What is Systems Design? (02)

A few months ago I was interviewing with Ubisoft Toronto. Naturally, the conversation is confidential so I won’t get into specifics, but I was talking with a creative director about the “types” of designers they hire and where I fit in. This was a refreshing topic as many, even within the HR departments in the video game industry, use “designer” as sort of a catch-all phrase. I, for instance, have very little experience firsthand with level design. Level design requires vastly different skills than does systems design which is what I consider my forte. However, by their very natures level design is much more visible than systems design. What is it and why is it important?

I can give you a very technical answer: systems design is the design of mechanics that result in desirable play dynamics. But that likely means little to most reading. Perhaps an example will help illustrate.

The purpose of having money systems in RPGs is what? It’s a very simple system to dole out empowerment to players at various points in the game. Have you played a game before where you had more money than you could ever use by endgame? The desirable play dynamic is an interesting level of challenge throughout the game. The means to achieve that dynamic is a system of rewards that scales well to achieve that end.

Two games I played this past year are, I think, great illustrations of successful systems design: Civilization V and Etrian Odyssey III.

Civ V, I think, could possibly be the most perfectly crafted example of systems design ever. There are literally dozens of systems: culture, influence, military power, research, wealth, citizen growth, happiness (and so on and so on) and most interact with each other. Civ V’s designers had to create these systems to satisfy numerous dynamic requirements: players couldn’t get nukes in the 1700s, no track to winning must clearly dominate in all cases, players cannot have too much money or get stuck in vicious loops. Somehow, miraculously with dozens of special citizens, wonders, units and players of varying levels of understanding of the systems and skill all these moving parts work together and create a balance such that decisions are interesting. In Halo, you pretty much go for the rocket launcher or the energy sword because they are the strongest. The decision of weapon choice isn’t interesting. In Civ V, you have dozens of choices in a turn, many of which do not have a clear answer and are as such interesting.

What should the Reproduction Rate be?

Etrian Odyssey III is another example. The dynamic aimed for is to clearly create a brutally difficult classic dungeon crawl. It would be very easy to make this game too hard to be beat or too easy to provide the challenge (the side most RPGs err on, see the new Golden Sun reviews). Players can be a dozen classes each with vastly differing skill trees – which will the player choose? How much experience with these systems does the player have? How much time will he spend grinding before he gives up? Again, systems designers have to test and retest dozens of variables to elicit the proper dynamics.

How Much XP?
How much XP do you give? How strong should monsters be?

In my previous life, I was designing a Facebook game. The game had to remain challenging, yet rewarding every time the player logged on. How often would they log on? What is a sufficient reward? What is an excessive reward? Will they play optimally? What if they don’t?

Let’s say a player gains 10% or 11% in their money per session. Not much difference right? Starting with 100 “gold”, after 100 sessions the 10% player ends up with 1,252,782. The 11% player ends with 3,068,844, almost 2.5x as much. You can see how tinychanges add up. Now imagine if instead of a simple 10% or 11%, the money is generated by a formula with six variables as it was in something similar to my my Facebook game: Money Gained = k * Visits * Tourists * Number of Animals * Rarity of Animals * Interest Level * Item Modifiers. Imagine little ripples in any of those variables and how they impact the composite variables.

Now you have an idea of the complexities of systems design. It’s a house of cards ready to be blown over by the first person to come along with an idea to “improve”. What is your opinion of Monopoly? Is it too slow to get to the interesting bits? Does it drag on too long? If so, you probably play with the “house rule” variants where you get money (for some reason) by landing on Free Parking and don’t auction unsold properties. Both of these mechanics are part of a well-balanced system. But the Free Parking modification floods the game with money, making it last much longer. The ignoring auctions modification makes the game take much longer to get to the interesting bits where players are competing for monopolies.

Free Money!
Free Money!

To outsiders, it seems like you just make up some numbers off the cuff and if they work, super. In reality, a good systems designer does a lot more work, testing and carefully tweaking values until either the tower balances or falls over.

6 thoughts on “What is Systems Design? (02)

  1. Wow, I was just talking about this exact problem with
    Monopoly with people the other day. Now I can link people here. I’m
    really curious where the Free Parking rule even started. Everyone
    seems to use it, without really thinking about how it slows the
    game down. I think people like the idea that someone losing could
    come back from behind with a “Lucky Lotto” roll, but that isn’t
    really how the game is designed to work.

  2. Monopoly already has a “Lucky Roll” system in the form of the better Community Chest/Chance cards, though they seem to be balanced so that a run through of the entire deck will reduce the amount of money left in the game.

    I’m still of the opinion that Monopoly can be bad even with all of the correct rules being followed.

    It can be good, too, but not in a way that endears you to your friends and family.

  3. I find myself drawn toward developing the systems which tie these complex sets of variables together. However, I have little to work with besides my own intuition and mathematical abilities; there isn’t much material to be found on the adventures, successes, and failures of others tackling similar problems.

    Are you aware of any communities or content on The Internet that deal with these sorts of puzzles? This stuff’s no joke, and it’s waaaay to complicated to discuss in depth anywhere like Gamasutra. My hypothesis is that there must be people talking about it somewhere. Will you help me find them?

    (note: found this page with a google search. My next step is to look through the rest of your posts for deeper discussions of your own experiences with these systems.)

  4. If there is such a place, I haven’t found it. The thing with the systems we develop is that they are so interdependent that we can only share heuristics between projects. Any sort of in-depth discussion would trail off into minutiae. Perhaps there are “social” game developers talking about it since there systems are all pretty similar, but I have not found such a place.

    There is an article by Richard Garfield where he talks about tuning systems (I believe), where he basically shuns tuning by degrees and instead uses a system where if a card is overpowered, he cuts in in half. Or vice versa, doubling the effects of weakened cards. The rationale is that you would be boiling a frog by degrees, slowly changing numbers and not receiving any delta that was meaningful. By doubling or halving, you can see exactly what effect the variable has at a glance.

    Beyond these kinds of rules of thumb, I have nothing for you.

Leave a Reply

Your email address will not be published. Required fields are marked *

Human? * Time limit is exhausted. Please reload the CAPTCHA.