|Trials, Triumphs & Trivialities #62:
Galactic Empires, Part One: Failing at Succession
February 28, 2002 - On April 2, 2001 Skotos Tech released Galactic Emperor: Succession, our first science-fiction game, and also what we hyped as a hybrid roleplaying-strategy game. The first incarnation of the game lasted less than two months before we shelved it, reworked it, and re-released it on July 2, 2001. Then, on February 11, 2002 we shut down Galactic Emperor: Succession for a second time.
Clearly, something wasn't working right here.
As game designers, we can learn as much from failures as successes, and that's what I'd like to look at this time around. How Galactic Emperor: Succession didn't work, why it didn't work, and what we can learn from that moving forward.
Stage One: Design & Development
I talked a little bit about the origins of Galactic Emperor: Succession in Trials, Triumphs & Trivialities #15, Welcome to the Twenty-First Century. As I noted there, Succession had its basis in a LARP named "The Galactic Emperor is Dead", created by talented game designer Mike Young.
A LARP, for those of you unfamiliar with the form, is a Live-Action Roleplaying Game. Like any other roleplaying game out there, from the tabletop variety to online venues like Castle Marrach, you play a character in a LARP. The main difference is that in LARPs you tend to dress up and roleplay live with a group of 10-80 other people over a short period of time (ranging from an evening to a weekend). As you might expect given that basis, the roleplaying is very intense.
In a LARP, each character has a group of very specific goals that he's trying to accomplish. It's what drives the game forward. The goals are usually carefully crafted for each individual character, a very time-consuming process. A LARP called Home of the Bold that I played in nine years ago now had goals such as:
As I recall now, many years later, I did manage to incite a rebellion, but my tribe was destroyed in the process. You win some, you lose some.
To make Succession a viable game we started putting it together as an online LARP (or a Stage) where everyone had the same goal: become the emperor. It seemed like a sound plan, and made it easier to allow for a large number of characters, because we didn't have to do all that tough individual goal tuning.
But as we worked on the game our dreams grew. We decided we could do more than a one-time LARP that showcased our systems, instead turning it into a repeating game. At the same time we continued expanding the game by developing more and more systems. There were, for example, two different psychic systems before the game got out of alpha, and in the end neither was used for a released game.
These changing and growing design goals resulted in our first mistakes on Galactic Emperor: Succession.
First, we never specified the initial design in a solid, agreed upon state. Thus it remained fluid for way too many months. We never really nailed down the central gameplay. We worked on systems that weren't important rather than making sure that core systems worked really well.
Second, we thought too big. We really should have released the game as that initial one-time LARP. Sure, it might have seemed like we'd wasted a lot of time just to host that one event, but in the process we would have learned a lot about building a new game, and we also would have proven a gameplay mechanism (or, alternatively, disproven one, which would have been just as useful). From there it would have been much easier, and much less traumatic to just drop the game if it was a failure ... or to figure out what to build upon.
We've encouraged external StoryBuilders to try a Stage first, before continuing on into a full game, and this is the reason. It lets a StoryBuilder focus his efforts and concentrate both his thoughts and his design, and also to show gameplay to real players, and see how it works out.
Stage Two: The April Beta
We continued forward, undaunted. Systems were added and removed on a weekly basis. Finally, we got near our April 2 release date, and things actually started to take a permanent shape. We had beautiful graphics to highlight the strategy parts of the game. Everything seemed to be coming together quite well.
On April 2, 2001, we released.
The first version of Galactic Emperor: Succession was very complex. There was the core goal, which stayed solid through every version of the game. You were trying to become emperor.
In order to be voted emperor, you had to be on the ballot, and in order to be on the ballot, you had to have sufficient money (the money requirement been a prestige requirement just before the release, but it amounted to the same thing). And how did you get that money? Now that was a little complex.
Your money was divided up into two parts: assets and cash holdings. The assets were based on the value of planets that you held and could be decreased if the safety ratings of the planets decreased related to attacks by the evil V'hurgh and cash disbursements and could be lost altogether if their unrest increased this rating partially related to safety, but also dependent on other factors. Your cash holdings on the other hand could be increased by levying taxes (which also increased unrest), by passing votes (which could also be used for several other factors), and by losing ships in the battle against the V'lhurgh (which could be stationed in certain kingdoms to increase their safety).
Are you with me so far?
Probably not, and that was the problem.
Within a few weeks it was obvious to us that we didn't have enough players in Succession, and those who were there really didn't understand the game.
From my description I trust it's pretty obvious that our big mistake this time around was that we made the game too complex. Too some degree that's a reflection of the earlier mistake I mentioned not trying out our core systems.
But, the complexity itself turned out to be a failing.
First, there was no easy beginner's path. People entering the game for the first time didn't have a simple way to get started one that they could have mucked with and eventually realized that it was less effective than other techniques. Instead they had to dive in and try and figure out the difference between a bunch of similar looking knobs and dials.
Another way of saying the same thing is: the game had no center. Sure, there was the gold ring at the end the emperor's seat but you arrived at it through money/prestige and there were so many ways to acquire that there was no clearly right path.
Our second big problem was that the game was opaque. Everything was done through multiple levels of indirection. If I wanted to make sure that I didn't lose a planet, I had to decrease unrest, which could be done through putting a family member there. If I wanted to increase my assets' value I had to increase their safety ratings and that could be done through votes for disbursements or by driving back the V'lhurgh invaders.
Because of these multiple levels of indirection, it was never obvious how cause (your doing something) related to effect (what you wanted). Worse, there was no reporting mechanism. I could put a family member on a planet, and perhaps see Unrest going down, but I never had a way of knowing if that was really what was happening. And, if Unrest was going up, how did I take that? Was it because my family member actually wasn't good at calming unrest? Or did some other factor like increased taxes, reduced disbursements or Vl'hurgh invasion cause the Unrest?
Without the cause and effect a player couldn't make meaningful decisions.
Interlude: The Big-Picture Problems
So, we closed down the first version of Succession and started thinking about version 2. At the same time, however, there were a whole bunch of big-picture problems that were floating around, and we were pretty much ignoring because we wanted to make the gameplay work right. None of these problems were show-stoppers, mind you, not precisely mistakes like some of the specific decisions we made but they were problems that we were facing and that we should have at least been considering.
First, we were producing a science-fiction game. Traditionally that genre isn't nearly as popular as fantasy and that problem seems to be even more pronounced among roleplaying games. There's only one tabletop roleplaying game that's done any good Traveller and it's actually gone through five major revisions and three publishers in its attempt to stay head above water. Science fiction MUDs are almost unknown. The first science-fiction MMORPG, Anarchy Online, seems to be doing mediocre.
We saw the problems with SF genre really clearly when we advertised. Castle Marrach sometimes resulted in a click through and Succession rarely did, even though I think the ads for the latter were snazzier.
Second, we were producing a hybrid strategy-roleplaying game. Who did it appeal too? Later statistics would say, not too many people.; We'd hybridized so well that both types of game ended up pale reflections of the type. We should have instead made one of the hybrid gameplay styles the center. But, even if we had, it would have been a hard sell.
Third, with both versions of the game we were faced with the fearsome critical mass problem. In short: if you don't have enough people to make the game fun people won't stick around and you'll never have enough people to make the game fun.
If we'd written a different type of game where there were good routes to success and fun gameplay that didn't require player interaction we might have done OK. People might have stuck around, done the solo gameplay, and eventually there would have been enough people to support a social community. But the game was all social, and we just never got enough people to make that work.
Castle Marrach would have had the same problem if we hadn't staffed it with a dozen people for the first few weeks and if we hadn't staged a big input of people at the very start of the game.
Anyone writing a social-based game needs to understand the problems that lack of critical mass will cause.
Stage Three: The July Beta
And finally, that brought us to the second beta version of the game, which was to be released on July 2, three months after our first attempt. I discussed the second version a tiny bit in Trials, Triumphs & Trivialities #39, Ah, Sweet Simplicity of Life!. To be brief: we simplified.
For beta v2 we decided that the core of the system was voting, so we tried to implement a more solid voting system that could truly be the heart of a game. And, to supplement that and add some complexity to the voting, we added trading too, so that players could collect certain resources, make certain types of planets, and get votes passed concerning those planet types in ways that benefited them.
I suspect the biggest problem this time around might have been the critical mass issue, which I just talked about. Since trading was now one of the two crucial systems, it was important to meet other people, and that just wasn't happening.
There were other issues too, though.
First, I think the simplicity bit us in the butt and made the game a bit boring. We'd wanted to simplify by decreasing the number of systems, and I still think that was a very sound decision. Unfortunately the systems that we produced voting and trading were simplistic in themselves. In trading there were four commodities with clear and set value. In voting votes tended to be either good or bad for each player. Period.
I don't really know how to solve this one yet. If I did, I'd go write beta v3 of Succession, and we'd be set. I do, however, know what we should have done. To be successful in a game that was partially strategic we needed to create simple systems that allowed for complex strategies. I can't explain quite how to do that, though next week I'm going to talk about a game that meets the criteria: Galactic Emperor: Hegemony.
The second big problem this time around was cheating. Trading was typically a win-win situation. There was never any uncertainty to trading, nor any drawback, and you could start trading the instant you got into the game. Which means that some players created extra accounts, trading with themselves, and came out with big benefits. Any game on the Internet has to acknowledge the fact that its players will be pseudo-anonymous individuals who may take advantage of a game by creating multiple logins. A good game tries to minimize this problem.
When it was all said and done, I don't think the second version of Succession was any less successful than the first. But it wasn't more successful either.
Eventually we had to pull the plug, so that newcomers to the Skotos site could instead gravitate to our games that worked.
A Listing: Lessons Learned
I've mentioned lots of mistakes and problems thus far, and think the lessons learned are pretty implicit. But, since the ground I've covered has been so wide I'd like to take the time to summarize them all.
In short, here's what we've learned from the whole Succession experience:
Whew. That's a darned lot for one game.
The Future of Galactic Emperor
So, where does this all leave us? To be honest, I'm not entirely sure. We're not really interested in giving Succession a third try ourselves because our own ideas are too deeply ingrained now, and we know they didn't work. But, we'd love an external StoryBuilder to come in, pick up the pieces, and give it a new shot. Anyone interested can mail me directly.
As for the universe of Galactic Emperor: we think that it's fun and evocative and worthy to be revisited. Thus, as we bring a new strategy game into our fold, we're adopting it to our Galactic Emperor background.
I'll talk about that next week when I discuss, "Building a Hegemony".