When I first started working at Interplay, my goal was always to move over to development, climb up the ranks, and work on Fallout 3. Though I had played RPGs and CRPGs all my life, Fallout was the first RPG since the original Pool of Radiance to really kick me in the ass and excite me (Darklands was more of a slow burn). Well, as fate had it, I wound up working on the original Icewind Dale. I wasn't particularly thrilled with how my work on it came out but hey -- one step closer to Fallout 3, man. I then did some work on Heart of Winter, and that managed to actually be worse than my Icewind Dale stuff. Icewind Dale II came out better than I expected, but still, not exactly awe-inspiring. Along the way, I designed the magic sub-system for Torn, gave worthless spell/feat implementation feedback on Baldur's Gate: Dark Alliance, and helped come up with some of the background story for Lionheart. All of this gave me a lot of time to think about the work I'm doing on our unannounced "Jefferson" project (which is not a Fallout title).

The setting and game system of the project will immediately load almost every reader's idea of what the game should or will be. For a short period of time, this mattered a lot to me, and I worked very hard on thinking of ways to deal with those pre-conceptions. As time moved forward on the project -- and inevitably stopped during Icewind Dale II -- it became increasingly clear to me that I needed to focus on making Jefferson a good game on its own terms. This can be a very difficult thing to do when you're so used to moving quickly on a project and accepting the rat-a-tat machinegun fire of user opinions.

A lot of users can't adequately describe why they like certain games -- or even why they like certain features of games. While it's certainly true that developers can listen too much to users, there's a frightening tendency for developers to respond to this deluge of opinion by simply ignoring the users. I think that the best thing to do is to listen to the users, but really examine the source of their praise and their criticism. Beyond that, it is very important to notice patterns and look at what goes unmentioned. Of course, user opinions on message boards shouldn't serve as a democratic republic for approving game design choices; quite frequently, the majority can have really bad ideas. Hear what people have to say, but do what you believe to be best after all is heard and honestly considered.

That rambling preface is out of the way, so I'll dive right into some personal thoughts about "good RPGs". I would like to believe that I am helping design Jefferson as a "good RPG": one where the player is free to pursue many different attitudes, roles, and paths through the game's plots and feel rewarded for their choices. This last portion is particularly important, as it is too often ignored in RPGs. Many RPGs give the player choice, but make that choice either irrelevant or "wrong". Icewind Dale II had some particularly bad examples of this. No one likes to feel like the way they choose to play their character is wrong. That's like picking the top hat in Monopoly and automatically losing.

The trick is to find ways to make player choice matter without the designers killing themselves to support all possible paths. Those tricks lie in the details of implementation, with which I will not bore you. However, there are a few basic guidelines that designers can employ on a large scale to help facilitate the feeling of an "open" game:

• Leave as many elements of the character's background and appearance open as is reasonably possible. Male, female, good, bad, ugly, black, white, mauve, young, old -- if you can reasonably allow it, let the player define it. If you have no good reason to restrict it, don't. I know it sounds pretty simple, but most RPGers like being able to make characters their way. If they want to make their hero a huge Samoan banjo player, give them your blessing.
  
• Don't try to reward the player/character for things they didn't do. This extends to "the setup" of the story as well. All developers have had their share of "Chosen One" storylines, so there's no need to point fingers. "Oh, wowie! I'm the Precious Special Lad!" Who cares? If the story does revolve around a sort of "Chosen" background inherent to the main character, allow the player to DO something with that role. Shirk it, embrace it, pervert it. SOMETHING. And when the character arrives in a new location, let them work to build up a reputation and gain quests -- or DESTROY a neutral reputation and gain/create seedy quests. Old Widow Mabel shouldn't ask me to rescue her cat unless the people of Sleepy Oaks know that Solosolo the the Samoan banjo player is a badass hero of the people. If you constantly reward the character for just being somewhere, it ceases to be a reward. Choice doesn't really matter, because everyone is just going to seek the main character out with their problems. Give the player the ability to make choices that determine how they are rewarded according to their own tastes.
  
• Make the story about the character's choices, not NPC-driven narrative. Most designers like to think that all of the characters they write up are totally bitchin' and everyone who plays the game will love them to tears. This usually isn't the case, and you'll find people blowing large holes in the chests of your beloved NPCs. Let 'em, and let them roll with the consequences. Try to keep "game critical" characters to a minimum, and design scenarios (or even just tiny responses) to the slaying of fancy people. If Solosolo breaks Mayor McGraw's neck with his large, meaty hand, you don't have to write an emotional ten node tree in every townie's dialogue reacting to the situation. A simple set of floats based off of that condition is usually enough to make the player nod in satisfaction. "Wow, I can't believe Mayor McGraw is dead." "Please don't hurt me, Solosolo!" "Yeah, I never liked the mayor, anyway." Mission accomplished. And if that changes how the main story plays out, that's even better. People live to hear Ron Perlman's voice say things like, "Lost without Mayor McGraw to defend them, the people of Sleepy Oaks were crushed and scattered by the rebel Brotherhood Paladins!"
  
• I mean it. Seriously, this also applies for characters who join the party and power groups with whom the character can interact/join. We all know the typical "power faction" setup: the Holy Faction of Blah Blah seeks out the player because they believe he will be helpful in Task X. Too often, this becomes a simple case of one-sided manipulation. Give the player the ability to turn the tables on those factions and squeeze them if he or she wants to. Holy Faction Member A: "Solosolo, we think you should go kill the dragon!" Holy Faction Member B: "No! Solosolo, we think you should go talk to the dragon and make him our ally!" Solosolo: "HEY, TWERP, HOW ABOUT I TELL YOU WHAT I'M GOING TO DO, AND THEN YOU ALL HOP IN LINE TO MAKE IT HAPPEN?!" Holy Faction Members: *cower* If two NPC party members start complaining to each other and getting into arguments, let the player step in a settle it the way he or she wants to (within the capabilities of their character).
  
• Always show the player when the choices they have made are tracked and used. If the player gets an option in dialogue that's the result of a high Speech skill, highlight that explicitly after the reply text. If the character earns a reputation as an arrogant bastard, show that on their character sheet. REPUTATION: ARROGANT BASTARD -- THE PEOPLE OF THE LAND KNOW THAT YOU THINK YOU'RE HOT STUFF, BUT THEY'RE UNIMPORTANT LITTLE JACKASSES, SO WHO CARES? When an NPC calls the character on it, let the ramifications of Arrogant Bastard-hood shine clearly across the land.

These things have nothing to do with the underlying game system or even with the setting of the game. You could write a CRPG with GURPS set in Isaac Asimov's Foundation Series and use all of these basic ideas. If the player can't see significant changes to the world based on his or her actions, it doesn't matter whether they blow something up, talk to it, or sneak past it. That variety and choice only matter if they lead to a satisfying conclusion. While the acts can be very satisfying themselves, most RPG players (even casual players) take even greater satisfaction in seeing their actions reflected on a larger scale. A good RPG doesn't need to be simplified OR complex. It just needs to support the player's ability to do what he or she wants within the greater, open story. It should be that freedom which directs the overall design of system implementation and setting presentation.