Thursday, August 27, 2009

Delegation In Game Design

An interesting conversation from the Arcen Games forums that I thought I'd just post here, because I thought it expressed an interesting point about game design for truly huge RTS games like AI War. A lot of people wonder how on earth you can manage thousands of ships simultaneously -- well, this is how.

Me (responding to another subject): More to the point, most RTS players find the economic simulation side of it to be something they would like to set up and forget.

Admiral (one of the AI War regulars): I must say, I find that an odd sentiment from an RTS developer. Isn't economic management and other logistical concerns a large/important part of the "S" in RTS? Isn't that what differentiates this genre from the RTT (such as e-Eraser's favorite NEXUS or the amazingly fun old game Myth)?

Calvin Southwood (Arcen techwriter/moderator): It depends where you want to emphasize the S. Many RTS games place equal value on a variety of areas, including the economy, others like AI War focus more on the combat aspect of the game, and have a relatively easy to manage economy.

Me: Right. And it's not that I don't want to have a robust economy -- that's not what I'm saying. What I'm saying is that I think that players (and myself included) want to be able to "set it and forget it" until we next decide to change something. It's like cruise control on your car; you can set it and not have to worry about keeping to the right speed while you focus on other things, but later when you want to go a different speed or drive in more dangerous conditions, you just turn it off and start readjusting.

There's a difference in being able to fine-tune the economy, compared to having to constantly babysit it. Even games like CivIV, if I am remembering correctly, do a good job at being hands-off unless you want to change something. That way you don't always have to have one eye on the economy, you know that it is left there doing the job you last set it to do until you come back to it.

Same with the military ships, honestly -- you can set them and forget them, to a degree, too. You might be able to get better results if you micromanage them to a certain extent, but that's neither required nor always possible. The goal of a design like this is to have the depth there in every area so that people can micromanage if that's their style, but a degree of automation in all those areas so that players can be more hands-off and let it do a pretty good job in their absence.

And of course, the players are very much having to manage the big picture no matter what, so I really think about this as more like delegation. Whatever aspects of the game are less important at the time given your situation or preferences, you can delegate to automation so that you can focus your attention on the aspects of the game that are more important right then. And when the situation changes, you shift what you delegate versus what you manage directly. That's the only way to control something so vast as AI War, but I think a lot of people don't really notice that sort of design in play just because it thankfully works well enough to be pretty invisible.


Lloyd said...

I have to buy your game one day!!!
The more I read about it, the more it looks like it should be good! ;-)

Scott Rycroft said...

Just heard about your game through and it all looks great. I'm a new grad programmer, and your interesting and detailed posts on the technical side were very interesting.

Just learning more about how the game was developed has got me hooked, and I'm probably going to pick it up in the future because of that.

The focus on military is similar to another game I've enjoyed, Sword of the Stars. It focuses mainly on smaller unit tactics and weapons though. Not much on the economy, but there is some if you dig.

I'm personally not a big economy guy, especially for real-time. For big name RTS's, if you don't build up in the correct order at perfect timing, you've already lost. I'd rather have everyone at pretty similar economies and have the tactical/strategic aspect be the focus.

Christopher M. Park said...

Thanks for the note, Scott. Glad the game seems interesting to you. I'm with you in that I think that some of the big-name games out there encourage a specific optimal build order, which gives experienced players a huge bonus over existing ones. I think this problem extends beyond the economy itself, and into the realm of logistics in general with those sorts of games.

For AI War, I really wanted to emphasize actual decision making rather than learned patterns, so the openings are highly randomized in the game to keep people from forming set habits. Making players actually react to the situation at hand, rather than just learning some pattern off of an FAQ and applying that, was a huge challenge but also one of my chief design goals.