ast week, I gave an update on a few elements of the game. The classes featured prominently since they are perhaps the centerpiece of the game. This week, I'd like to talk a little bit more about how we're approaching class design.
Our approach can be broken down into a few key concepts we use to guide things.
Use What We Have: In some ways, designing classes is pretty easy. Let's take the bard as an example. The bard has been designed four times before, so I have a lot of good work to start with. My first step is to review prior versions of the bard and take note of what has remained consistent. A few things stand out:
- The bard has always had access to spells.
- The bard has always used music in some form.
- The bard has always had some sort of ability to recall lore.
- The bard has always had access to pretty good weapons and armor.
- The bard is pretty good with skills.
Those might all seem like obvious starting points, but they're a helpful reminder of what people expect. When we design a class, it's important for conversions to stay on target with what a class means in D&D.
In some places, this task can bring up some interesting questions. For instance, is the paladin's ability to smite key to the class? What about the druid's animal companion? That's an ability that rested in the spell animal friendship. Is it absolutely critical to the class's identity? The next step speaks to this issue.
Bridge Gaps as Needed: For some classes, we can make a list and implement what's on it. For other classes, we need to figure out if everything on the list is important, or if other parts of the system can handle it. For instance, a ranger might lack any specific mechanics for archery or two-weapon fighting. Instead, a player can choose to take the appropriate specialty or venture into a different one. The feat system and the ranger's standard attack improvement shoulder the load here.
In some cases, this step affords us the chance to try something new. Take the bard as an example. Music and performance are always highlighted as keys to the class, but the bardic music ability lags behind spells. We could look at combining the two, treating the bard's special music abilities as the class's key expression of magic. Mechanically, the warlock's invocation mechanics might be a better match than spell slots, with a bard infusing magic into songs when appropriate to control emotions, inspire heroics, and so forth.
For something such as animal companions, we might take a step back and look at how they fit into the game as a whole. Giving a class a companion is tricky. Some players like the class but don't want the complexity. Balancing it against other characters is difficult. If the companion is too weak, why bother? If it's too strong, have we given one player two characters' worth of power?
This might be a case where we sidestep the issue. For instance, our rules module for followers could include an option for rangers and druids to take on beasts, werebears, and fey creatures as companions. A cleric or paladin might enlist the aid of an archon, a wizard could bind a lesser demon or elemental, and so forth. Such companions could gain XP and level up along preprogrammed paths, making them equivalent to characters. By making those rules a module, we create an option for groups that have few players (everyone takes a follower) or that let a druid player have a fully featured companion while giving the DM clear guidelines on how to balance that (the DM simply treats the companion as an extra PC). The DM or the group consensus determines how best to handle the issue. Meanwhile, we can create classes without worrying about companions (animal or otherwise) as a default option.
Keep Options Open: The final step involves looking at our various campaign settings and how the classes fit within them. How do settings typically change classes? What exists within the class that might be specific to a culture or world?
Wizard traditions are an excellent example of where this approach leads. We arrived there by thinking about how the Forgotten Realms' bladesinger, the orders of high sorcery in Krynn, and Dark Sun's preservers and defilers could all fit into one class. Traditions gave us a way to swap out class features based on a wizard's arcane order. A ranger's favored enemy, the specific music a bard masters, and so on help us ensure that we leave room in each class to customize it to a setting. The good news is that this approach also gives DMs the tools to shape classes to fit their worlds.
Experiment: Finally, it's worth noting that having such an enormous, energized group of playtesters means that we can push things ahead a little more aggressively than we otherwise could. Does bardic music as a supernatural ability replace spells? We can design a prototype, and you can read it, play it, and let us know. Without the playtest we'd have to tread more carefully.
Speaking of the playtest, it looks like we should have another playtest packet out by the end of the month.
Mike Mearls is the senior manager for the D&D research and design team. He has worked on the Ravenloft board game along with a number of supplements for the D&D RPG.