Rebooting the RPG


In the Beginning

Many indie game developers start with RPG’s, and thirteen years ago, I was no exception. With hundreds of hours of classic RPG playing experience and RPG Toolkit 2.0 freshly installed on my Compaq Presario, I sat down to create my answer to Final Fantasy.

Months later, after I’d made a game that not even my mom pretended to like, I went online to ask experienced developers for advice. I'll never forget what they said:

"Forget RPG's and start with PONG."

It’s like Joel Spolsky says in his article Back to Basics:

If you want to teach somebody something well, you have to start at the very lowest level. It's like Karate Kid. Wax On, Wax Off. Wax On, Wax Off. Do that for three weeks. Then Knocking The Other Kid's Head off is easy.
So I did just that - I worked on simple games, and slowly improved my skills over the years.

Ever since then, I've steered clear of RPG's, my first love, and never looked back. Fast forward to today, and I've somehow realized my dream of making games for a living. Whereas I once hoped to work for Square or Nintendo, I now run my own company, and instead of working on RPG's for home consoles, I develop strategy and educational games for the web and PC.

All these years later, I have a confession. I’m working on an RPG. But this time, with everything I've learned, I'm going to do it right.

The Appeal of RPG’s

When I was six years old, video games were difficult. Action games in particular were nearly impossible. They required quick reflexes, and if you reached a stage that was too difficult, you could easily get stuck and give up. RPG’s offered time to think, as well as a safety valve. If the game was too hard, you could grind for a while until you reached a higher level. In those days, beating an RPG was a mental exercise, making it the thinking man’s game. RPG’s were also about the only games that offered stories. Other games often had a light premise, but if you wanted an interesting narrative, the logical choice was an RPG.

Anatomy of an RPG

RPG’s typically feature several systems, including:

· Overworld · Towns · A battle system · Inventory/equipment system · Experience/level-up system · Character classes · Dungeons · Boss fights · Mini-games · Cutscenes

Each of these was originally created for a specific purpose, but over the years, they’ve become standard features simply because “that’s what you do in an RPG.” With all these elements competing for attention, it’s easy to lose focus—a significant risk for any developer, and indie developers in particular.

Focusing on the Battle System

The core of any RPG is the battle system. That’s where players spend most of their time, and if it isn’t awesome, it weighs down the entire experience. Even so, many RPG battle systems use a traditional formula—simply copying what came before. Take, for example, many JRPG battle systems. They’ve been around for decades, but most of them still look a lot like this:

Just look at Persona 4 (2008):

The standard turn-based JRPG battle system was designed to deal with graphical and memory limitations that don’t exist on modern machines. Furthermore, this system is pretty shallow (for further reading on this, try this great article by Indie RPG designer Zeboyd ).

The basic strategy flowchart for most RPG battles looks like this:

Developers have added plenty to this format, including magic systems, special attacks, elemental weaknesses, etc, but none of this can compensate for the system’s main limitation—its complete lack of geography. Board Position is one of the key elements of almost all games, including everything from checkers to Civilization.

Physical space, then, is the chief innovation of tactical RPG’s. Rather than lining up on either side of an abstract battle zone, characters in a tactical RPG are deployed and arranged in a tiled environment, similar to board games—tabletop war games in particular.

Tactical RPG's add depth...
Due to the increased complexity of these battles, tactical RPG’s feature fewer battle sequences than JRPG’s, as well as more characters.

This introduces a problem of scale. It’s exponentially more difficult to manage fifty characters than it is five. When tactical RPG’s borrow too many elements from JRPG’s, players can end up spending more time tracking inventory than fighting battles.

...but also lots of potentially tedious detailDistracting Systems RPG's advertise 60+ hours of epic gameplay, but only a fraction of that is spent on doing interesting things. If the core of an RPG is the battle system, then the other systems should compliment it, adding richness and depth.

Here's a rough estimate of how you spend your time in RPG's:

Let's break these down one by one:

  1. Overworld Games like Zelda do a good job with the overworld, but in most RPG's it's usually just empty space to march across. Even better, it's full of random, time-sucking battles.
  2. Towns Towns are often just bad menus. I just want to buy potions, but instead of clicking "buy potions" I walk to town, walk to the potion store, talk to the potion clerk, etc. I spend more time physically moving between errands than on the errands themselves. Towns can add realism, but only if they have some purpose other than selling equipment to adventurers. If that's all they do, then give me a way to buy my groceries in less than an hour.
  3. Inventory screen Inventory screens are a great place for interesting decisions - how do I best outfit my group? Unfortunately, RPG designers tend to cram in so many variables that choice paralysis sets in as you try to decide whether a "+15 Ancient Blue Halberd" is better than a "+10 Exquisite Rapier of Greater Frostbite."
  4. Random battles RPG's insist on interrupting you with hundreds of random battles. Often, you don't get your health back after a battle, which ensures you'll face the boss in an anemic state. Even if you get strong enough to survive them all, you're still being constantly interrupted.

    Rarely does the player have to make an interesting decision here.

  5. Story battles These are the interesting battles that actually have something to do with the story! These would be awesome, except for one of two things. You either used up all of your resources slogging through random battles, OR you leveled up enough to get here unscathed, and now you're so strong the boss isn't even fun to fight anymore.
Surely, we can do better than this. Let's make RPG's fun again.

Introducing Defender's Quest

We're making a new game. It's called Defender's Quest.

We decided to make a streamlined RPG where the battle system is the core focus. We chose Tower Defense (TD) as our base mechanic because of its depth and interesting strategic choices. Since we're an indie team, we can't afford to waste time on extraneous systems - so everything else in the game was designed to compliment the battle system as efficiently as possible.

In effect, we're taking RPG's back to first principles. Here’s a quick screenshot of the battle system:

Typical Defender's Quest battle screen Let's break it down. The lady in the middle of the screen is the player character and the game's McGuffin: to win you have to keep the monsters from reaching and killing her.Protect the McGuffin! The other characters are defenders. The McGuffin can summon and place them in the battle one by one. Essentially, your party members are like towers.

Like most TD games, your defenders will automatically attack the enemy, and killing enemies will recharge the McGuffin's PSI energy, which you can use to summon more defenders or cast spells.

Kill the bad guys!Six different defender classes fulfill roles such as ranged, melee, support, etc. Some TD games include light RPG elements, such as the ability to upgrade an entire class of towers, or research new spells, etc. Defender Quest, however, goes far beyond that.

Whereas in most TD games all towers of the same class are identical, in this game each defender of a certain class is a unique party member. That means that each defender can be customized individually as they level up to suit your needs. Archers, for instance, can specialize in doing more damage, hitting more targets, or inflicting status effects like poison. Since you'll have several defenders of a given type, you can tweak them in different ways to deal with different strategic situations.

This gives the player a lot of meaningful choices both in the short term and the long term.

For instance, in the short term, you can either spend PSI to summon more defenders, or use that energy to boost (upgrade) existing defenders. Summoning extra defenders puts more damage on the board quickly, and also increases the amount of space you can cover on the board. Boosting existing defenders unlocks their higher-level attacks and increases their stats.

The McGuffin boosts a defender In the long term, you face a dilemma each time a defender levels up. Each level confers one skill point to put anywhere in the skill tree - but where?

Each defender has five technique or "tech" skills, which will be used in battle. The first tech is the weakest, but is available immediately after the defender is placed. The last tech is the strongest, but isn't available in battle until the defender is boosted four times. Concentrating on low-level techs makes defenders more useful right away, but the splashier high-level techs are essential late in a battle. Furthermore, defenders can spend skill points on trait skills, which are "always on" abilities, such as adding poison to every hit.

There's much more to the game than just this, of course. We've got several other systems that drive the meta-game, all designed with the idea of keeping the momentum going. And of course, we have a cool story. Rather than subjecting you to my own aborted attempts at literature, however, we found a guy who actually majored in English to write our gripping tale of semi-epic, nearly-heroic proportions! Thrilling, I know.

We'll go into more depth in future posts as we talk about how our game addresses some of the core problems of RPG's, while building on the appeal that makes this genre interesting in the first place.

See you there!

-Lars out

This is a companion discussion topic for the original entry at