Hussar said:
Flavour First: The designer comes up with an idea first and then tries to create mechanics to best fit that idea.
Game First: The designer looks at how the game is being played at the table and creates mechanics to best facilitate that. Flavour is then added afterwards to justify the mechanics.
This is a neat way to look at game design, but I don't think it's the source of the problem you describe. Like others up thread, I think flavor should come first but that a positive feedback loop between both is necessary to end up with a playable game.
Your problem though is primarily with PCs that were designed for a different campaign then the one they find themselves in. Or maybe your problem is more generally with class design that's not universally useful in all kinds of campaigns.
After all, a Ranger in a borderlands orc-war campaign is going to be pretty bad ass. The Paladin though is going to be frustrated by always having to leave his horse and plate armor behind when it's time to spelunk or brachiate. Is that a problem with flavor first class design, or campaign design, or just a player-dm communications breakdown. I think it's the last one.
As a simple for-instance, I had to reengineer my character in my Iron Heroes group because I had designed him expecting more wilderness adventuring. Since we were all city, all the time it was frustrating that several of his feats and skills were never used. That was my fault for making a PC poorly made for the campaign, not the game's fault.
Is it possible to make a character that will be useful in every possible setting? Is it a good idea? Do they have a distinct flavor and character left if you do that? I think the answer to all three is: No.
Hussar said:
We've had flavour first mechanics for the past thirty years. I think that's more than long enough of a kick at the cat. It's time to let game first rule the table. Find out what works at the table, what results in the most fun (yes, I'm going to use the "f" word here) and then design to that.
I disagree. We play D&D for its flavor. Throw that out and we might as well be playing chess or TF2. It's nice to have good rules, but the flavor
must be right. When push comes to shove, flavor should win. Many people seem a lot more tolerant of minor rules inconsistency than flavor inconsistency. Just look at the thread on "Are Fighter Exploits Magic?". The rules must be well designed
and support the flavor you're looking for.
Mallus said:
Are you familiar with Champions or Mutants and Masterminds (both are effects-based point-buy superhero RPGs)? They are fine examples of systems which focus on providing mechanical frameworks for character abilities, with mere suggestions on how to define ("flavor") them in-game.
How does M&M handle flavor/rule conflicts? It seems to me that the flavor can either be anything or nothing, but not in between. If the rule say "Power A can do X, Y and Z, and is limited by 1 and 2." then only certain things (and maybe nothing!) can fit that description. Power Suits run out of fuel; electricity grounds when fire doesn't; boomerangs are slower than bullets; etc. When you're playing M&M (or any game where flavor is "tacked on" by the players) you need to decide ahead of time "Okay, when flavor and rule conflict (which is inevitable, even if rare), which wins?"
I say flavor wins, alway. What does M&M say? If it says "Rules win", then what's the point of your Cheops Drop Pyramid when it's only a pyramid "most of the time"?
My takeaway points are:
1. Great roleplaying needs great flavor, so flavor first. There's a reason
The Lord of the Rings inspires more D&D campaigns than ... well, anything, or that the new
Battlestar Gallactica is more likely to have RPG success than the 1970s version.
2. Fun times are helped by playable rules, so keep an eye on those; but they aren't the reason you play an RPG.
3. Make sure PCs are suited to the campaign. Paladins in Lankmar or Pirates in Mongolia just aren't good ideas.