There's been a lot of talk lately about the role flavor should play in design and how much weight designers should give to how something subjectively "feels" if there are game mechanic or balance concerns involved. I wanted to give the community a sense of just how complex and multilayered this process is. It isn't as easy as just "find a way to make sure the flavor wins--create a new ability if you have to." So here are some of the things we go through in designing pieces. This is after we've already chosen a set list, often 9 months or more before the set comes out.

1. Research--especially if we've decided to make a character that isn't familiar to us, a lot of time on wookieepedia and other sites trying to get a sense of the piece and what it might do in a game, what abilities might be appropriate, even whether or not it should be competitive. This might also include looking at customs of the piece on Bloomilk made by other players.

2. Initial stats--someone posts a stat block including all the information that will eventually end up on the card. Ideally, they also include text explaining what the piece is, where they imagine it fitting in the game, and what power level it should be aiming for.

3. Initial discussion--the rest of the design team chimes in with any thoughts or ideas. Steps 2 and 3 are where ideas for new abilities typically come about, but not every piece needs new abilities. In fact, the glossary is overflowing with abilities. So, many designers are cautious about creating something new when an existing ability or combination could have the same effect. There's also some reluctance to make a "one-off" ability that will only ever appear on one character. As much fun as it is to tailor a stat card, for the most part, having abilities that appear only once greatly increases the complexity of the game without much functional payoff.

4. Detailed discussion. So let's say we decided to put a new ability or mechanic on this piece. How does it work? How do we get it on the card in the most natural way? Say we want to link this character to another one we're creating. There are a few options for that--Camaraderie, synergy, CEs. Camaraderie has been somewhat overused in the design process, so it might be something we want to avoid unless there's a particular reason to use it. Synergy depends on range and usually provides only a small boost. A CE might be okay, but what if you want this character to be a follower? You might need to add Willing to Serve. Do you want it to be disruptible? There are ways around disruptive, but suddenly you have a stat card that's all about making a CE function instead of making a character come to life. Or maybe you want a character to do something entirely new, but the rules gurus tell you "That's not how the Force works." So you have to figure out another way.

5. Interactions. This is when you think about, say, whether the piece you're making should be able to get Twin Attack, or what other boosts might be available to it. (Please note that 3-5 are taking place more or less simultaneously and throughout the design process). What squads are people most likely to build? Where's the damage max on this guy? Is that okay for his cost and faction? How about his defense, hit points, attack value? What do we want him to be able to do? This is the place where flavor and function might intersect the most. We're thinking about which pieces we want the new character to work well with, which can be in large part about flavor. We're also thinking about the function of the piece, where it fits in the competitive game as a whole. We're comparing the piece to similar pieces in the same price range to see if the power level feels right. We might also be designing in tandem with other pieces, trying to get the balance of flavor and function as right as it can be. This process can involve a lot of compromise as each designer figures out where they stand on a particular piece or interaction.

6. Playtesting. Our creations are sent out to the playtesters, who come back with reports that tell us not only how the piece worked, but whether it felt like the character we were trying to represent. Playtesters might tell us that a new ability is unclear or that a different ability would feel more like the character. We take these comments, including any about power level, into consideration and...

7. Redesign. After we have some playtests on a piece, we return to the design thread to discuss what we learned. This is the time stats get raised or lowered, and we consider ideas for additional abilities or interactions. We talk about what the playtesters said and whether it seemed like the piece functioned how we wanted it to. Sometimes this discussion can get extremely detailed, as there can be many ideas about how to achieve the desired flavor while not sacrificing making a useful piece. If changes are made, the piece returns to playtesting.

8. Finalizing design and QC. Once the team has enough playtest reports and the timeline is winding down, it's time to start thinking about finishing pieces. If QC hasn't been involved until this point in the process, this is when they come in to make sure everything is in order. They may also make suggestions to insure that abilities work the way they're intended to. The stats are inputted into card templates so we can see the finished product. At this point, if it hasn't already come up in the design process, the designers might realize that there's just too much text on a card and have to either leave some abilities undefined or choose to drop something. More compromise to practical concerns. Once the cards are finalized, the PDF is sent to the printer and the waiting and teasing begin.

My point in all of this is just to say that there is a LOT of work that goes into the design of a set, a lot of practical concerns we have to weigh when trying to get flavor right. It involves constant analysis, knowing when to stick to your guns and when to compromise, and weighing the opinions of lots of other people who may or may not agree with you. It is an incredibly complex process, and with that complexity comes room for error. But it also breeds some truly great ideas and pieces. When everything comes together and a character you had a hand in really works and feels right...it's one of the best feelings in the world.

Flavor accuracy is another reason we need more people to become active play testers. As we keep moving forward and have continued to increase the quality of the play test process, man power is what we need now more than ever.

How you could possibly implement an AI to help with this process is lost on me.

It honestly wouldnt be too difficult to program a "player" to play the game piece in Vassal. It would just take some time, and effort to get it right for each other piece in game and vs each piece it may play. But in time, it could be done.

Once done you could let it run in multiple games vs various squads. Then read the a report that could be generated detailing what each piece did, attacks/abilities used and the results. (much like a report of a chess match but, obviously much more detailed).

Also, great post Laura, as always!

Very true, however if both AI are playing at the same level, at least it would be a bit more ballanced as far as we can review things on a bit more equal terms. :)

Cool post. It is interesting to see what the process looks like (on paper at least).

As for flavor, has anyone researched card manufacturers who offer scratch-n-sniff technology? 😉

+1

And Graham is an Ace at that