Monday, June 20, 2011

Realm of the Mad God Released!

A little over a year ago, Alex Carobus and Rob Shillingsburg from WildShadow created a prototype of a re-imagined MMO for the TIGSource Assemblee game jam. David and I started working with them a few months later through Spry Fox, helping flesh out with the monetization, interface and cooperative mechanics. The game has been in public beta for a year, rigorously tested by thousands of passionate players. The metrics look good. The tech is in place. Today, after a huge amount of work by a truly talented team, we are officially launching on the Chrome Web Store.

As with any online game, this is just the start. On the feature side, expect to see the game grow and evolve substantially in the coming months. On the distribution side, we'll also be slowly be rolling the game out to thousands of portal sites. Limber up your shooting finger and give Realm a go. Don't forget to dodge.

Indies Innovate

Realm of the Mad God is unique. If you were to toss various gaming labels at it, you'd get the following:
  • Fantasy MMORPG
  • Bullet Hell Shooter (so many bullets!)
  • Co-op only
  • Real-time combat (with 80+ players on screen)
  • Permadeath
As added spice, it runs in a browser using Flash, has a scalable cloud-based backend, sports user generated art and is rather effectively funded by a free-to-play business model.

It turns out that only an indie has the ability to reinvent the MMO genre from the ground up. Certainly no large studio could have made Realm. The original idea contained risky design, challenging hardcore gameplay, hitherto unseen technology. Heck, who even thinks of greenlighting a modern game with permadeath? Yet because of exactly these moonshot design constraints, Realm is fresher than 90% of the games that ever get released. Core gamers who love great games will love Realm.

Realm falls into the emerging category I think of as "Top Shelf Indie": highly playable games that offer deep, polished experiences that are just as much fun as more bloated titles but come with the distinctive spin that only a smaller independent team can manage. Pulling such a feat off requires three ingredients that are rare in our industry:
  • A highly talented small team
  • Freedom to pursue an innovative unified design vision
  • The ability to rapidly release and iterate on playable builds.
Simple stuff. If you love making games, you should look at your current project and ask yourself how many of these ingredients you have in place. (Yes, you deserve to work on a team with all of these.)


Here are a couple of the big design challenges:
  • Co-op is a pervasive design philosophy, not a tacked on feature.
  • Rival goods are a form of PvP

Co-op is a pervasive design philosophy, not a tacked on feature.

The rule of thumb for everything is "Does this mechanic make the game better to play together?" Here are a few of the systems in place the encourage cooperative play.
  • Experience is shared. All you need to do in order to earn experience is to be close by when an enemy is killed. There is a simple yet effective benefit from playing with others. More players means enemies die more quickly, so players working together level and earn our time-based currency (fame) more quickly.
  • Teleporting is cheap. Players can quickly and easily teleport to other players. In early builds, most players would wander for hours without seeing another human being. A smaller percentage of players liked this, but most simply assumed the game was single player and then left. We also wanted to avoid requiring a complex guild and grouping system, especially for new players. As it stands now, it is difficult to play for long without another player jumping to your position and playing alongside you.
  • All experience levels plays together: Level 1 players can easily play alongside Level 20 players. They increase their chance of dying, but there are few artificial barriers if you want to group together with friends of a different level. It also helps that the leveling curves are more shallow than is typical. he most powerful player is only about 30 times more powerful than a first level player.
  • Twinking is encouraged since loot is plentiful, inventory is limited and there isn't a direct means of cashing in items for a more liquid currency. As such there is a culture of players sharing stuff they've found with other players.
Crowds form organically and you'd find many small bands merrily adventuring through the world.

Yet even with all these core systems in place co-op can still be quite fragile. One incredibly cool (and powerful) character is the rogue who has the ability to go invisible and rush in for a sneak attack. Yet, since this class can still be hit by stray bullets, the rogue prefers to work alone where he have better control over exactly how enemy attacks are triggered. As a result there is now a large population of solo-focused players that often complain bitterly when anyone from the larger community attempts to play cooperatively. Other players resent the aptly named rogue for playing selfishly and stealing loot they would have otherwise had a chance to pick up. In essence, an individually fun class actively pollutes the core intent of the game.

So lesson I've learned is that co-op works best when every system in the game is tuned to encourage inexpensive and easy cooperation. This can mean tossing many traditional concepts like expensive travel or requiring guilds for groups. It may mean nerfing or changing a fun and delightful mechanic if it somehow damages the community as a whole. In multiplayer games, finding the fun isn't enough. You need to maximize the fun without poisoning the experience of others.

Rival goods are a form of PvP

An early decision made for the first prototype haunted the game for far too long. Loot in Realm is a rival good; to paraphrase Raph Koster, "If I grab it, you can't." While the rest of the game is all about mutual benefit, loot as a rival good brought out the bloodlust in our players on a number of levels.
  • Initial looting: A team of players will cooperate beautiful to bring down a large boss. They'll coordinate, send happy messages and joke with one another. Yet as soon as someone snags loot that another person desired, the conversation turns acrimonious. Inevitably one player grabs far more loot than others and the predominate emotions of the game decay into greed, distrust and a deeply felt lack of fairness.
  • Death: When someone dies, a percentage of their items stay on the body. This yields a huge incentive to steal from your fellow players. I'd love to run some experiments where all items on a body go away or are soulbound and see how that changes the mood of the community.
What is tricky about both of these is that both looting and stealing from corpses is surprisingly fun. Players treat it as a form of survival of the fittest PvP and a mercenary minority have perfected their skills as efficient looters. Yet as a whole, these 'fun' moments for individuals create a dysfunctional society.

We are methodically solving for these issues by awarding players with soulbound loot. This removes much of the competitive nature of picking up loot and turns loot into a non-rival good. Showing up and helping out gets you loot, not stealing from others.

Other lessons

The game does some other things worth noting that are common to other games I've mentioned here. These help ensure that a small team can release a game that traditionally requires the efforts of hundreds:
  • Design from the root: Instead of accepting genre conventions, each new feature was vigorously questioned to see if it fit the core concept of the game as a co-op MMO shooter.
  • Procedural generation: All the maps and dungeons are procedurally generated. This means content is quite cheap.
  • User generated content: The initial graphics used in the game came from the wonderful Oryx set. However, Alex invested a good amount of time in building a full featured sprite editor that allows users to make new monsters, items and animations. Now almost all new visuals end up being sourced from the community.


MMOs are intensely complex games where even simple systems blossom into a thousand layers of culture and community. The rules of the game create explosive economies, ecosystems and power structures that deeply intertwine with the lives of players. I'm really not surprised that so many MMOs are essentially clones. The genre canon is a tightly wound mechanism where even small changes can destroy your game. Every MMO team faces this particular minefield and must ask themselves if they have the guts to mess with standards that have been gingerly polished over decades.

Most back away from the challenge with fear in their eyes. Instead of making something new, they take the coward's path and desperately try to differentiate their me-too creation with pointless cut-scenes, laborious writing and gaudy graphics.

Alex and Rob took the challenge. Yes, rebuilding many of these systems from a unique starting place is an epic undertaking. But it worked. We need more teams with the guts and the ability to reinvent genre conventions. We need more games like Realm of the Mad God.

take care,