Author: Robert J. Schwalb

On the road to see a flick in Nashville, my non-gamer friend, Trey, was asking about how RPGs worked. Having never played an RPG before and probably having never had much interest in the hobby in the first place, this was all new territory for him. I gave the usual pitch—you create or choose a character and then make decisions for that character, blah, blah, blah. You know what an RPG is, so I’ll move on.

In the thick of the discussion, my sobriety-afflicted little attic wiggled and seized on an idea, one that crystallized why and how the experience one has playing the game has evolved to the point that the way we play “mainstream” RPGs would be unrecognizable to our older selves. I think it all comes down to the problem of increasing complexity.

The big coup that made 3rd Edition such an innovative game was that it took all the little independent systems lurking in D&D’s dark and scary corners, brought them into the light, and pounded on them until they all used the core task resolution mechanic. Old D&D felt like it was cobbled together from a variety of resolution systems made ad hoc to resolve situations that cropped in play. Attacks with melee weapons and ranged weapons both rolled a d20, but they took place at different points in the round and each had special rules. Dodging a fireball, resisting a spell, or withstanding poison all used saving throws, but the numbers varied by class and level. And don’t get me started on psionics or unarmed combat.

Distilling everything down to a common mechanic obviously made the game easier to learn. After all, once you figure out how to add a number to a random number, you can play D&D. Sort of. Initiative, attack rolls, saving throws, and everything else you would do that had an uncertain outcome relied on the common mechanical procedure for determining success and failure. Slick, sexy, and fun to play, and which is why the game is still being played in one version or another almost 15 years after it first appeared on the shelves.

As awesome and as powerful as the 3rd Edition rules were for D&D, I think it came with a deeper and far more insidious transformation. And that transformation finds its roots in character customization outside of class and race. Introducing rules exceptions by methods outside of the big choices made the game fuzzier and, in my opinion, harder to master. Non-weapon proficiencies may have been the initial culprit, but I would argue that it was the line of complete books in 2nd Edition that gave the final push that would drag the game down the rabbit hole into the murky and strange world in which we now live.

The difference between now and then is that though game-complexity started higher (the learning curve for AD&D was pretty damned steep), it stayed relatively flat once you figured out how it worked. You could create a dozen fighters, with different races, alignments, personalities, weapon assortments, but they all played more or less the same. The more you played a fighter, the more second nature it became until you could always play your fighter effectively. Your big contribution to game-play was high accuracy and consistent damage output most of the time. That might not seem like much fun, but given the necessary abstractions to make combat simulation manageable, the mechanics did the job.

Obviously, spellcasters were complicated. Keeping track of all the spells is tough and so players who chose to play magic-using characters had their work cut out for them. But, again, once you get over the hump, does complexity really ramp up? I’d say no. Even if you’re preparing 35 spells each day, you eventually settle into a pattern after you figure out what spells work well and what spells don’t. There’s not a lot of mystery to magic missile after the first time you cast it.

At the time, 2nd Edition felt like a radical change from 1st Edition and, in hindsight, the nerd rage I felt then is kind of amusing when I look at the rulebooks today. The guts of that system are basically the same and, I would even say, better. We lost a lot of the really arcane mechanics and the system made great strides toward normalizing how you did things in the game. And, best of all, you could still enjoy the same experience of system mastery you had with 1E in that a fighter was pretty much a fighter every time you played. But the Complete books changed that. Not only did they help push the non-weapon proficiency system away from being an optional set of rules, they made class functionality and definition fuzzy. You might play a myrmidon in one campaign and a gladiator in the next. Both are fighters, yes, but the fighter identity—what I do in the game—changed by allowing players to select a kit. More options are good, yes? Maybe? Maybe not.

Kits just kicked open the door for more customization. Since there aren’t moving parts in classes aside from spells, the game needed other elements for players to choose, elements that would become skills and feats. Kits from 2nd Edition didn’t really have mechanical weight. They just added flavor. Since people gravitate toward those things that make their characters more powerful, skills and feats needed and got mechanical weight, which, in turn, raised their importance for making an effective character. And, since the game needs to reflect growth, the system had to parcel them out over time. The net result is great texture and mechanical weight attached to things that used to just be how you described or roleplayed your character. But it had a huge price tag.

From 3rd Edition on, your character, regardless of whether you chose a fighter or a wizard, always becomes harder to play as you gain levels. A character is a veritable catalog of permissions (I can attack with this weapon, I know this thing) and exceptions to the rules (I can sacrifice accuracy to boost damage, I can maximize my spell damage). By the time you reach the top of the ladder, you have so many things to track and remember, so many widgets, what should be simple—I attack with my sword, is a veritable labyrinth of yes and nos. The result is a slower, clunky game that more often than not comes screeching to a halt as everyone gathers around to sort out just what widget Q from sourcebook 76 actually does.

I know, I sound like a cantankerous old man and I’m really just whining about bygone days. But seriously, the prevailing opinion that greater complexity over time terrifies me. When did it become good to make a game harder to play the longer you play it? And I’m not talking about harder challenges; I’m talking about making basic decisions and resolving them as you take on those harder challenges. Why would we ever build-in a disincentive to play the game more if the goal, as producers of product, is to encourage customers to use our form of entertainment instead of something else?

The genie is undoubtedly out of the bottle. The existing audience demands cool and interesting rewards to act as carrots for continuous play. However, the methods we use to deliver these rewards and to determine what awards ought to be is critical for controlling complexity bloat.

One of the things I liked a lot about 4E was the attempt to control complexity by having higher-level powers replace lower level powers. Of course, the game shot itself in the junk by piling on extra powers that were additions rather than replacements, but the core idea is sound. Widgets that retain their core functionality and just improve at later points would be an easier way to deliver these extras. Maybe instead of getting 7 feats, you gain 3. And at the points when you would normally gain a feat, you upgrade the effects of the feat(s) you already have. Power Attack, in this model, might be a flat penalty to the attack roll to get a flat boost to damage. Let’s say –2 for +4 damage. At level 6, the penalty might stay at –2 and the damage would jump to +6 or +8. Just a vague idea, but I think something like this would demonstrate improvement, preserve character customization, and control the excess of complexity by piling more stuff on top of the stuff you already have.

So, what do you think? How often does increasing play complexity end your campaigns compared to finishing the campaign’s story? Is complexity a problem in your games or is it something you work around?

I’ve not made a secret my disdain for skills in d20-based games. They always feel tacked on, a sop offered to players who need more numbers to truly realize the character they want to play. Rather than just say, “My dude spent a lot of time in the king’s court” and let the GM interpret what this means, most players feel it’s important to have mechanics that represent capabilities beyond the core task resolution system. Today, I want to talk about why I feel this approach is bad for d20 games and probably make you super upset as a result. But I have a solution and it comes at the end.

Where to begin? Every RPG needs a clear and easy way to interpret whether what a player describes happens or doesn’t happen. Every description can be distilled down to a yes or no. While some games explore ideas of “yes, but” and “no, but,” the “but” only slaps on a consequence for not quite getting to the positive outcome or dipping into the negative and provides a bit more nuance and story fodder for folks who want it. It’s still a binary outcome though. Ultimately it happens or it doesn’t.

In old editions of D&D, the game had individual mechanics for tasks and sometimes those mechanics employed completely different systems based on whether or not you were a member of a particular class. Cast a spell? The target might make a saving throw against spells. Swing a sword, roll the die, add your modifier, and tell the DM who then checks the table. Want to pummel a dude, try something else that’s not quite so arcane or involved please. Proficiencies, introduced in 1st and spread to 2nd as an optional rule, provided yet another method for resolving tasks. It was a great, big complicated mess. When 3rd Edition showed up, all slick and sexy with a unified task resolution mechanic, it was a revelation to me and it’s still remarkable how much the game changed when the edition meter rolled over to 3.

As awesome and as powerful as a core mechanic was for the game, it fundamentally altered the role ability scores played in the game and in a manner that reduced abilities to nothing more than descriptors. In the older editions, the six abilities pretty much told you everything you needed to know about your character and the game system told you how those abilities plugged into specific tasks. I know when I ran 2nd Edition, whenever someone described an action that wasn’t an attack or a spell, there were three possible outcomes: yes, no, and roll a d20 to see what happens. If the player rolled under his or her score, the action happened. Otherwise it didn’t. This technique comes straight out of the game’s proficiency system, so I wasn’t doing anything revolutionary. But the main thing is that the abilities were always where the player looked to determine whether a described action happened when the outcome was not certain.

From 3rd Edition on, ability scores primarily served as the method to derive another number, the modifier that plugged into the core task resolution system. Aside from a few weird corner case rules, the scores only functioned as another kind of hit point pool. A special attack could damage or drain a character’s score, and thus reduce the modifier. And 4th Edition didn’t even have that.

So what’s the problem? If the modifiers were the only thing a character added to his or her task resolution roll, I probably wouldn’t have much of a problem aside from the fact that the score itself is just character sheet clutter for much of the game (True20 solves this problem by just jettisoning the score). But the modifier is not the only thing you add to your task resolution roll and skills bloat the numbers until they are the only thing that matters.

Both 3rd Edition and 4th Edition effectively abandoned abilities as the primary method for task resolution by attaching numbers to ability modifiers that were larger than the numbers themselves. If you can have 23 ranks in Use Rope, the +3 bonus from Dexterity seems paltry. The most important component of the calculation to determine if you manage to tie the knot properly is the bonus from the skill. And if the skill ranks are the thing that matters most, why bother having the ability modifier at all? Wouldn’t it be simpler to just use a range of skills and have a character choose a couple of descriptors (Strong, Smart, Alert) that plug into the skills and increase the skill bonus? For example, you might have the Athletics skill and it might start with a modifier of 0 to 4 that can grow as you gain levels. If you have the Strong tag, you might get a +1 bonus to Athletics checks. While similar to what 4E did with skills (granting a big, heavy flat bonus instead of mucking about with ranks), it dispels the illusion that abilities matter.

But, carving up the corpses of abilities into smaller pieces encourages specialization and, when those pieces directly interact with the game’s predominant play-style (combat, intrigue, exploration, whatever), survival impels players to focus develop those pieces. We see this all over the place in 3rd and 4th. Choosing the History skill in 4th Edition over Perception is a stupid decision unless the game is focused on dredging up dusty bits of trivia instead of finding treasure and butchering monsters. Similarly, putting ranks in Knowledge (Architecture and Engineering) is foolish if you take those points away from Listen/Search/Tumble. Such focus undermines the whole point in having skills in the first place. To realize that wonderful character you’ve imagined in your head, you have to make bad build decisions about your character. Efforts to mitigate this problem by making flavor-skills more mechanically useful only managed to create weird narrative problems. If having History gives your dude a big fat bonus to attack rolls every now and then, your game is likely to be filled with historian-fighters. That seems weird.

Let’s wrap this up. D&D is chock full of big abstractions and the biggest is in how the big six abilities claim all the character’s descriptive real estate. The game, at various points, tells us that ability scores are a combination of natural talent and formal training. Intelligence isn’t just your character’s IQ, but it’s also your character’s education, wits, and recall.

The non-weapon proficiency system is a long way from perfect, but what it did was plug into the existing system architecture without diluting the importance of ability scores. Simply, you roll a die and if you equaled or got under the score (+ or – the check modifier), you succeed. Boom. Dead simple. And if you want to improve your proficiency, you can just take it again, gaining a +1 bonus to the check for each time you took it after the first.

Adopting a system like this for your d20 game isn’t that hard, but it would mean stripping a fairly big chunk of the task resolution system out of the game. The big thing is to flip the chance for success. You’d need to double the modifier for all checks (not attack rolls or saving throws) and make the baseline DC 10. Drop the DC to 5 for easy tasks and jack it up to 15 for hard tasks.

Rather than have a slew of combat skills like Spot, Athletics, and so on, you just lean on the abilities. If you want to get better at climbing, make darn sure you put your ability score increases in Strength. Skills would become things you have to have special training to pull off: Animal Handling, Picking Locks, and so on. And those skills would modify the DC based on their complexity. Feats would just cover everything else.

Well that took a while to get to, but I think this is pretty cool. What about you?