To sum this up effectively and as simple as possible: A sandbox game keeps the player (or players) busy with a chain of events and tasks. There is a certain ‘rhythm’ based on the intervals between whenever a task (doesn’t matter if it’s a ‘chore’ or a ‘carrot’/'goodie’) pops up and gets performed, when an event is triggered and when the player has to reorient himself, aka “What am I going to do next?”. In an ideal situation, the player doesn’t have to reorient himself at all.
The diagram has to separate branches. The red one is the very simple ‘isle’ sandbox design. After having performed a task, there is no reaction of the environment around the player, nor is there development outside the player’s microcosmos. Usually, there is a time span between the tasks in which the player is pondering about what he will do next. Projected onto a space game, the factor that defines the outcome of the reorientation phase lies in what the player’s basic occupation is, so whether he favors trading or a more martial style of playing, and how many possibilities the game holds open in that situation. This is usually shaped by the price of items and ships, the galaxy map and occasionally by racial differences. I have noticed that as for space sims, it is quite hard to keep up the “Aha”- and “Wow”-moments high in number and concise in appearance. Frankly, there are hardly any “Aha”-moments in current space games (the exception proves the rule), as most mechanics are either obvious - a prime example of that would be the rule of thumb in a simple combat ship design, in which ships with more firepower or shields are simply better than everything else - or built on top of simple, easy to see through mechanics - yes, I’m talking about you, Freelancer weapon balancing. But after all, there may be a way to analyze that and come to a solution (thinking game design atoms). That will mean a lot of handwork though.
Back to the diagram: The plain black branch is what I like to refer to as the “waterfall” sandbox design. It’s a lot more laborious to design, let alone to implement and to maintain - but it is more interesting to design, of course, and a lot more fun to play in the end. It limits the importance of the player as a parameter for the game world, so the environment (factions, economy, etc.) may take a completely different direction, hence the “waterfall”. Though the player is able to manipulate his environment; if he participates in raids of a faction against another faction in the conflict; if he’s a driving economical power (again, space stations, trading, escort/police service, blah, blah, blah); and of course if he has a huge fleet, an enormous wallet and can just gank everything and everyone in the universe.
In the diagram, the normal arrows point at events or tasks that get triggered or are opened after player interaction or rather the performing of a task. The diamond arrows on the other hand point at events or tasks opened up for the player, without the need of player interaction - they just happen or appear. An example of that would be the beginning of a conflict over a system rich on sources. No player interaction needed there. Although, the faction may now offer missions regarding this conflict (raids, defense, recon) to the player. The goal would be to make the waterfalls and occasional isles as long and as rich as possible.
Another trick beside the usual waterfall and isles, and of course, making the basic elements of the game - trading, combat, exploration, story - as fun as possible, to keep the player at it is by adding fun activities or “pretended” fun activities. Say, space racing, gambling, etc. The key there is to add a money bonus to it.
Other than that - my typical credo - : Increased interaction between the player and his/her (AI) environment. A martial player will eventually have somebody holding a grudge against him at a certain point. The player’s standing towards one or more factions will eventually be so low, that the faction will send ruffians to attack his fleet or his stations (if he has any) as an act of retaliation or rather aggressiveness. That would also be a reason for the player to continue shaping his environment and attacking other factions or teaming up with some.
I hope I have cleared things up a bit…
PS.M64 mentioned in his Blog that one of the issues that FOSS game development has to tackle are huge, bloated “game designs” (quotation marks by me), which try to imitate commercial games to some degree. FOSS games should rather focus on casual games with smaller designs, so he says. That is true for most cases, but in the end, everything gets down to the game design, and if there is none, a bigger game can’t be played consistently nor be really fun. I want to prove that 
Tags: FOSS, FOSS game, FOSS game design, Isle design, Sandbox designs, Sandbox game, Simulation, Space game, Space sim, Task, Waterfall design
#1 by Tranberry on February 8th, 2009 - 6:55 pm
Great article Sindwiller! Now I would like more of this, game mechanics and the like. hurray!
#2 by ghoulsblade on February 9th, 2009 - 2:26 am
yep, interesting article =)
i have to think about escape velocity here, there you could do always do simple random generated delivery missions between planets, but if you did a few there was a chance that after you completed one of the missions, you’d be offered a bigger mission. i think those were hand-designed. then you’d do a few more random missions before getting the next big one. if you did a few of the big ones, they’d give you access to a few private jumpgates that allowed you to take shortcuts.
i think this is a nice way to “stretch” hand-made missions with random generated ones, and i also like the reward, it’s luxurious and cool, but you can also live without it if you don’t do that trading at all.