the players to ask about my availability to play rather than trying to chase them?

Well what if, given the wealth of OSR knowledge that I had absorbed, I made the contract of my campaign specific in the way that is time tested:

In this campaign, the player characters will explore a megadungeon for treasure. They will frequently return to the outlying town or towns to restock and rest, and may occasionally pursue non-dungeon adventures when the desire strikes them. New players and characters can come in or be swapped in at any point while the party is in town. No individual player needs to attend every session because no individual character needs to be on any given delve.

I, the DM, promise to prepare for that. The players promise to bring characters that will eagerly pursue that cadence of play. It doesn't create any unspoken requirements to attend every session for the sake of keeping in the loop of the plot. It doesn't mislead players as to what they will be doing. And most importantly, it takes a hard line on the boundaries of the play area, but within that play area, there is real choice with real consequence. No Quantum Ogre, no DM fiat.

And so, if I designed a ruleset around that contract, and DMed the game around that contract, I would be able to actually improve the game from session to session because I and the players would be pursuing the same goal. Which finally brings me to Delvers of Dàrkmesa.

Once I landed on the idea of the megadungeon as a distillation of the time-tested campaign promise, I thought about making my own take on Dwimmermount (if only for the supremely euphonic nature of the name), but decided instead to go completely homebrew. That's how Dàrkmesa came to be. I was struck by the double meaning of mesa (a huge rock formation that could easily fit a megadungeon inside, as well as the Spanish/Latin root for table), and what would a dungeon be without darkness? The Delvers part is just to drive the point home. Both in the sense that the characters will be primarily engaged in exploring a dungeon, and in the sense that the game is about them and about their exploration, not some arbitrary story I have concocted.

The rules and classes and example spells to include all fell into place now that I had a gameplay goal in mind other than "have fun." Many of the strangest parts of old school games that I could never understand before started to make sense and became incredibly attractive aspects of the experience for me.

Of course the game should be deadly, because that is a large part of what creates drama in a campaign with no forced plot. Of course random encounters were necessary, to keep the players and the DM honest as they all pursued the campaign goal: random chance is what keeps a non-narrative game alive. Of course two hundred pig-faced orcs appearing in a dungeon cavern is a good encounter, because it forces the players to bring to the table what makes them good players, not well-statted characters. And finally, of course the dungeon is the primordial soup from which all good games emerge, because it is the most apt compromise between what the DM needs to do, and what the players need to do. It creates strict boundaries that allow the DM to prepare more of the right thing, but when done right it forces no specific action on the players other than to engage. They want the treasure, and however they get it is by definition a fine way to play. That's why we all play this game, to get the treasure that exists hidden in the forgotten halls of our collective imagination and good company. So why hide the ball? Why not be clear about how agency works in the game right from the beginning? Why not embrace the canvas that was originally conceived for the game, and was so important that it became the first word in its name?

Luckily, it only took me ten years of searching to figure that out. Now I have the rest of my time to play.

Disappointing Combat in D&D

SPOILER ALERT: I'm writing this while watching the newest episode of Critical Role (ep. 88). If you don't want to have minor combat details ruined for you, wait to read this until you've seen the episode.

Read at your own peril.
--

In this episode, the party engages in an underwater battle with a kraken. The setup is short but sweet, and then, as much D&D combat is want to do, the play slows to a crawl as soon as initiative is rolled.

Now, given that this is all submerged, I understand slower movement speed. But that's not what is going on here. Each player's turn takes one of two forms:

1) excruciatingly long because they are trying to figure out how the rules apply to the unique circumstances.

2) breathlessly short because they are grappled and fail to escape.

The short turns are effectively not turns at all, they are just more pauses in the action. Here is an example of an excruciating turn: Grog is swallowed by the kraken and he is blinded, restrained, and slowly burning in acid. When it comes to his turn, he doesn't have anything to do but "swing his weapon" to deal damage to the kraken from the inside. Of course, the whole reason he is restrained is because he is being squeezed by the kraken's insides. He doesn't have nearly enough space to swing an axe or hammer, but because the rules tell him he can do nothing else, he is forced to make nondescript attacks that don't reflect the narrative reality at all. Needless to say, the turn is slow and uninteresting.

Worse than that, it's how you are supposed to escape. If you deal enough damage to the kraken from the inside, you might get puked out. But as we've already established, that makes little to no sense. Like I've said before on this blog, if there is a rule for how to do something, people are much more likely to use the rule rather than make something up, even if the rule is lame. We think within the box most of the time.

I can't think of an opportunity for more dynamic and exciting play than when a player in swallowed whole by a huge creature. Unfortunately, that's not what happened on that turn or the next. It took an incredible leap of logic by Grog's player to pull out his magic jug that makes oil. Then, when Keyleth the druid is also swallowed immediately after, the big risk of setting off a fireball and purposefully igniting the oil in the jug pumped life back into the encounter. The kraken pukes them back up and the real battle begins.

About an hour later, the real battle has ground to a halt again (before the actual encounter ends) and it becomes a game of how to escape through a portal when everyone keeps getting grabbed and restrained by the kraken. The party's goal is to leave without killing the beast, but it's "stickiness" and the underwater environment make this goal extremely difficult. After several rounds of the party trying to break free and getting pulled back, Vax the rogue is swallowed while unconscious.

This is followed by another grueling turn for poor Keyleth who is barely able to keep the all the alter-self, animal shapes, and druid beast-shapes straight.

The issue here is not Critical Role, the players, or the DM. It's the rules. The rules are designed in such a way as to punish any "get-in-get-out" encounters. Every enemy is sticky in D&D, and the kraken is the king of sticky things. It has two average parties-worth of tentacles which auto-grab and restrain after dealing damage on a hit. The party members that are restrained lose approximately a third to a quarter of all their actions during the fight. They just fail their escape rolls and do nothing. That's not even counting the swallow ability, which is essentially a nigh-inescapable grapple.

It's painful to watch the party members on screen look utterly exhausted by their lack of options. The combat ends in an intense way, but that's all thanks to the roleplaying and DMing that are superb. They were succeeding in entertaining themselves (and us viewers) in spite of the rules, rather than with them.

I haven't thought about D&D from a design standpoint in a while, but these same issues are always on my mind when I do. There's got to be a better paradigm for handling combats like this, where everything devolves into repetition of two or three optimal actions until math saves the day for one side or the other.

Why the ranger class sucks these days...

To understand the fall of the modern ranger, it is helpful to study the rise of the modern bard. When I first started playing RPGs over a decade ago, the bard looked like this:

Now I first want to say that I mean no disrespect to the artist that worked hard on this piece. I think it is a great piece of art, but a bard this is not. The musical instrument is a hardly visible afterthought (he's wearing a lute on his back) and his outfit doesn't scream "storyteller" or "poet" but rather "rogue" or "eccentric."

And this perfectly describes the bard class (D&D 3E) that this picture was used to represent: a mixed bag of uncertain purpose. It has middling attack bonuses, a middling spellcasting ability, mediocre spell selection, and its class-specific features were so situational and odd as to be laughable.

Hence, this comical representation of the bard throughout 3rd edition's reign.

But now! NOW the bard looks like THIS:

Holy moly! Look at that badass fantasy character! A sweet guitar that sticks out like a sore thumb (in the best way) the subtle wood flute hanging around the neck, the lack of any weapons or aggressive garb. This bard is something unique unto itself. The colors are bright and stylish. Even the character's haircut is edgy and cool. This is a socialite. This is a world-traveler.

And just like the picture, the 5e bard is the king of classes. Highly versatile, interesting, and competent. A skilled magic-user, a healer, and combatant, and even a battlefield inspiration.

It is admittedly a bit overpowered, but at least it's playable in an unironic way.

So what has the ranger done in this same timeframe?

Well, the first ranger I ever saw in a D&D book looked like this:

This is someone I can picture roaming the wilderness for months on end. The armor might be a little overdone, but it also looks rugged and that is important. The weapons are sharp and deadly, there are torches on his back next to his bow, and he is wearing a green, elvish-camouflage cape. He's ready to face the wilds. And he even looks like he has, given his bronzed skin tone (for an elf).

This 3rd edition ranger was not as playable as it seemed, but it was playable nonetheless. Favored Enemy is a very tricky feature to make work in a standard D&D campaign, and that was the main draw of the class. Worst case scenario: your 3E ranger had the high attack bonuses, decent hit points, some fun combat perks to start with, and nice skill points and spells to fill out the weaker areas of the character. If your party needed to track anything, or survive in the wilds, you needed a ranger, no question.

What about the ranger now? Well........

She's either a drow, or a purple elf. Neither of those scream "wanderer of the forests" to me like the tanned elf from before. She is also wearing a bright blue cape and puffy white sleeves. Not exactly camouflage. 

Something tells me that this piece was just supposed to be a warrior-type drow, and when time came to put a ranger picture in the 5e Player's Handbook, somebody grabbed this one, drew a bird onto the bow, and called it a day. Not cool.

And worst of all, the 5e ranger is by-far the least popular class. Only three skills? Someone that survives on their own in the wild with no help only gets three skills from their class? Favored Enemy gets nerfed, even though it was already mediocre at best. And don't even get me started on the current state of the animal companion... what a joke. It's so unpopular, in fact, that WOTC has made it a point to update it completely. Are the updates enough? Eh. I'm on the fence.

The Upswing

So what is wrong with the ranger? Why does it blow nowadays? Why is it "weak"?

Because it doesn't do anything. Sure, it can do all the normal character stuff fine. In fact, it can do all the normal character stuff (moving, hiding, fighting, seeing, hearing, etc.) more reliably than any other class. Move through brambles unhindered, fight invisible things, not lose its way in the woods, etc.

But those things aren't exciting. If a rogue could only hide, and not sneak attack, it would suck (by modern standards). Well, the ranger is exactly that. Sure, he or she can run through a field of pricker bushes without getting slowed down or hurt, but once on the other side of the field, they will put you to sleep with their lack of unique proactive choices.

Doing things was the focus of my Ranger Class without Math from a few months back. If you don't wanna read that through, here is the lowdown.

I give them two major weapon choices in combat: bows and knives. They get to choose whether to engage from afar or close up in each fight, not as a one-time character feature. No more melee OR ranged rangers. Each ranger is both.

I give them an awesome and totally independent animal companion to control, which is just one more sweet and proactive choice they get to make each turn. Every ranger gets one.

I give them the ability to make natural traps, and search for natural traps. Boom. Interesting choices in non-combat time other than tracking.

I give them herbalism and bush medicine to make them healers because if rangers don't know that isht then they die all alone in the woods. Just makes sense, and is uber useful for the party.

See how easy that is? But just like the bard became more capable and proactive between 3rd and 5th editions, the ranger fell from grace. How ironic that the tracker and wayfinder class has lost its way.

But hey, now you know how to fix it.

5 Goals Your NPC Organizations Shouldn't Have

There are three types of NPC groups in fantasy RPG worlds:

1. Groups the PCs can fight. Example: the cult of an evil demon lord.
2. Groups the PCs can receive service from. Example: a merchant guild.
3. Groups the PCs can join. Example: giant-slayer mercenaries. 

No DM has trouble with the first group. They are the easiest to think-up and use in the game.

The second group is easier than the first, in that their motivations need not be world-shaking or nuanced. They can just be a trading band, after all. But they are also more difficult than the first group, because it is much less natural to roleplay a merchant than an evil cultist, considering how scant merchants are in most fantasy novels and D&D plot-lines. It's very similar to writing: the mundane stuff is the hardest to make believable because you haven't ever given much though to it.

Where many D&D worlds fail, however, is with the third group. I'll give you some examples.

Black-Flame Zealots: If these aren't the coolest divine group ever, I don't know what is. They are Assassin's Creed meets ninja meets Nepalese holy-warrior. If there was a novel or video game about these guys, I'd buy it in a heartbeat. Unfortunately, they are a clandestine order of monks who work primarily out of city hideouts and only battle with the enemies of their god. They are like a thieves guild, but with more smiting and less casual sex.

How in the world do you DM a PC that is a member of this group? How do you convince the PC that going off to a random dungeon and looking for treasure is a reasonable quest given his profession? If an old lady says she has ROUSs in her basement, why on earth would a member of the BFZs offer to help her? Don't they have a dark god to keep at bay?

How about the Cavaliers? Mounted soldiers? Really? That's the kind of character you are going to play in this dungeon-crawling game? One who needs a horse at all times to have any fun?

The deeper you dig into all these supplemental classes and groups, the more you realize that over half of them cannot be played.

Sure, you can take the Purple Dragon Knight prestige class from 3.5's Complete Warrior, but take a look at their description, and the paragraph-long disclaimer about player-character members...

D&D 3.5, Complete Warrior, p. 70, Wizards of the Coast

In the end, a PC Purple Dragon Knight has no rank, no authority, no experience, no service record, no commitment to future service, nothing. They have nothing but an honorary title and a slew of class abilities that seemingly appear out of nowhere, given how little the PC actually has contact with their affiliate group...

These disclaimers are included because the designers know that players want the cool features and titles, but none of the responsibility. The PCs still plan to explore and search for treasure and kill monsters as their primary lifestyle, so being a full-time (or even reserve) member of some army is not on their to-do list.

And why should it be? This is Dungeons & Dragons, not Soldiers & Sergeants. If you want to play a medieval warfare simulator, play one. D&D is not that.

This is a reality that DMs and PCs need to accept: limit yourself so you can focus your game and do it well. Players don't need access to infinite character concepts to have fun, they need access to a handful of character concepts that fit the world and the tone the game is going for. DMs don't need infinite NPC groups from every walk of life, they need a couple dozen groups, the majority of which should be adventuring groups that PCs could easily join while still maintaining their normal day job.

There are a million reasons why an NPC group would be invested in dungeon-crawling and monster-slaying. I'll put a bunch into a blog post sometime. But until then, here are five goals your joinable NPC groups shouldn't have, because they create direct contradictions with the adventuring lifestyle:

1. Operate in a particular area, e.g. a specific city, forest, or even country.

If the PCs ever want to leave, either the PC members have to quit, get some sort of negotiated leave time, or create a convoluted reason for why it is relevant to the Elvish Defenders of Leafwood that Joe the Ranger go off into a desert and dig up some magic scrolls about turning lead into gold.... These groups are type one or type two: enemies or service providers.

2. Research a particular thing unrelated to dungeons, e.g. feywild specialists, divination experts, very specific monster-hunters, etc.

This shouldn't be confused with something like a weapon-specialist group. I'm talking about any group where actively adventuring would cut into their study time, rather than be study time. Sword masters-in-training can hone their skills anywhere, but students of Fey lore have nothing to learn while infiltrating the underground city of the Derro. These kind of academic groups are also type two.

3. Serve a particular (non-deity) leader or country, e.g. a king or empire.

Unless he is the king of dungeon exploration, I don't want to hear about any PCs being in the kingsguard.... These are type one and/or two.

4. Pursue sedentary or mundane professions, e.g. blacksmiting, fishing, etc.

You don't have time to smith while you are slaying dragons and stealing their treasure, so smiths don't do that stuff. They stay home and smith. If you are a smith, you aren't a PC. If you are a PC, you aren't a smith. No ifs, ands, or buts. Type two again.

5. Enact a modern moral or ethical sensibility.

I'm not saying every group should be made up of assholes, but if your PCs come across the Loving Sisters of Peaceful Coexistence, do NOT let them join. Peace and non-violence are for farmers, not adventurers. Sad but true. The players can agree with their message in spirit, but they have to realize that their entire lifestyle (i.e. the entire game) is premised on violence and stealing, whether it is against other sentient races, monstrous races, or just plain ol' monsters. Evil in D&D is real, and it is out to get you. Monsters exist, and they don't only attack when threatened by oil spills or aggressive human expansion. And since the monsters are evil, might as well take their stuff when their dead. Right? Peace can be an ideal in your world, but it cannot be the mainstream. This could be type one or two.

Now go make some NPC groups that love spelunking and looting. I'd join that shit...

How to Escape the Railroad D&D Game: Get off at a station

Quick announcement: Haven't posted in a while (got a full-time job), but plan to keep posting as often as possible. Will most likely shift my focus from game system design to adventure design, as that is where my head is right now.
____________________________________________________________

"Some players, Mr. DM, just want to watch the world burn." 
 - Someone afraid of sandbox games

"Players don't panic when things go according to plan. Even if the plan is horrifying."
- Someone afraid of railroad games
____________________________________________________________

Oh that unanswerable question: railroad, or sandbox? So short, and so mystifying. Within its seemingly binary choice are paradoxes and contradictions enough to fill entire college courses. See: Free Will vs. Determinism.

But let's boil the idea down to this: every DM wants their players to feel as though they are making choices that impact the world, but how do you do that without needing to literally create an entire world spontaneously?

Some will argue that you make a little "sandbox" and let the players do whatever they want in that little area. Easy. My response is this: even generating a tiny area that is a "free-roam" zone for the players is an unimaginably huge task. The idea doesn't ever scale down. PCs can cover a lot of ground in one session. You can't pre-generate all those possible paths, even in a very limited sandbox. And if you're thinking, "you don't generate everything the PCs can do, just the things they are likely to do" I say, welcome to a railroad game....

Meanwhile, if your argument is that you create some basic details and improvise the rest, you've entered into a similar paradox. If you magically incorporate your details into any path the PCs take, then you are running something VERY close to a railroad game. Best described as the "Quantum Ogre" effect. On the other hand, if you are willing to entirely abandon what you have prepared for the sake of maintaining player agency, then why prep at all? What are the odds that your players will choose exactly what you have prepared for them without railroading them at all?

To further confuse things: For every player that will purposefully avoid a pre-built story for something they think will be more exciting, there is a player who will do the opposite: ignore any hook or moment of inspired DM roleplaying for the never-ending slog through what they see as the "story."

The only way to appease both of these player types is to give the entire decision process over to the players, and then hold them to their choices. Your game must have cause and effect relationships, or the "watch the world burn" types will have no boundaries. Just so, the game needs to have very regular moments of choice, or the "go with the plan no matter what" types will quash everyone's creativity.

Okay, enough theory. Let's talk about putting this into practice.

Step 1: Prepare shit. There are no two-ways around it. You have to prepare sessions if you are going to give them any substance. Dynamic combats with interesting monsters, unique NPCs with believable spots in your world, competent antagonists, and many other parts of great D&D only thrive in the pre-game creative process. Maybe your random tables, acting, and improv are so good you don't really need to prep. If that's the case, I hate you.

Step 2: Accept that session 1 is railroad-y. Every D&D party has to suspend disbelief for long enough to accept that they are together, that they are at least allies in some respect, and that they will continue to work together until further notice. This is not an easy narrative step when your players are supposed to be brimming with "free will" and "agency." As the DM, you have to put your foot down and tell the PCs, "Welcome to session 1. Your characters are are all gathered in ______ for _______ and are planning to do _________. Go." Get the game going, regardless of whether every player is 100% into the mini-plot you have for the first session.

Step 3: Everything in the world has space for a hook. You have to bake hooks right into the body of your game world. Every corner should conceal another hook. But you need to expand your definition of what a hook is. Magic items are hooks. (What is it? Who can tell us? Where are they?) Information is a hook. Overheard conversations. Letters. Gossip. Everything. (The baron's raising an army? Why? Where? Someone mentioned the God of Socks. Who is that? Where can we learn about such a deity?) People are hooks. The dark spells cultists use are hooks. Statues are hooks. Paintings are hooks. EVERYTHING CAN BE A HOOK. Write a sentence for each good one you think up, and remind your players that these hidden treasures of adventure are everywhere if they pay attention and ask questions.

Step 4: Pace your sessions correctly. This will end up being a whole other blog post, but the short version is this: be smart about how long to spend on things around the table, and when to initiate them. Big combat after 4 hours of roleplaying session? Probably a poor idea, especially if it's getting late. Remember: you can always stretch things out with good roleplaying and description. You cannot, however, speed things up arbitrarily. When players focus on something, the time management is under their control. Prepare enough content to fill 50-75% of your session, then stretch it as needed.

Step 5: Finish your sessions at a choice, and have the players make it before ending. This makes #4 doubly important, because step 5 depends of successfully executing step 4. Each session should contain minimal "choice" points at the beginning and middle, because the more your players can veer off the path mid-session, the harder it is to make that session tight and satisfying. BUT, a major choice about what to do next that is placed at the end of a session allow players to tell you exactly what to prepare for next time. This choice point doesn't have to be "multiple choice" either. It can be "open response." So long as you have a week to prep what the PCs choose to do, they can do anything they want. Now THAT is agency.

Hooks are not choice points unless the players have nothing more urgent to do. Hooks can be kept in the back pocket of the party until some adventure time is freed up. This maximizes player agency, because it allows the players to choose the next part of their story organically, based on what most interests them about your game world. Free yourself from the trap of the escalating railroad campaign! You are better than that!

But let's be real: at the end of the day, even this style looks like a railroad. Pre-planned sessions that the PCs can't really change much once they are prepped by the DM.

But railroads can take you anywhere. Railroads (in the sense of pre-prepared sessions and stories) only become constricting when the players get an urge to jump off mid-way between stations. Solution: put in more stations. One at the end of every session. Keep the sessions focused and let them flourish into more depth instead of more plot if you have extra time in the moment. Your players will naturally want to follow mini-stories and plots to their immediate ends. Keeping players focused on a plot for one session is not an impossible task. 

Trying to milk that one major player choice for three or four sessions of material is asking for trouble.

Sneaking Around In Tabletop RPGs

Stealth checks are a mess. I have yet to meet or hear of a GM that actually uses stealth checks in any deterministic way when it involves more than a single "sneaker." Party stealth is the most common kind of stealth check made, and as far as I have experienced, the party just needs to roll a rough average above the difficulty to avoid detection. Like a skill challenge a la 4E. How else could you actually do it?

From OOTS #90

If you determine stealth individually, then certain party members (read: cleric, fighter, etc.) will fail virtually every check, because the game assumes they are in super noisy scale or plate armor. This means that even if half of the party moved into position undetected, the enemy is no longer caught unaware, even if they are still surprised or sneak-attacked by the rogue and ranger, for instance. The game tries to go easy on the sneakier classes by still giving them damage bonuses for being stealthy, but the mechanics have made a completely reasonable tactic (stealth) entirely devoid of utility except for those who specialize in it.

This bears mentioning, because we have all snuck up on our parent, sibling, friend, etc. in real life, and scared the ba-jeezus out of them. Sneaking up on someone who is not paying attention to their surroundings is actually pretty easy. Even cats and dogs, with their superior animal senses, can be scared out of their furry wits by a sudden jump from around the corner.

So why does the game assume you will alert your enemies to your presence most of the time, if you let the fighter or cleric anywhere in their earshot? Surprising the enemy shouldn't be so much about silence as about timing.

Take the Siege of Osgiliath, for example. Regardless of how the build-up was shot (which I feel left a lot of tension to be desired), the great thing about this scene is that the Gondorian soldiers still manage to get the jump on the orcs by hiding just inside the entrance, despite half of them having plate armor. They get a surprise round, even though the orcs know full well that there are enemies inside the walls. In fact, a lot of the stuff about this scene that bugs people actually would make for great rules-of-thumb around the table.

Yes, plate mail rings like a bag of tin cans and wind-chimes when you fight in it.

But, there is always some kind of white noise going on that will lessen the odds that someone who isn't actively listening for sounds notices a person in plate mail walking carefully.

Yes, soldiers run across the orcs' line of sight like, a dozen times just before the boats dock.

But it is dark and misty and foggy and the GM has to be honest: the orc saw a shadowy figure dart across the entrance. They know that Osgiliath is occupied, but they haven't heard any war horns. Does that really give them enough information to avoid the ambush? I don't think so.

It is a strange vestige of incorporating perception and stealth checks into the game that many GMs and players operate under two very unrealistic and paradoxical assumptions:

1) That nearly every non-living, physical detail about the world is shrouded in some kind of impenetrable fog or obscurity until such time as a good perception check is rolled.

2) That nearly everyone is fully aware of each living creature within eyesight and earshot until such time as a good stealth check is rolled against them.

Those assumptions lead to this being part of every D&D session ever:

(GM) "You guys begin trekking through the woods"
(Players) "We all go stealthily" *rollrollrollroll*
(DM) "Uh...yeah, okay you go pretty quietly on your way"

How is this even a thing? You roll one stealth check each for a whole day of travel, when earlier that day the rogue had to roll a whole stealth check just to take two steps behind an orc without it noticing? That's like saying the fighter can get through a whole day of random encounters with one attack roll, but has to roll for every attack when he encounters a pre-planned fight.

Are your characters seriously checking behind every tree and around every hill and padding every footstep to avoid contact with sentient beings? Or are you just keeping your voices low and going slow? Cuz you shouldn't need to roll for going slow, bro. Slowing down and shutting up is easy.

Are your characters trying to move fast, but not alert anyone? Tough shit. You get one or the other, unless you have some bombdotcom spell you wanna blow for super quick silent movement, or you are a high level ranger with some class feature that gives you crazy stealth stride. It is a basic fact of life, and one that needs to be preserved to have an interesting and problem-solving oriented adventure game: the faster you go, the louder you are and the more likely someone will notice you. The slower you go, the easier it is to move about without drawing attention to yourself.

Stealth is probably one of the wonkiest mechanics simply because it exists as the veil between the free-range character action outside of combat, and the turn-based character action inside of an encounter. Some might think of initiative when I say that, but initiative is "Step One" of combat, not the connecty bit. If you fail your stealth check, typically, combat begins. Maybe the target starts running away, maybe they pull out an axe and start swinging, who knows. Either way, it's now a two-player contest, and turns start being taken. If you succeed on Stealth, you remain the only "turn-taker" in the game, which means it isn't even a contest yet. No combat. Yet.

What would happen if we relaxed the combat structure of hard turns in a specific order, and instead made use of stealth merely as a means of getting the element of surprise on your side? In movies, for instance, when the action hero gets the drop on the guard guy, the hero snaps his neck or muffles him and slips a blade betwixt his ribs and that's all she wrote. In D&D, anyone but the rogue has almost no chance of taking the unsuspecting guard out in one-shot, because once combat rules kick in, damage is supposed to happen in little chunks over a long period of time. I don't see why the fighter couldn't shove his longsword through a dude's back and take him out (perhaps not dead, but certainly down), regardless of his HP. But again, if your combat is very rigidly segregated from your "non-combat" play, then it will be hard to justify anyone but the rogue being capable of shanking a dude.

And so stealth also brings up trouble with regard to how hit points work. One of the primary benefits of sneaking up on an enemy combatant is the opportunity to quickly and quietly dispose of them. However, most enemies in D&D type games have more HP than a fighter can deal in damage in one round. This winds up being "role protectionism," for the rogue. Even if the monk, or ranger, or fighter, or paladin, etc. snuck up on the enemy, they would not be able to kill them quietly, or quickly, thus giving their little stealth operation a close to 0% chance of success (where success is taking the enemy dude out without alerting the other enemies).

But enough about the problems with stealth. How do we fix it?

Well, there are a number of things:

1. Stop screwing characters in heavy armor. Seriously. Give them penalties to swimming, give them a hard time when they try to run more than 50 yards, but stop forcing them to auto-fail stealth checks with your gratuitous penalties. Armor is heavy, armor is expensive, and those two things together are more than enough to counterbalance the bonus to a character's defense. In reality, D&D-type games have been biased toward not wearing much armor for a long time. If you are regularly in hand-to-hand combat, wearing no armor is suicidal. All those leather-pants-wearing thieves out there who have complained enough to be allowed into front-line fighting are messing things up for their armor-bearing allies. Reality checks should be used very sparingly when it comes to fantasy games, but this is a spot where one is sorely needed.
   
2. There are two types of "sneak attacks." Ones where you plan to eliminate your target, and ones where you plan to get in a solid first strike against a big evil guy like a dragon. Take that into consideration when you are stating up your NPCs and monsters. If the standard ol' city guards have 5d8+12 HP each, then only the rogue can ever hope to take them out with a surprise attack. Don't do that to your players. Let your player characters outrun the rest of humanity in terms of HP and bonuses and such as they level-up. That way, a normal city guardsman always has roughly 1-8 HP, and even the bard could kill him silently with a good stab in the back. If you were hoping for the city guard to put up a good fight against your magically equipped and time-tested PCs, you are designing your encounters wrong. Simple as that.
   
3. If a PC critically fails a stealth check, they should be seen, no question. However, a regular fail on a stealth check should be interpreted more loosely as "a complication is added to the scenario." The castle guards stop to chat, and the careful timing you did of their rounds is now totally useless. A new NPC shows up to speak to the guard, or give him some food, or something else. A third party, like a wild animal or a beggar or something, shows up and draws attention in a direction you didn't want it drawn. The dog smells you and starts to bark from his cage, but the guards are just telling it to shut-up for now. This way, when the party rolls stealth all together, there are three possible outcomes: everyone succeeds and no complications arise, some people fail and complications arise, everyone fails (or at least one critically fails) and the PCs are discovered before their plan can occur. Now, the fighter and cleric can come along just like everyone else, but each additional person is another chance for a crit-fail to show up. Bigger party = harder to go unnoticed = perfectly reasonable to me.
   
4. Be real about how aware the enemy really is. A small camp of orcs that is feasting and wrestling are essentially unaware of their surroundings unless they have posted a lookout (which should not always be the case, unless its wartime or something). Assuming there is no lookout, the PCs should be able to sneak around their camp (in the trees) without rolling anything. If they want to foot-pad into the camp, enter tents, steal food or gold, etc., then stealth checks are needed to see how well that goes, since a stray orc could easily befall them as they reach into the chests of treasure in the middle of the camp.
   
5. Allow the players to solve the problem without rolling. If a rock thrown in the opposite direction of the entrance can draw the lookout away from his position long enough for the party to slip inside, then the party shouldn't need to roll stealth at all so long as they figured that out and did it. An invisibility spell + slow walking should allow a character to move through an open doorway without a problem, even with a guard a few feet away. Maybe a dexterity check might be needed if someone was to almost bump into them, but you get the drift.

Those are my best pieces of advice. If you have any other tips, or you've got the ultra-genius way of resolving stealth that I could never have conceived in a million years, please leave them in the comments. Pretty please?

How to Handle Traps in D&D

I hate they way traps are presented in the D&D I've played. Because they advertise this:

Source

But that is never what it feels like. That elf rogue's player literally spent two seconds on each of his last two turns.

Turn 1: "I search for a control panel" *rolls a nat 20, clearly*... DM: "Wow. Okay. You find one in the wall behind one of the face plates."
Turn 2: "I disable the device" *rolls something we don't know yet based on the picture*

There is no drama here! The only thing that is going on here is dozens and dozens of dice rolls until enough high or low rolls are made, and then the DM either tells the players "You shut of the trap/escape" or "You die".

It's almost laughable how much this picture overcompensates for the lack of tangible drama the D&D trap mechanics actually cause. Look at all the bull$hit going on in that picture! I don't care what system you are running, that situation is unplayable.

So if that is how you do traps wrong, how do you do them right?

Traps are not a game of dice. Traps are a game of information. Adventurers survive when they have enough information, and then process that information correctly. For my money, information should never be rolled for, it should be role-played for. What information do the players get? Well, that depends on what kind of trap we're talking about.

Big Trap?
or...
Small Trap?

A big trap involves a major change in circumstances: falling into a deep pit, being run-down by a huge boulder, getting caught in a room that is filling with water, etc. Interacting with these traps and their aftermath could last as long as a combat encounter.

A small trap generally involves deterring a character from interacting with something: an explosive runes spellbook trap, a flame-jet trapped treasure chest, a poison covered doorknob, etc. These traps are short and sweet, and should be avoidable in mere minutes, if the party has the resources to do so. A poison doorknob can be avoided with a spare rag. A trapped spellbook may require traveling back to the city to get the assistance of a high-level magic-user.

Now, let's get something out of the way right here: there is no real reason why traps should be detectable. Just like there is no real reason why that level 1 dungeon shouldn't have twelve cave trolls in it, instead of one.

But this wonderful hobby of ours is about letting players risk their characters for highly rewarding (and uncertain) outcomes. If a situation clearly spells doom for the party, then it shouldn't be sprung on your players. Sure, they could always goad the ancient dragon, and then all get eaten as their just desserts. But at least they knew what they were getting into. Opening a door and being thrust into an encounter that is impossible to win, or a deadly trap that is impossible to avoid, isn't fun or engaging for the players, it's pointless.

So when it comes to traps, sure, you could make it undetectable except on a DC 48 search check, regardless of whether someone says "I scan the chest for signs of traps." But if you did, you'd be a dick.

Really no two ways around it.

You need to make sure that every trap you put into your game has at least a few warning signs that become apparent upon a players saying "I check the [insert whatever here] for traps/anything weird," or even just saying they do something that would get them close enough to notice.

Example: the flame-jet trap on a treasure chest.

This trap could be mechanical, or magical. If it was mechanically constructed, you could say that there is a VERY thin, almost invisible wire that runs from the latch at the front, across the top of the chest, and into the space at the back next to the hinges. Upon a player saying they inspect the chest, you tell them about the wire, and how it is under a lot of tension.

If it's magical, perhaps there is a rune inscribed on the side of the chest so that half of it is on the upper part of the chest, and half is on the lower part. It makes one symbol when the chest is closed, but opening it will split the symbol in two.

And that's it.

You don't say "This looks like a flame-jet trap." You don't say "Something is probably going to shoot out of the chest when it's opened." You just describe the warning signs. Then, the party needs to figure out what they are going to do. Cut the wire? Open the chest with the fighter's greatsword? Cast antimagic field and then open it normally? Pick the chest up and throw it across the room? Bring the closed chest back to town and figure it out later?

And the best part is that all of this can be role-played. There is no roll that will make or break your party's strategy or chances of survival.

Another example: a room that locks itself and fills with water upon a character turning a knob attached to a false door in the wall.

What are the warning signs here? Well, the water has to come in somehow. Sure, they could be totally hidden before the knob is turned, but what fun is that? Have some metal slits at the top of the walls that circle the whole room. Have them be slightly wet upon inspection. And the water has to drain somehow after the trap is sprung, otherwise this is a one-use room. The drains would need to seal off, but you could have some puddles on the floor. Where did those come from, I wonder?

And that false door should obviously be false upon inspection, but that's it. They get nothing else from looking at it. If the rogue says "I check the door for traps," I reply "The door appears to be fake. You suspect there is just more wall behind it," done. There is no way that by simply inspecting the knob a person could deduce the water trap. But once they know the door is fake, but the knob seems real, and if they think about the puddles and the damp wall slits, then a picture starts to come together.

Once water starts filling the room and the doors magically seal, the players focus should now be on how the trap shuts off. Plugging the slits in the walls is probably not going to work, unless the wizard has some powerful spell they are willing to spend.

Maybe there is a small sapphire on the ceiling that hangs down a few inches, and when the water touches the sapphire, the traps shuts off and drains opens up in the floor? Now, a player frantically looking over the room as it floods can put two-and-two together and try to splash water onto the gem to preemptively shut the water off.

And if they don't figure that out, someone needs to think of a way to trap some air inside the room to breath for a bit after it fills up. Again, the water has to drain at some point, hence only puddles being left when they walked in.

Be sure to preset the rate at which the water fills the room, so that there really is an objective time-limit to the PC's attempts. That keeps you, the DM, honest, and allows you to answer accurately when the party asks how deep the water is each round.

So to quickly summarize:

1. WARNING SIGNS WARNING SIGNS WARNING SIGNS!! One at least, more for more dangerous and intensive traps.

2. Give out the warning signs freely. If they waltz right by, then they miss it and get the short end of the stick. But if they tread carefully, the information should fall in their laps.

3. Role-play it. Anyone can try to plug a hole, or smash a chest, or cut a wire so long as they have the tools to do so.

What "Say Yes" Really Means, Part 2

If you haven't caught up with part 1, please do.

So, we were talking about Saying Yes, and then after hundreds of words, we got to an example where I said a human fighter and dwarven cleric ought to be able to jump a chasm without a check, even though the elven wizard spent a fly spell to cross it while carrying the halfling rogue.

The catch was that the DM told the burlier, armored player characters they could jump the chasm if they ditched their backpacks and weapons, where normally, a DM playing D&D 5 or the equivalent might just ask for a skill/ability check.

And the question I left you with was, "Why?"

Why not use that rule that the game has put there specifically for this case instead of making up some gamble you want the players to take instead?

The answer is complicated, but I will try and express it as best I can.

Essentially, your players should NOT be lame. They should be the stars of the story, and the stars of the story don't just miss a jump and die for no reason.

When the fellowship was fleeing the Mines of Moria, they weren't constantly falling off the very thin and precariously angled stairs, even though they totally should have been in real life. Why?

The reason is two-fold.

Firstly, the Lord of the Rings is a story, and stories are different from real life in a few very important ways, not least of which is that if slips, falls, and deaths are not built up to and satisfying on a narrative level, then they are cheap and lame and shouldn't happen.

Aragorn falls off of a cliff while tied to a warg because it builds up narratively to a sweet comeback moment at Helms Deep. And the plot reason for his fall isn't because he misses his footing. He's dragged off the edge by a leather buckle caught on his bracer. Nothing he could do about it.

Secondly, adventurers are awesome. Not immortal, or invincible, or the most powerful beings in the world, but they are awesome. Gandalf is awesome, that's why the only plot reason that would work for his falling into Khazad-dum is that the Balrog is pulling him with his fiery whip. And the narrative reason for Gandalf falling into Khazad-dum is so he can have an epic fight with a demon while hurdling toward the center of the earth and then die and come back even stronger.

They did NOT fall anywhere because they failed a regular ol' jump check at some random moment during the game when the d20 decided to roll a 1.

And guess what!?

That hypothetical D&D situation with the chasm and the extra weight in nearly identical to Gandalf on the Bridge of Khazad-dum. He had a grip on the ledge, and the rest of the fellowship was there. They might have been able to pull him up to safety, but if they had tried to, they may very well have been turned into pin-cushions by goblin arrows. That would have been exactly what the DM told them at the table.

DM: You can leave him, or you can try to save him, but if you do, each of you is very likely to be targeted by at least two goblin archers a piece.
Frodo's player: "We have to save him! I go grab him and help pull him up!"
Aragorn's player: "I stop him from moving forward. We'll never survive that many shots."
Gandalf's player: "Fly! You fools!"
Legolas's player: "Yup, that decides it. We're outta here."
Frodo's player: "I resist! Opposed strength checks."
DM: Nope.* Aragorn is much stronger than you and he's already got a hold. You're going with him whether you like it or not.
Frodo's player: "NOOOOOOOOOOOO! GANDAAAAAAAAALLLLLLF!"
DM: Woh, dude, chill out.

*See there? That's another place where the dice would have gotten in the way. If the DM allows a halfling to try and escape from his strong human companion who's already grabbed him to pull him to safety, then all you are gonna have is inter-party conflict. And then one character is biting another character's nose off, and it is just a mess.

I guess if you are looking for a hard and fast rule concerning say yes, to make things more concrete, here it is:

Say "Yes" or "Yes, but..." as much as possible outside of battles. That is the part of the game where it is easiest to create excitement and drama with simple give and take/cause and effect. "I want to do this." "Okay, it will cost you this." "What if I did this." "Okay, it will only cost you this, but this might happen too."

When you get to a battle, there is an assumption that neither the DM nor the players are not in total narrative control. You can't adjudicate combat with a give and take method like you can exploration, because the goal of combat is to leave with as close to what you came in with as possible. Yes, there may be treasure at the end of combat, but the combat is not necessitated by the search for the treasure (unless it is dragon skin or nightmare ashes or something like that literally ATTACHED the thing you have to fight). At the end of a combat, someone is going to lose a lot, and the great thing about having rolls decide who loses and who wins is that it keeps the story interesting, for both the DM and the players.*

*Quick note: this requires that your combats be meaningfully setup in the story. If you just throw some monsters in for no reason but to be in the way, then whatever fair and randomized result the dice give you, no matter how heart-breaking or mind-blowing, is not going to be narratively interesting or earned.

So I say roll all the dice you like for combat. In that instance, they are like salt: mostly about taste, though way too many will definitely kill you. Outside of combat, try not to roll any unless you, as a DM, can honestly say, "I don't know how I would narrate this fairly or appropriately without some randomized result." Then sure, roll a die or two. But no more.

God I hate this...

What I think this implies for game design is this: skills/specialties/whatever you call them are actually worthless mechanics from a roleplaying perspective. If you give certain characters "skill" in something, it sends an immediate message to the players who don't play that kind of character that they cannot succeed at that particular thing unless they have that mechanical bonus.

If you think your character should be aided by their level of experience in a particular area, then make it known that your character is experienced in that area. Your DM should just allow you to do more things related to that experience without any issue. You were a renowned blacksmith back in your home city? Then you can smith a sword, no check necessary.

Ahh, so you want to smith a special kind of blade that you've never made before? Roll me intelligence! Let's see if you become even more renowned over night...

What "Say Yes" Really Means, Part 1

I am an RPG player and rules tinkerer who came into the hobby in the 'new school' age. D&D 3/3.5 was my first tabletop, pen and paper RPG playground. From there I went to D&D 4, Pathfinder, every other OGL system under the sun, and now I am playing mainly D&D 5.

In the design diaspora that was D&D 4, its aftermath, and the announcement of D&D Next, I began wandering the internet in search of RPG wisdom. Little did I know just how vast the halls of the RPG internet realm were, and how much the players of 'old school' games had to say.

Let me tell you, the blog posts were titillating, and the house rules, desirable.

But coming from a new school background, and having no old school players nearby to join in with, I had to take all this information and decode it like some ancient codex. What was it that these old-schoolers were so crazy about, and how did their games appear to have such a laid back and easy-going feel to them, while simultaneously sounding deadly and challenging?

Well, you will have to ask them, not me. I'm sure I couldn't explain it half as well.

But! There was one old-school rule that has clanged around in my head for ages, which I have only recently come to understand. Sort of like how Buddhism has seven stages of enlightenment. Or maybe not like that at all. I don't know.

That rule is this: Say Yes.

So simple! And yet so complex!

At first I thought that Say Yes meant you should never shoot down a player's idea. If the player says they want to attempt something, don't tell them "No, you can't do that" or "No, your character wouldn't do that."

Weeeeeell, that's not really it. It is important to have an open mind, as a DM, and remember that the game is at least partially about entertainment. That doesn't have to mean silliness, but it does have to mean engaging your player characters in events and challenges that they find interesting. So, if you prepared a dungeon crawl based around helping the bartender get rats out of her basement, but your players want to accompany some randomly generated NPC, who they met in the bar, on his trip to some randomly generated town because it sounds like fun, weeeeeeeell.... you better start thinking up possible road encounters.

But the Say Yes rule goes even deeper than that. Say Yes is a rock-bottom credo about where the drama and action are in your game. The more you Say Yes, the more you focus the challenge and drama onto the most important parts of the game, and onto player skill.

Here is a little hypothetical D&D type situation: the party (a human fighter, halfling rogue, dwarven cleric, and elven wizard) is fleeing the underground goblin city, and they are being chased by a half-a'gabillion goblins. They come to a chasm that is juuuuust slightly wider than is comfortable. The wizard has cast fly on himself, picked up the rogue, and safely made it to the other side. The fighter and the cleric remain.

In a world without Say Yes, they ask the DM "Can we jump it?" and the DM says: "roll jump checks." The fighter probably has a 50% chance to clear the chasm. The cleric will almost certainly fall to his doom like a stubby-legged anchor.

What if they asked the DM about their odds of clearing it? He/she might say "eh, not great." But then what? Grappling hooks? Magic? Pitons and rope for godssake? If the DM wants to keep the adrenaline pumping, there is no time for that, in-game or out of game.

And you can't toss a dwarf, obviously.

So what is the solution? Not put any chasms in the game? Oops, we have a heavily armored dwarven cleric, I guess we can't ask the party to jump anywhere this campaign.

In a world with Say Yes, the DM assumes that the characters can accomplish the task before them, though a sacrifice may be necessary. Does the jump have a DC? Some target number that they have to roll a check against? I don't know, maybe.

But should it be relevant? No!

When the players for the fighter and cleric say "Can we jump it?" the DM should say something like "YES, but you think you'll need to ditch the backpacks and weapons to clear the distance."

See what this does? It gives the players a whole scope of options to fiddle around with. It says "Here is the situation, and here are all the variables. Figure out a solution, if you can, and preferably one that doesn't get everyone killed."

Now, those two players are practically running around the table figuring out a plan, "Quick! Toss your backpack over the chasm. You guys on the other side will catch them, right? Can we toss our weapons? What if we tie rope to our belts and then to the ends of our weapons? Could we jump the gap and then pull our weapons over from the other side? Let's try it."

Now you've got compounding variables too, because you have allowed your players to do something other than succeed perfectly or fail perfectly. The DM makes a roll, and the goblins show up just before the party can pull their warriors' weapons over. The goblins grab the weapons and begin pulling the opposite way! Now what? Does the party cut their losses, literally, and lose their weapons? Do they get into a tug-o-war with a whole cadre of goblins? Will one of them be pulled back to the other side?

"But wait!" Some of you may say, "How can they jump the chasm without a jump/athletics/acrobatics check?" Isn't that what those skills are for?

Well, I'll have an answer for you in part 2.

I'm for Random Level Up Bonuses

Power-gaming requires the ability to plan your character.

Playing stereotype characters requires the ability to plan your character.

Games that emphasize balance require the ability to plan your character.

So, because I like my games kinda old school, I don't care much for the ability to plan your characters.

Stats should be rolled randomly, straight down. I also allow the player to swap any two stat values for each other before getting on with the rest of character creations. This allows a meaningful but minimal amount of control over the character's trajectory.

Then I randomize the number of items characters get. Usually basic clothes + 1d6 other things. This isn't a rule for how much the character has to their name, but rather how much they have on them at the beginning of the game. Rolled low? Well, maybe your character was just robbed by bandits, and strolled into town without two coppers to rub together.

Race and class are the player's choice, but since stats are mostly random, this is another instance of meaningful but minimal control. (Combine that with VERY simple races and classes that only grant abilities at first level, and you have a nice open ended character on your hands.)

Lastly, when characters level up, I have them roll on one of two tables. Even levels is a 1d6 table for increases in HP, extra spells, etc. Odd levels is a 1d6 table for increasing one of your stats by 1.


I want to give an example of how you can use this to enhance your games, in case you aren't already sold on the idea.

EXAMPLE: Grimbug, the orc warrior, had just achieved level 2. His player rolled nice strength at character creation (hence the warrior class), and he has a nice dexterity. In my game, that means he is pretty good at sneaking around. But he is a warrior, so he doesn't think about sneaking all that much.

That is, until Grimbug's player rolls on the d6 level up table and Grimbug gets a +1 to stealth (which translates into a 5% increase in stealth check successes in my game). Well, suddenly Grimbug's player is gonna think about stealth more, and probably try it out more often.


See, random level-ups denote what the character will be better at in the future. They speak to a part of your story that has yet to be told.

Level-up bonuses that are set and selectable are indicative of what has already happened...which has already happened...and can no longer help make your game more fun.


A friend of mine was recently bemoaning his players' cliche character and race combos. Dwarven fighters, elvish rangers, halfling rogues, etc. He's running a 5e campaign, but this player trap has been around for a loooooooong time. I'm not saying my style of game is a 100% fix...

...all I'm saying is that when Tordek the Dwarven fighter starts developing magical abilities he doesn't understand (and his player didn't foresee or plan) it certainly makes him a lot more interesting.

Why I Don't Let Players Roll All the Dice

"In my game, the players roll all the dice."

It's a nice thought, and it certainly has its benefits.

-It makes the GM seem more like a storyteller and less like an opposing team.

-It speeds up the GM's turn, keeping players engaged.

-It removes the need for a GM screen, which makes the whole experience a bit more casual and comfortable (imho).

But I disagree with it on theoretical grounds.

You see, game mechanics shape the story you tell at the table. Even if you don't think they do, they do. Let's take 3.5 as an example:

When a fighter swings a sword, the fighter's player rolls the dice. But when the same fighter is trying to avoid a sword-swinging orc, he doesn't roll anything. He uses his static Armor Class number and hopes the orc rolls low. Why? Isn't the fighter still trying to win? Isn't the player served equally by attacking and defending? Why not give the player agency (or rather, illusory agency in the form of a random die roll) over protecting their character as well?

Here's why: if your game allows players to impale themselves with their own roll, then your game probably isn't gonna be that fun.

Truth be told, there is a chance this happens in my game. If a player fumbles an attack roll, there is a random fumble table result that allows the monster an auto-hit counter attack that could kill the player, I suppose. But that result requires very low HP, and then a fumble, and then rolling one particular result on the fumble table, and then rolling enough damage on 1d6 to kill the player. Rare enough to warrant the shock value? I think so.

Anything more common than that and I think you have a problem. I can see low HP moments feeling like those god awful video game quick-time events which force you to press a certain button within one second or die. "Uh oh, the third orc is attacking you! Better roll AC again, or DIE."

As a player, I can accept when a monster rolls well and KOs me. But if it was my roll that took me out of the fight, I'd feel stupid.

Another example from 3.5: saves vs. magic.

3.5 had a near universal mechanic of "offensive actor rolls 1d20 + modifiers." Why would the game suddenly reverse this rule for saves against magic? Was it a mistake? 4e seemed to suggest so....

I don't think it was. Where magic is concerned, expending a limited resource (spell slots) if price enough for casting a spell. Simultaneously asking the player to risk a total miss, which would then waste the resource completely, is just cruel.

Now, asking the defending monsters to roll instead doesn't guarantee a hit, in fact it hardly changes the probabilities at all. But it does transfer the illusory agency of a miss from the player to the monster.

Wizard rolls low when casting a spell: "Wow Magic-Bob, you suck. Learn how to be better."
Goblin rolls high when saving vs. a spell: "Wow Magic-Bob, that goblin sure is tricky."

See what happens? It was the same exact mechanical outcome, but a different narrative structure, because when the magic-user's spells all go off with roughly the same DC to avoid, no one can say the caster has suddenly become crappy at his one job, instead, the monsters just feel more capable.

And having monsters that feel more capable is ALWAYS better than having players that feel less capable.

He had a Gaping Vortex in his Chest

Enter O'Malley. An undead dwarven justicar with an eldritch vortex of infinite pain taking up most of his torso.

He was an NPC that I'll never forget. And he is the result of an encounter that was too high level for the PCs.

This reaches back to the infamous Nick's Campaign (3.5 D&D) that I mentioned in my introduction page. I was a player, and I was a monk, and I hated it. But Nick was such a fantastic DM, and my friends were so swept up in the awesomeness of the sessions, that I couldn't help but get swept up too. Ah, nostalgia.

See, Nick had done something that many, many DMs refuse to do: he posed us, a group of rather inexperienced players, with an encounter that was way, WAY out of our league. He let us prepare for it. He let us think that, with the (then living and quite friendly) Justicar and some townsfolk with pitchforks, we could handle it (what the actual encounter was I no longer remember).

We could not handle it. Not at all. We had to flee. And the worst part, we had to leave the Justicar behind.

To die.

To be raised from death.

And to be controlled by a terrible, horrible, evil wizard from Ravenloft named Ian Mcalister.

O'Malley returned a few sessions later, looking bad, but not so bad that we saw his as a threat. He walked right up to our wizard in a tavern an double crit on a surprise attack.

Bye-bye wizard head.

Ian Mcalister had sent O'Malley after us, to hunt us down and kill us with the very man who embodied our failures as a party. O'Malley was our past sins pulling the strings of our future fate.

We fought against O'Malley, his wizard overlord, and countless other abominations in a final-stand fort-defense battle. Our wizard (resurrected, though painfully), used some dark magic tome to cast Finger of Death on Ian Mcalister and save us all.

But O'Malley got away. No longer was he a puppet, but he still harbored a hate for us that could not be satisfied.

And we had sinned again. Our wizard had given in to the dark side to defeat evil, and this set in motion an entirely new course of events.

We were not heroes. We were survivors. We had abandoned people when it put us in too much danger. We had used arcane horrors to fight back the darkness. We had delayed and allowed the forces of evil to muster this far.

But such is war. Lord of the Rings may feature shining heroes without a bit of red in their ledgers, but we've all heard Lord of the Rings. It's over now. Good vs. Evil is no longer an interesting story when the lines are drawn so clearly.

Fantasy books and Fantasy RPGs have gone beyond fairytales and bedtime stories and become literature. They have become about people and their faults.

So give your players something to fail at. Give them situations with no easy answer. Give them hard choices. It shouldn't be a cheap shot, or contrived. It should be organic, and the guilt and regret should come from the players recognizing that their character took on a responsibility, and failed to live up to it.

We tried to save O'Malley's town. But we couldn't. We wanted to. We really did. But we misjudged our enemy. And Nick, the DM, let us do that. He let us walk into a death trap.

Sure, in the long run, fleeing meant losing a battle to win the war. It was for the greater good.

But the greater good should always have consequences in the now. That's why it is the greater good. Because it takes honest roleplaying and hard decisions to achieve it. If it were easy every time, or if the encounter was always winnable, then it wouldn't be a story about the greater good.

It would just be a story about regular ol' good. We have real life for that.

Feats of an Unorthodox Nature

Although my knowledge is limited, I can pretty safely say that no tabletop rpgs have rules for doing the mundane things that the players themselves do on a regular basis in real life. Walking on solid ground, casually speaking with someone in a single language, buying food, eating food, sleeping, navigating a small town, etc..

The vast majority of rpgs, on the other hand, devote most of their rules to those things that are considered extraordinary, but within the bounds of the setting. (For a wonderful discussion of the quantity of rules in a game, and what it means and doesn't mean, check out this post over at How To Start a Revolution....) This would include (to taste): martial combat like swinging swords and shooting arrows, magic in nearly any form, character health and performance, character death and destruction, nuanced or uncommon skills like riding a horse into battle or singing a beautiful song, avoiding bad things be they fireballs or spiked pits, navigating magical or intensely wild landscapes, etc..

Then there are my favorite things, the Feats of an Unorthodox Nature. When designing a game, you have to decide whether this category of actions will be a very long list, or a rarely referenced, highly obscure group of things that no one ever does. A perfect example, last night I was flipping through my friend Max's 5e dmg, and I came across an optional rule at the back: climbing on larger combatants. In our 3.5 games, this was unheard of. No one ever even considered it a viable option. In 4e, I seem to recall one Rogue class power that could be taken that allowed for climbing combatants. No one was going to play a rogue simply to have that power at level 17 or some-such, however.

Now here we are in 5e, with an optional rule that would allow anyone to attempt this radical maneuver.

Personally, I don't care one way or the other whether you allow climbing opponents in your game. But the way you rule it in or out affects how the players will use it.

Think about it. You sit the players down for the first 5e session and declare the "climbing combatants" rule is fair game...and suddenly it's climbing season. You won't be able to get your players off the big boss monsters. And why not? We're all here to have fun! Why wouldn't we use the rules that sound the most fun?

But you don't want your players climbing your monsters constantly! You want them climbing monsters when it's cool, and fits with the flow of the encounter, and when they aren't just trying it for the attack or defense bonus it grants.

Here's how I fix it. Don't write a bunch of rules for every cool but crazy thing a player could do, just write one good rule for this category of unorthodox actions. My rule is:

• Want to do something crazy awesome? (Or maybe just crazy?) Roll the check with Disadvantage. Success works as written, but failure has two states: If a single die is too low (i.e. you failed due to disadvantage) then you fail. However, if both dice results are too low, then you fumble. 

I was inspired by the 5e rules, and by this Called Shot Mechanic by Zak S. Later, I actually found the same exact rule over at The Last Gasp, but without the advantage/disadvantage terminology.

This fixes the problem of non-use and abuse of crazy maneuvers by giving the players a "blank check." You aren't pushing the players (unconsciously) toward a particular choice of action, but you are also opening the door for narratively-appropriate inspiration.

If that troll is begging to be climbed on, then the players will try it. But if there's something even cooler burning a hole in your player's imagination, they'll try that instead. That's the difference between a rule that provides players agency, and a rule that provides players with a script.