One of the first rules of user interface design is that good design is invisible. This also applies to game design. An excellent book on the subject of user interface design, and indirectly on game design, is Scott Krug’s Don’t Make Me Think. Krug’s primary thesis is that reducing options is the key to a clean design. Have your design anticipate what the user will be doing and re-use design elements and methodologies. Above all else, minimize options.

A common ‘failure mode’ in game designers – especially beginning ones – is throwing everything into a game design that the designer things someone somewhere might want, often without asking “is this ever going to be used?”  For example, RIFTS vehicular combat uses an entirely different set of mechanisms than RIFTS personal combat, which uses an entirely different set of mechanisms than RIFTS psychic combat.

Unfortunately, there are strong business incentives against following Krug’s advice.

We’re going to look at the rise, rebirth and implosion of D&D 3.5, and how it changed game publishing at the top of the market for RPGs

D&D 3.0 came out after a four year lull in the brand. There was a lot of pent-up demand. There were marketing surveys run by Ryan Dancey that showed that there are five basic types of roleplaying players, and a good viable group needed to have at least three or four of the five. There was marketing data that showed how long a typical campaign ran in years, with sessions per year, and that was used to determine the roughly “13 encounters per level” metric.  They also made it harder to die, and easier to recover from being dead; in some ways this damaged the narrative fiction underlying D&D style fantasy – if you could haul Ferd the Fighter’s arrow-ridden corpse to any city of about 10,000 people, and pay 6800 GP, he could be raised from the dead, and two installments of 1280 GP later, the two negative levels could even be erased.

D&D 3.0, unlike any of the prior editions, was truly designed.  And by designed, I mean that it used consistent ways of doing things.  It got a massive public playtest, one of the first in the industry.  It unified a lot of clunky 2e mechanics. And after being on the market for three years and four print runs, they did a major revision to the game engine with 3.5, spreading some of the rewards further out, adjusting a number of mechanics that just weren’t working as well as they’d like.

WoTC worked to keep 3.5 mostly compatible with 3.0 materials, and they focused on the OGL. This meant that they could let a million flowers bloom, a boom of creativity, removing the barriers to entry for thousands of game publishers!

What WoTC anticipated was that third party publishers would go after the adventure niche…which tend to be labor intensive, and sell adequately, but not sell as well as other products.  An adventure takes more work to write and get publishable than twice as many pages of new spells, magic weapons or character creation options; a GM can cherry pick character creation options; an adventure, if it’s going to be a worthwhile investment for the fellow working behind the screen, has to work with minimal adjustments.

Adventures will sell about 1/4 as many copies for that work investment: You sell adventures (and bestiary books) to the person who runs the games. You sell the character bling books to nearly everyone at the table looking for a new way to trick out their characters.

A side effect of character bling books is that each one published makes an adventure harder to write, because of the new abilities someone might have as a player. Even worse, since D&D 3.5 has most monsters written up as player characters, the amount of work needed to generate stat blocks is comparable to making a player character…whose job is to come on stage and be a hit point sponge.  By adding more options, the game became harder to run.

However, character creation options sell. Third party publishers figured out where the money was. And we got splat books. And books of options. And advanced books of options. And prestige class books, and master-class books…and WoTC did the same thing on their production schedule, and within a year of 3.5 being released, there were so many options to choose from that the game was getting heavy and bloated.  Adventures also got asymptotically harder to write, at least around the idea of 13 encounters to level, because adventures that worked fine in 3.0 were getting steamrolled by characters built around newer options.

Pathfinder came about as, functionally, D&D 3.5 in 2008, after WoTC managed to kill the market for 3.5 third party materials, and retracted Paizo’s license to publish Dungeon and Dragon magazines.  (Paizo was, for all intents and purposes, the only publisher that made the 3.5 “third party adventure publishing” niche work well at all, by using a subscription model.  You bought the first book of an adventure path, and subscribed to the next five volumes which came out roughly every three to five weeks, so you could stay just in front of your players.)

After six and a half years on the market, Pathfinder is also getting chromed up with character creation options…and playability is declining; if you don’t have the latest stuff, the newest scenarios are hard to play. If you use all the latest stuff, the oldest material is boringly easy.