Sunday, March 31, 2024

Trapfinding and "rules elide" as a design goal in 3e D&D

 This started as a question on the queer OSR server (discord invite) about if a rogue's trapfinding ability in 3e is an example of Jared Sinclair's "rules elide" in effect.

I think rules elide matters a TON for 3e D&D. In fact, it (although not under the same name) was one of the primary design goals.

A lot of people think of 3e as the first "modern" D&D. It's the first one produced by WotC instead of TSR, it's the one that really made a break with old-school play design and goals, and it's the one that really started on stuff like character builds and balanced set piece combat that characterize "modern" Dungeons and Dragons.

But in reality, like every other edition of every RPG, 3e is a response to what was going on in the game immediately before it, and what the problems and issues were with 2e as designers, players, and the people who made 3e saw them to be. When we really think about 3e as a "modern" D&D, most people are thinking about 3.5, especially later 3.5 when a definitively "modern" approach to supplement and game design had been developed (an emphasis on character balance, set piece combats, and character creation with "builds" of interlocking parts.)

But 3.0 is a different beast in some ways.

And one of the fundamental problems with 2e as the game designers of 3e saw it was just how inconsistent 2e was as a ruleset. (My source for this is the 3e preview articles in Dragon Magazine, 1999-2000.) 2e was an iterative improvement upon 1e AD&D before it, and 1e AD&D was an iterative improvement on OD&D (largely by Gygax as an attempt to formalize and integrate the rules into something suitable for fair tournament play.)

In neither 2e or 1e had there been an attempt to completely strip down, refashion, and rebuild D&D from the ground up as a refresh of the core game design. Indeed, if you compare the two, you can see that they substantially have the same math and game concepts, but 2e adds more options, clarifies various parts of the rules, and rewrites parts to make them for "useful." (My quotation marks.) For example, the 1e DMG has huge portions devoted to random monster tables for various parts of campaign/adventure design, while the 2e DMG has far fewer specific tables and instead devotes that same space to teaching a DM how to create them with specific mathematic models (a 2-20 range you get from rolling a d8+d12) and filling them with the much larger, much more diverse array of monsters 2e had. However, those same monsters, with minor conversions you can do in your head on the fly, are usable in either version. (This also extends to player content: the Complete Wizard's Handbook for 2e specifically says it's usable with 1e games as well.)

Why did neither 1e nor 2e do this complete refresh? It's pretty simple: the rules worked well enough. There was a strong enough play culture and a strong identity of what D&D "was" that even though the rules were wobbly sometimes (math not quite working, the vagaries of divining what Gary Gygax meant in the 1e core ruleboooks), the overall core play experience was good enough for the vast majority of players. (2e, if you've never read it, does an absolutely terrible job of actually teaching you how to play or run D&D, but most of the rules are there, it just doesn't tell you how to put them together into anything. It assumes you already know, from someone else teaching you or experience with Basic/a starter set game.)

And the designers of 3e didn't disagree with this. Instead, they had a problem with the second-order effect of it (not how this core design of AD&D previously worked, but what it created in things following from it). And that second-order problem is pretty easy to see: if you look at 2e supplements, they are, universally, a MESS. Not each one individually, but taken together, they are completely indigestible. And one of the biggest problems is that they had to keep reinventing wheels. 2e had a much, much larger publication history than 1e did, and each "core line" rulebook was also interacting with the flurry of individual settings, each with its own variant rules. It was impossible to keep everything straight, so 2e supplements are full of notes like "this supplement uses the psionicist class as developed in The Complete Psionics Handbook" because they had to specify all that stuff. But wait, that version of the psionicist was made obsolete by the later version from Player's Option: Skills and Powers. What if you wanted to use Skills and Powers and the Dark Sun stuff that depended upon the Complete Psionics Handbook? Good luck.

It wasn't just the issue of overlapping, substituting publications though. Each of these takes on say, psionics, or seafaring and weather, or adding additional armor types and functions, would be going back to the 2e core rules and then building from there. And because the 2e core rules were a bit wobbly, the resulting rules in these supplements would be just a bit more wobbly, and these rules would conflict with each other from supplement to supplement. (If you've worked to ever make something physical, like a box, you're familiar with this: get one dimension slightly off, then everything else is off just a little bit more to match the first wrong dimension, and everything ends up fucked and your box wobbles.)

So when the grand refactoring of 3e occurs, putting everything possible into d20+modifier versus DC as the core mechanic and solidifying all the other parts of the game from that basic core mechanic and shared, common terms, it's attempting to create a solid foundation for 3e, not just in the core rulebooks but in any other supplements. And as a 3e player who spent a lot of time staring at these supplements, it worked pretty well. 3e isn't perfect, not by a long shot, but you can look at the vast majority of its pieces right from the 3.0 core rulebooks in 2000 to the final releases in 2007 and they use the same rules and they have the same mechanics underlying them. (You do still have the problem sometimes of overlapping, substituting publications though. It was possible to have a "legal" option from early in the edition that wasn't legal later on.)

How does this relate back to the trapfinding ability though? Well, it's pretty simple and yet also illustrates the design problems that 3e did struggle with. If you have a hammer, everything looks like a nail, so it made sense to tie the traditional "thief skills" into the overall skill system that 3e added. (That system was itself an evolution of non-weapon proficiencies in 2e, which were one of the biggest problems with wobbly overlapping rules. 3e offering a mostly-authoritative general list with the same core mechanic as everything else made a big difference in clarity and consistency.)

So, great, rogues are the best at skills, they've got their same skills as they've always had, but they'll be the best at them in terms of having the most and being good at them. And this also makes adjudicating those skills easier, because now there's no difference between a thief's climb walls thief skill or the Climbing nonweapon proficiency, it's just one Climb skill. (Thief in this case meaning the 2e or earlier class, and rogue being the 3e class.)

Except. If other people can take thief skills now. If they're all just the same skill. Why take a thief at all? You can just spread out the thief bits and have the wizard do it or something. So the rogue's trapfinding ability is specifically included as a form of "role protection," saying that you must have a rogue to disable these kinds of traps and no one else can do it. This trapfinding rule exists solely to protect the rogue's role in the party and make it so no one else can replace them.

There's actually very few rules like that in 3e that are specific to one class to the point of excluding others. Some feats work like that, but by and large 3e is explicitly built to allow you to get to whatever character build goal you want from a variety of starting points. All the way back in the 3.0 DMG at their introduction, it tells the DM that prestige classes shouldn't require specific classes to take, but instead have character statistic requirements that can be met in a variety of ways. (So instead of having a prestige class require being a 5th level barbarian, it requires the attack bonus of a 5th level barbarian and a barbarian's rage ability; this means you could meet the requirements with 1 level of barbarian and 4 levels of fighter, for example.)

This example is actually relevant to trapfinding, which is that it's a 1st level ability for a rogue, so you can just grab one level in rogue and do the rest of your levels in say, ranger and still be able to find traps just as well as a character of the same level who's only got rogue levels. (And you very specifically want that level in rogue to be your very first level, because 3e's skill system is a goddamn mess and if you do it any other way you will be catching up literally forever. But I digress.)

Back to rules elide. Yes, 3e D&D has a very formalized system for trapfinding and disarming. To refer to Jared Sinclair's post, it's got a system that's like his Example 2 about lockpicking. But I think the how of how 3e gets to that formalized system also tells us a lot about the game, as Jared discusses with his examples 3 and 4. Indeed, 3e wasn't telling us that locks or traps aren't important: one of the slogans used for its release as 3.0 or around then was "Back to the Dungeon!" - WotC wanted you exploring ruins and disarming dangerous traps and fighting monsters and wresting magical treasures from them. That's very clear if you look at the adventure design guidelines in the 3.0 DMG, 3e is absolutely trying to be a classic dungeon crawl game yet again. (This is a response to 2e, which got away from the dungeon for grand plots and frankly weird, stumbling, adventures stemming from the popularity of the Dragonlance module series.)

But trapfinding still isn't the only way you can interact with a trap in 3e. Indeed, the trapfinding ability itself notes this - most characters can try to find traps using the Search skill, but only the rogue can find the vast majority of them (basically anything harder to spot than a pit trap covered with a rug). Most people can try to disarm traps, but only the rogue can disarm magic traps. If you've played a dungeon crawl though, not all traps are disarmed. Players do other things: you go around a trap, you avoid it, you set it off hoping you'll be out of the area of effect, you hack it to pieces before it can go off. And 3e does allow for these interactions, because it very clearly has dungeons mapped out on graph paper with 5 foot squares (yes, 5, not 10 feet) and its traps do describe their area of effect and use the common mechanics for how to resolve them (so an attack roll versus AC, or a saving throw to get out of a trap's way.) Again, we're seeing that refactoring 3e did in the design here: 3e doesn't have special rules for resolving a trap's effects because wherever possible a trap uses the same rolls and rules as standard attacks or spells. Breaking a trap works the same way: there's no specific rules for breaking traps, but there is a common set of rules for breaking or attacking objects, and traps follow those rules so you can resolve breaking a trap with those same mechanics 3e has already given you.

I think the reason that I was asked if "rules elide" applies to the trapfinding ability is because the trapfinding ability says a lot about how the game is supposed to work in 3e. It does formalize using Search and Disable Device to interact with traps, and it does carve out that place for the rogue and only the rogue to be doing this activity. However, this isn't really any different from previous editions of D&D, because thieves also had Find/Disarm traps in their thief skills. 3e has just expanded these to general skills that mostly anyone can use, it's just that the trapfinding ability is required to make the best uses of them.

However, I do think "rules elide" applies differently to interacting with traps overall in 3e, and I do think it was an intentional choice on the part of the designers. If we go back a couple paragraphs, you'll note that I've listed the standard ways most PCs interact with traps, and how 3e does provide the mechanics for resolving those interactions. But what about other interactions? What if you want to do something non-standard? What if you got a giant lodestone and just rolled it through a dungeon and set off all the traps that way? The 3e rules can't account for that, can they?

No, they absolutely can't. They admit as much, on page 9 of the 3.0 DMG, that "often a situation will arise that isn't explicitly covered by the rules." with some tips on how to resolve such situations. However, we've just spent a lot of time and space discussing how these standard interactions are accounted for, and I think that is actually the elision: a flip in game design principles.

Think about how an OSR game is written. You write up your classes, your races/ancestries, your spells and your monsters. Writing an OSR ruleset isn't really a tabula rasa experience, but what you're doing is essentially trawling the dream sea: you're picking things you like and explicitly putting those in the rules and saying "okay, fighters in my game work like this and fireballs explode like that and dragons are this dangerous." Anything you don't say isn't set in stone. You're not saying people can't play other classes or have other spells, but this is the start of what you've put together.

Writing the 3rd edition of Dungeons and Dragons is a little different, because there's no tabula rasa at all. It's the new edition of D&D, for crying out loud! So when you're making those same choices and putting those things in the rules, you're dealing with a lot of existing expectations about what D&D should contain in its races, classes, and so on. (There was a huge outcry over the 4e PHB 1 not containing gnomes or druids, for example.) And 3e contained many of the classic standards, including ones that hadn't been included in 2e's core rules: half-orcs, monks, assassins.

It's not just limited to character building blocks, though. It's also the process of figuring out how people play the game and either facilitating that or telling them what needs to be done differently. And again, 3e contains the rules for the common ways to deal with traps - and no uncommon ways. To echo Jared's point, by providing resolution mechanics for the common interactions with traps, 3e is also suggesting that other ways aren't meaningful to the game's rules. They're possible (you can lodestone the trap) but you just do the fallbacks for situations that aren't covered by the rules. And I think this was intentional by the designers. Go back up and look at the discussion of how 2e supplements kept reinventing the wheels of mechanics like sailing and weather; each one of those was performing its own elision about how to resolve sailing mechanics for your ship campaign, and none of the pieces fit together because those elisions weren't the same. In contrast, 3e is making the choices about what to elide right now, in the core rulebooks, and is just leaving the rest out.

In other words, 3e's design principle isn't "provide some stuff, and let the players fill in the rest of the possibility space" it's "provide everything we can, to make the possibility space manageable collectively for everyone playing the edition." And that is a deliberate choice, because 3e was responding to a previous design model that hadn't been working for years. Now, this assumption that 3e's designers can provide everything might seem arrogant, but I think it's less arrogant than it appears. 3.0's designers (in contrast to 3.5's) were a lot of old TSR hands. WotC essentially absorbed the entire TSR design staff, got them to do 3.0, and started their infamous Christmastime yearly cuts AFTER 3.0 was released. So the people who made 3.0 had a lot of experience designing D&D already, and were able to draw upon that experience when providing everything they could, because they had a fairly good idea of what most D&D players did or wanted or required from the game already. They weren't blank-sheet thinking about "how do PCs interact with traps," they had decades of experience telling them what needed to be accounted for. Therefore, the common interactions were accounted for in the published rules. 

(As a related aside, this approach lead to a misconception about 3e's design philosophies: a lot of people look back at it and think it was some sort of simulationist attempt to create the guiding physics and metaphysics of a complete D&D world because of everything 3e covers, and then those people get upset when the pieces don't come together to create a complete world simulation like if it was GURPS or HERO System. [Common examples are the peasant railgun or the value of ten foot poles and ladders.] That's a misunderstanding. Instead, the pieces 3e provides are meant to account for the games and experiences people had been playing D&D for, and are meant to be consistent as game mechanics, but there's no effort to simulate a world beyond the players themselves. [Note the word simulate here. 3e cared about the world beyond the players but it didn't attempt to mathematically or philosophically simulate it as a complete system. The game doesn't concern itself with providing details on how much money a town of NPCs makes, it tells you how to build that town as a play experience and have the PCs go to a tavern or do some shopping or have a quest there. That distinction is often lost because people get confused by the existence of say, trade good tables. Trade goods exist because they're treasure or plot objects for PCs, not to simulate a whole economy around the PCs.])

Now, does the rules elision around traps in 3e matter? It absolutely does, because the decisions of what statistics apply to traps and how to interact with them means that there's player characters who are good at traps or not good at them. Related to "role protection" is the idea of "spotlighting," or who gets to spend time being the important character doing things in an RPG session. If the rogue has the trapfinding abilities, then the rogue more frequently gets the spotlight time about traps because they're the best at overcoming that obstacle. They get to be the center of attention with traps. (This isn't solely reserved for the rogue though. Adamantine weapons in 3e D&D can cut through basically any non-adamantine material like butter, so there's often a fun experience of a PC getting an adamantine weapon and then resolving traps or locked doors for awhile by just cutting walls open and going around any obstacles. Same obstacle but a different PC gets spotlighted.)

This issue is even stronger when you think about how 3e does character building, with characters accruing new and better abilities as they level up through a series of player choices. There's absolutely a social pressure to give the spotlight on dealing with traps to the rogue when not only are they the rogue, but they've taken a feat to make them extra good at disarming traps and spent their extra money last downtime on a set of magic thieves' tools. Consequently, the other common options of how to resolve traps fall to the wayside, because the social spotlight goes back to the rogue. And even more so, the uncommon options get completely forgotten. (Don't need to carry the lodestone when the rogue can disarm anything short of a god's own chastity belt.)

And this is why people feel that 3e and later games encourage "playing from your sheet" in terms of just using the abilities there, instead of being creative. The sheet improves as a function of player choice, and people want to see that choice made meaningful in play, so it demands the spotlight. A felicitous imagination of what to do is replaced.

I don't think that's any different in OSR games, though. (At least traditional ones, like S&W or OSE. Your game may vary.) PCs still gain levels, and magic items add new capabilities. Eventually, players do figure out a best option or a best character for doing something. They decide upon their own procedures, and they do carry the lodestone around to each and every dungeon. A lucky roll on a treasure table gives the wizard a scroll of passwall, the wizard makes the roll to scribe it into their spellbook, and then the whole party can enjoy just getting around traps instead of having to deal with them for awhile. Eventually, the same social pressures and spotlights will emerge, and in my experience it doesn't even take very long; a few sessions in and an OSR group will probably have figured out a go to mapper and caller and those get spotlights the same way.

The distinction is just that there's less explicit rules elision in these OSR rulesets, generally, and that there's less agency in OSR games to allow players to craft the specific characters they want to get the specific outcomes and spotlights they desire. (It's the long-term effect of rolling a random character and getting attached to their story, instead of creating an OC and being attached to their story before seeing it in play.)

To conclude, rules elision is a thing in D&D in general. If you don't want it for trapfinding, you're going to have to go back to OD&D and skip the Greyhawk supplement. And that's a legit way to play! But it's certainly not what the designers of 3e D&D were doing 20 years on, and I don't think it's reasonable to think that they would. Is 3e still D&D though? Yes, absolutely.