Not Actually Why Fourth Edition Died

The problem with Fourth Edition Dungeons & Dragons, arguably the biggest problem that any edition of D&D has ever faced, is that the people who were designing it didn’t know a goddamned thing about Design Space.
- Kris Hanson, "Why Fourth Edition Died"

A Pre-emptive Critique of D&D Next, part 2:

While pondering the next point in my own in-my-head series of essays on the coming D&D Fifth Edition and why I'm not looking forward to it, I ran across a pretty impressive essay by a Canadian game store owner named Kris Hanson. He argues persuasively, in both the linked essay above and a followup to it, that the failure of 4th Edition D&D was a failure of design, and specifically, a failure of design to do the sort of things that are routinely done in Magic: the Gathering, Wizards of the Coast's (WotC's) other major game product.

Hanson isn't saying that D&D needed to be more like Magic -- rather, he's arguing that the design function in D&D should be seen as much more like the design function in Magic, and that the D&D designers simply failed to mine their design for as much interesting stuff as the Magic designers have. I find the argument fascinating because, at heart, Hanson is arguing that 5th Edition should incorporate exactly what I think WotC will incorporate for 5th Edition, and is precisely why I think 5th Edition will fail. Specifically, I think this is one of the two pitfalls I talked about last time -- that the designers might fail to define the game sub-systems specifically enough to create a truly modular game.

Hanson is right about one thing -- Mark Rosewater, the head designer of Magic: the Gathering, talks a lot about Design Space. Specifically, he describes how Magic design can fill two basic and seemingly contradictory human needs -- familiarity and novelty -- by taking a familiar game mechanic and tweaking it in new and hopefully interesting ways. Hanson then goes on to list a bunch of things that D&D could have done in this 'design space' vein, had D&D designers been thinking like Magic designers:

The design space here is mind-boggling. Off the top of my head: a whole list of Encounter Powers that could be made better by giving up a Daily; Encounter Powers that take two full rounds to use; Daily Powers that change when in certain circumstances (underground, flying, in the sunlight, whatever); a list of At-Will powers you can give up using for the rest of the encounter to make way better; a list of At-Will powers that are as strong as Daily powers, but use up the character’s healing surges as a cost; Lifetime powers, powers that can only ever be used once, ever; uncontrollable powers that go off in certain circumstances, but the player has no control over them; come-back powers that allow you to do incredible things for the cost of an action point; counter-powers that allow a player to stop another player or NPC’s powers from going off as an immediate interrupt; mind-drain powers that steal another character’s powers for a turn, or treat that power as spent; dramatic powers that allow you to influence story events in subtle (or, for dailies, not-so-subtle ways)…

For starters, I'll point out that some of these 'design spaces' are actually being mined, but by module writers, not the D&D core designers -- for instance, I'm running a D&D Living Forgotten Realms adventure tomorrow morning that incorporates an 'uncontrollable' power for those who possess a specific story award from a previous module. The core problem with Hanson's thesis, though, is that D&D designers can't think like Magic designers, not always, and there are good reasons for this. Hanson himself already intuited one:

What the designers of D&D tried to do was create a homogenized environment in which none of the powers effectively broke the rules, and that’s really the wrong way to build a game that’s made out of a bunch of crazy powers.

If you have a game, like Magic or Pathfinder, where the game is about characters that have a bunch of crazy powers, then perhaps that kind of design perspective is desirable. (Although frankly, the ultimate 'characters that have a bunch of crazy powers' game has already been written, so why bother?)

I'd argue that 4th Edition D&D is a game about characters doing their jobs for the greater good, and as such I am absolutely convinced that the designers of 4th Edition D&D had a well-defined structure for all the game-mechanical sub-systems of their game. In some cases I know this because the designers told me so -- for instance, in the Player's Strategy Guide, there's a chart that lists the expected values of the game's core numbers (AC, attack bonus, etc.) by level. In other cases I can intuit this based on the choices that the designers made for some game sub-systems, and more specifically, the choices that the designers chose not to make.

Let's take feats, for example. Feats were first introduced in 3rd Edition D&D, and I can honestly say that the designers really had no idea what they were doing when they created the feat system:

• Can feats give you skill bonuses? Sure. (+2 to two specific skills [Alertness], +2 to any single skill [Skill Focus])
• Can feats give you saving throw bonuses? Sure (+2 to Fort, Reflex, or will, depending on which feat you take)
• Can feats change the way you interact with core game rules? Sure. (Blind-Fight changes the way you interact with the concealment rules, while Ambidexterity and Two-Weapon Fighting change the way you interact with the 'fighting with a weapon in each hand' rules)
• Can feats change the way your spells work? Sure. (Metamagic feats, which were an evolution of the AD&D spells that modified things like spell durations)
• Can feats allow you to do things other characters can't do? Sure. (Things like Spring Attack and Whirlwind Attack, which I think are specifically what Hanson, and many D&D players, is specifically thinking about when he talks about feats.)
• Can feats grant you class abilities?

It is specifically that last item that convinces me that the 3rd Edition designers really hadn't defined the concept of feats and were just using the idea as a 'catch-all' for cool concepts that didn't fit anywhere else. For example, your character's weapon proficiencies are determined by which class you take (fighters can use any weapon that isn't designated as 'exotic', while wizards have only a few specific proficiencies), or to a degree by which race you take (a hold-over from old-school D&D, where what are now races were once classes). By taking the proper Weapon Proficiency feat, you could represent the ability to use without penalty a weapon that your specific class or race wouldn't normally be able to use (such as a wizard wielding a two-handed sword). But the real kicker, in this respect, was Weapon Specialization.

In the 3.0 Player's Handbook, the list of fighter class abilities includes something called 'Weapon Specialization', which is an ability that fighters could gain at 4th level or later. But the actual ability was defined as a feat -- one that a fighter could take either with a fighter bonus feat, or by using one of his general character feats that he'd gain every third level. And because Weapon Specialization is a feat, it is listed in the feat list and described in the feat descriptions just like any other feat, even though only a fighter can take the feat. (This would eventually be 'fixed' in Pathfinder, when Paizo's designers realized that assuming that every fighter wanted to spend feats on Weapon Specialization, Weapon Focus, Weapon Mastery and the like was just silly and reconfigured all these abilities to be fighter class features.)

In 4th Edition, meanwhile, feats were much better defined. If you took a feat at a particular tier (heroic, paragon, epic), you knew what kind and size of bonus it should apply (+1 to attacks per tier, +2 feat bonus to skill checks per tier, etc.). Most important, if you had an idea for something cool, you could tell if it should be a feat at all, and if so, what kind of feat. If it was something that modified your ability to interact with the core game rules, it might be a feat if it was accessible to any character; if it represented a particular talent that a given race had for something, it would be a racial feat and be listed in a different feat list. If it represented a modification to an existing class feature, it would be a class feat and be listed in yet another area. If it allowed you to combine the abilities of two classes, it would be a multiclass feat, etc. And, if it didn't fall into any of these buckets, it wasn't a feat and you'd have to find somewhere else to put it -- perhaps make it a power and define it for certain classes or make it an advanced skill option.

It is this last thing -- the ability to say 'this cool thing does not belong in the feat list, we need to find where it does fit' -- that gives the game both structure and extensibility. Because feats do certain specific things and not other things, you can create a concept like 'themes' (a combination of what would otherwise be powers and class abilities) and discover that you haven't just created something that makes feats irrelevant or that is irrelevant when taking feats into account. (Contrast with the late 3rd edition concept of the 'skill trick', an idea which basically boils down to "these things look like feats but are too weak to be 'real' feats, so we'll make you pay skill points for them instead", except that any of them that simulate actual feats will both appear to be and actually be pointless when compared to an actual feat.)

Similarly for powers -- I've never seen anything that specifically defines how to construct a given power, but I'm convinced that there are at least ground rules if not a full design document for power creation. The rule would be something like this: an at-will power deals 1 die of damage and has some additional minor effect (applies a status condition until end-of-next-turn, can be used against a small number of additional targets, etc.). An encounter power of level X does Y dice of damage; subtract dice for each of the following that applies: it can be used as a close burst or close blast, it can affect multiple targets, it applies status conditions, it applies those conditions as 'save ends' rather than 'until end-of-next-turn', it has increased effect when used with a particular game-mechanical object or if used against an opponent which a particular keyword. The laundry list of 'design space' that Hanson lists seems superficially interesting, but without a clear analysis of the relative effectiveness of the various power options, all you're doing by providing such a list is giving players a chance to data-mine the powers list for min/max options -- the chance to figure out which ability gives the best positive punch for the smallest drawback. (And nowhere is this more sickeningly true than the munchkin-friendly 'merits and flaws' systems that Hanson seems to like, but I've learned to react to with the same level of revulsion as if I'd found a rat skeleton in my restaurant stew.)

However, there's an even bigger problem with Hanson's determination that D&D design needs to be more like Magic design, and it has to do with this comment from the original essay:

Understand that the only interesting powers in the whole of the game were those in the Player’s Handbook and those in the Player’s Handbook 3. The originals and the _only_ set of powers that were at all different or strange. One could argue for the powers from Essentials, but seriously, fuck one.

No, really, there's a reason that the Essentials powers (and to a degree other game elements) have such a different feel to them than the main powers, and though I hadn't considered it until I read Hanson's essay, I know believe that they were D&D's attempt to become more like Magic.

When Hanson thinks about Magic card design, he's thinking about the splashy cards that break the rules of the game -- the cool rares, in other words. But there are lots of cards in Magic, including a lot of cards that are deliberately bad. When I think of Magic design, though, I don't think of splashy rares -- I think of Volcanic Hammer and Shock, and what they mean.

You see, the very first printing of Magic: the Gathering included a card called Lightning Bolt. Later, the same company both designed and printed cards that were strictly worse cards than Lightning Bolt -- Volcanic Hammer did the same damage but cost more to cast, while Shock cost the same to cast but did less damage. Neither of the later cards had any mitigating condition that made them comparable to Lightning Bolt -- anybody looking at the three cards could easily see that Lightning Bolt was the best and would use it in preference to the others. Then why were the other cards even printed? Because you don't always get to use Lightning Bolt.

Magic: the Gathering has successfully promoted the idea of 'environments': specific groups of cards and card sets which can be 'legally' played. In theory this only applies to tournament play, but since the concept of an 'environment' also creates some sense of competitive balance, many casual players choose to abide by a particular set of environment rules as a way of limiting what cards they can expect to see out of their opponents' decks. At one extreme, an environment can consist of 'only the cards we open at the table during this event', while at the other extreme, an environment can be 'any Magic card ever printed by WotC in any format'.

While some chose to view Essentials as WotC's attempt to stealth-promote 'D&D 4.5' (I did at the time), a better analogy might well have been that WotC was trying to produce a 'D&D Standard' environment as opposed to the 'D&D Modern' of 4th Edition. This makes even more sense when you consider the character creation rules of the D&D Encounters program -- the characters for Encounters are meant to be created using the Essentials books (the 'core set' for D&D Standard), plus the most recent published book (the 'expansion set' for D&D Standard). The older 4th Edition books, and the books that were no longer the most current 'expansion' books could be used in 'D&D Modern', but not in 'D&D Standard', and some DMs might well say that the Essentials books weren't legal in his 'D&D Modern' environment (which helps explain why there are whole series of weapon-related and defense feats in Essentials that are strictly better than anything published in 'core 4E').

The problem with this is that many people playing D&D have been playing for as long or longer than they've been playing Magic, and the idea of separating D&D into 'environments' is silly to them -- in effect, all versions of D&D are 'Legacy' where anything published for an edition (and anything published for a previous edition, if it can be converted) is fair game.

In a sense, this makes the call to 'unify the editions' more sensible -- if the people buying and playing your game want to be able to use anything ever created as fuel for that game, then it would seem that an edition deliberately designed to allow for that would be a huge hit. But there's a peril in that, too, and it has to do with the other pitfall I pointed out in my first essay on the topic -- making things too specific that the modules have to be used in their entirety or not at all, limiting their appeal. More on that next time.