Ever since Paizo announced this book and the playtest back in August, I have been thinking a lot about classes and my philosophy about what concepts deserve their own base classes and which don't. Personally, when creating a class, I feel like the designer should ask themselves three questions:
- Does the class' concept have enough "meat" to build a full class around it?
- Does the class' concept fill a role that is currently missing in the game?
- Is the class necessary to play the concept from the 1st level?
Before you start working on the mechanics of the class, you need to figure out if the concept has enough "meat" to build a class around it. For example, if you're building a class for Pathfinder, there needs to be enough there to fill out twenty levels. If you can only fill out a few levels, it is probably better off as an archetype or a prestige class.
Also, I believe its important to design classes that fill missing roles in the game. The reason why I believe this is important is that it gives the new class a niche to fil. For example, you could design a class that creates and utilizes technological devices for Pathfinder. Since there really isn't a class that fills that niche (if you don't count 3PP classes), this new class would have a unique place in the game.
Finally, the concept should be something you can't already build with the current options at 1st level. For example, you can build a Swashbuckler character using the current Pathfinder rules. However, you will most likely have to wait a few levels, choose a bunch of feats, take a level in the Duelist prestige class, etc. However, a Swashbuckler class allows someone to play a popular fantasy archetype from the very 1st level instead of waiting until level 5 or so to play it.
Now, I readily admit these are only my personal opinions on the subject and I know a few people will probably disagree with me. However, I stand by these three questions for the most part and they will be on my mind when I'm reviewing the playtest document tomorrow.
What are your thoughts on designing new classes? How do you judge if a concept is worthy of a full class? Finally, what concepts do you believe deserve their own full class?