Jeff Freeman apparently has time for random blogging again (sheesh, Jeff, you don’t see expansion crunch slowing me or Scott down, do ya?), and has gotten around to finally updating his blog. He’s responded to a post I had forgotten I made in Corpnews (motto: Fuck mottos). The discussion is systems vs content. Here’s the money quote from my post:

<] Inside the industry, we separate out what you guys are calling ‘content’ into two categories: systems and content. ‘Systems’ are new code and/or game functionality that create large subsets of new behavior: crafting, gambling, and minigames all fall into systems.’Content’ refers mostly to stuff that is art and data. Anytime you say ‘more’ of something, you’re generally talking about content – more monsters, more levels, more areas, more skills, more spells.

Of course, the two categories interact quite a bit. Coding a quest would be content, but coding a quest engine would be a major system. Major systems frequently require being ‘filled’ by content, although as developers, we love systems that provide great replayability without requiring a ton of content (such as the Resource System we just did in Shadowbane).

Jeff responds with a description of what they learned developing SWG and how they applied it to JTL. I found Jeff’s post interesting, as I agreed with most of what he was saying, even though I’ve long been an advocate of more and better systems, while Jeff describes himself as an advocate of what he calls ‘Content Driven Development’. A snippet from his post:

So SWG shipped without a lot of content…For example, there was a Smuggler profession, but there was not actually any smuggling…For another example, we built 10 16×16 kilometer planets, and then sought to “fill them with content”. They weren’t what you’d call “packed” with interesting, fun things to see and do.

We took a different approach with JtL, and that was to essentially let content drive the process. We started with the ability to add 14 new zones. So we said, “We only want to add ten. Rori is Naboo’s moon and Talus is in the same system as Corellia, so they can share. If we come up with more content and no where to put it, then we’ll start adding additional zones”. Additionally, we started with 4×4×4 kilometer zones. We didn’t increase the size of them until we needed to (and we did need to, twice).

We applied the same philosophy to the skill systems themselves. Rather than adding a pilot profession and then thinking of different things for pilots to do (or worse, not thinking of things for pilots to do), we added content.

I agree with most of his sentiments, but not all. I was trying to put a finger on what, exactly, I thought about all of this. I came up with what I call the 10 commandments of content vs systems, a name designed to think I put more thought into the topic than I actually did. My thoughts.

1) Experience is king. This is really what I read Jeff as saying, and as such, I vote to change what he calls ‘content-driven development’ to ‘experience-driven development’. What does it FEEL like to play as a Thief in Shadowbane or a Smuggler in Star Wars? Does it satisfy your expectations? What systems are needed to make it feel right? And what content? How can you deepen the feeling that you are occupying this role? What sacrifices do you have to make to allow a thief (a purely antisocial class) in an MMO, and is it still worth doing if you cut too deeply?

2) Less good content is better than a dizzying array of nearly identical choices. You can always add more races and more classes in the future. Put in a limited amount of content, and hit a home run with it. Once you nail the first class, building the rest will be easy. If you try to build all the classes at once, you will fail to take a holistic look of the game’s development (as Walter Kim alludes to here, among other things). Jeff himself says “[you want to] ensure that you don’t have more canvas than paint.” What he says about the space zone size in JTL is instructive – make a small, tight area of content, and once that’s nailed, if you still need more, spiralling out more will be trivially easy. On the other hand, if you try to create or balance all of your classes in one pass, it will be hard to ensure that any of them are balanced correctly.

3) Iterate your systems and content together. As game projects have become larger and more complex, game producers and schedulemonkeys have been trying to fit game design tasks into smaller, more manageable bits. “We’ll finish combat at the end of July, and the designers will fill in the data in August. Done.” Wrong. Fun takes iterations, which means that both the systems and the content will need revisiting, massaging and tweaking. Plan to iterate. Cut other content and systems if you need to nail the experience of the ones that are really important. Understand what your core is, and give those systems more love.

4) Do the right systems first. Not to pick on SWG too much, but most players logged in to become Han Solo and Luke Skywalker, yet both Jedi and Smuggler were unplayable at ship, in favor of such professions as weaponsmiths and bio-engineers as well as a matchmaking service which compared your blood type to random walkers-by. In the interest of fairness, I’ll point out that Shadowbane shipped with many features it didn’t necessarily need, but with a ‘not quite ready for prime time’ Siege system. We’ve made great strides with it since, but if Shadowbane really is “The Siege Game”, shouldn’t we have done that first and made SURE we got it right? Damn straight we should have.

5) If you can’t complete it, don’t expose it. Incomplete content simply creates an appetite that will never go away. The original SWG launch was not troubled by having too many systems, but rather, by displaying non-functionaly skill boxes to the fans for some classes – and as such, had too FEW systems. If the fans know something was SUPPOSED to be there, they’ll always complain, even as you try to deal with other issues, and often whatever you implement will not live up to their own expectations. On the flip side, Meridian 59 (a fantasy MMORPG) never had any sort of thievery skills or rogue classes. But we never announced plans for it, never put it on the web site, and never put a greyed out button saying ‘coming soon’, and got little complaint.

6) If it doesn’t create a new, fresh and interesting experience, it’s not really new content. This is a simple rule (see #1: “Experience is King”) but is also where I challenge what Jeff wrote the most. Quoting:

Anyway, content driving the systems would work like this:

* You add a guild of warriors called the Thundering Pounders.
* Fictionally, you define that warriors guild: What is it? What do people who join it do? What do they wear? What is their fighting-style? Who are their enemies? And so on.
* Now you need a warrior class, so you add that.
* Add another warriors guild, but this one’s called the Darknight Assassins.
* Fictionally define that warriors guild as a completely different sort of thing than the first one.
* You already have a warrior class, so just use that. Ah, but these warriors are different from those warriors, so maybe the class system that you’ve added needs to be tweaked so as to accommodate the distinction between these different kinds of warriors. Maybe these guys use poison and the other guys just hit harder. Whatever.
* You don’t add a thief class unless/until the content demands it.

If I’m playing a fantasy RPG, and there are three kinds of warriors and no thief-y option, I’ll quit (and I know that directly contradicts what I said about Meridian 59. Shush).

Players are very savvy, and quickly realize that a Dark Paladin is a Warrior with a black coat of paint. While cheap wins via data-driven design are always nifty, the actual experience of playing one vs the other has to really FEEL different and provide a different sort of experience to really succeed (see “#1: Experience is king”). Often, this will require adding or expanding a system. Players are also quick to note that every mission in City of Heroes has the same Coke machine in the lobby, and that randomly generated terrain creates hills that all look kinda the same (in fact, i very much think that EQ’s hand-crafted zones were a large part of its dominance over Asheron’s Call’s endless expanses of rolling hills).

In the Shadowbane Expansion Pack, we added a race and two new classes. Our stated goal was to make three player choices that felt like nothing that already existed in Shadowbane. The Vampire’s powers all cost blood and he’s got new ‘fortitude’ abilities that block damage in unique ways. The Necromancer is the first pet caster to be able to summon swarms of pets. The Nightstalker is… okay, it’s an Assassin with a fresh coat of paint, but the paint makes the Nightstalker the master of fast attacks, something the Assassin was never good at. We came to these classes via ‘experience-driven design’- i.e. how could we make the experience of playing a vampire and necromancer cool and starkly different than the other play experiences that exist in Shadowbane? We determined that the answer was, in this case, a couple of new systems to buttress what we could already do in data (i.e. with content).

In retrospect, perhaps Jeff and I agree exactly (quoting again):

It just so happens that Storm Squadron is an Imperial Navy unit, and the end result is that one does, in fact, join the Imperial Navy and become an Imperial Navy Pilot. The difference is that there are things to see and do as an Imperial Navy Pilot, because the things to see and do – the content – drove the development. The Imperial Navy Pilot profession was added because the Storm Squadron content needed it, not the other way around.

If you replace ‘the content’ with ‘the experience’, he’s right on.

7) Pure, mass amounts of content has the ability to be soulless. If you have 500 food items that all do one of three things, it actually makes the world less interesting than having 10 food items that have wildly different effects. Similarly, if you have a 10 mile by 10 mile planet and can only drop enough interesting Points of Interest to fill two city blocks, players will remember the emptiness more than the content. You’d be better off with a planet that was four city blocks large.

A wise man once said that the best Interface design was the one where you never noticed the interface – the logic being that if you noticed the interface, it was non-intuitive and therefore failing. The same can be said for content – if your players notice that your content is repetitive, or huge swaths of your game are barren of it, they will hold those failures against the game design as strongly than your successes.

8) It doesn’t take a lot of systems, or very complicated systems, to create big effects. Systems, as code, interact in lots of interesting and difficult to balance. The best computer game of all time, Civilization, has a very low number of systems. However, those systems interact in interesting ways that create huge depth and strategic possibilities.

9) Content is consumable, systems less so. Most people don’t like to play the same content over and over again. Content usually carries a sense of discovery of the content – interesting characters and plot twists in fiction – but going through it the second time is always less compelling than the first. Think, for example, of how often you could really stand to play a game of Myst.

Systems, on the other hand, lead to a much more repeatable experience. Civilization 2 is low on content and high on systems, and the end result is a game much more repeatable than Myst. Given our business model depends on players playing for a long time, having a game that’s engaging every time you play it is a goal worth striving for.

The problem with both is that they can run together. The City of Heroes gang, for example, was very proud that they produced 400 hours of content. I probably played 20 hours of it before it all ran together and felt the same. I was craving some gameplay experience that was less controlled and less on railed – systems that created more unpredictable gameplay experiences.

10) The best systems turn people into the content. Shadowbane’s Resource system is a good system in that it needed very little design content- it’s an extremely simple system which gives people something to fight over, and makes it easy for people to find and get to the fight. There, the players – and all of their skills, their strategies, and their alliances– become the interesting content. And since player skill, strategy and alliances change every day, it means the player experience changes every day.

The most innovative system in Star Wars is the mental fatigue/entertainer system, simply because it gets people to talk and/or cyber with people they otherwise wouldn’t do so with. Little content was needed to make it work, because people (and all of the strange and eccentric things they do) ARE the content.

For all of City of Heroes’ innovations, it’s most interesting feature was its character creation feature. The most fun to be had in the game is standing around town square looking at the fruits of other player’s creativity.

Look, at the core level, people play MMOs because other people are there. If they liked systems OR content more than other people, they’d all be playing KOTOR or Baldur’s Gate. Even the players that prefer to play MMOs alone are making a decision to play there instead of playing Morrowind. Why? Because the mere existence of other people walking by adds enough spice and interest to make up for lag and bugs that too often afflict our titles.

So ‘other people’ is our unique selling point. But we forget that too often. The future of MMO design is, fundamentally, not continuing to try to mimic the content and systems of single player gaming, but rather doing more to emphasize the ‘massive’ in Massively Multiplayer gaming.