I mentioned in the last post that I expect WoDO to be a single-shard game in a similar way to EVE Online. Scott from Pumping Irony, an outstanding blog that you should be reading if you aren’t already, called me out on it. This post is my answer. I’ll make three concessions before digging into my reasoning:
- I could be wrong.
- This expectation may be wishful thinking on my part. The MMO commenting community is rife with this kind of projection onto almost every in-development title. I think I recognize this more than some, but that doesn’t mean I’m entirely immune to the phenomenon.
- I may be underestimating the engineering challenge involved in making it so.
With that out of the way, I still expect WoDO to be a single-shard game. Sic cogito:
- A single shard has enormous advantages for players, and in itself can be used in marketing, at is has been in EVE’s case. It is thus a desirable outcome of the design.
- Prior to EVE Online, the idea of keeping all your players on one server was thought to be impossible unless your total population was so low that one server was all you needed. EVE was the first to try it, and while it has certainly caused problems in places and at times, by and large it’s worked.
- Over time, the problems have lessened and the single-shard technology (barring hiccups,) has worked better and better. In other words, CCP is not only the best in the business at this, but they have an entire team dedicated to making this work better and better.
- I submit that the people who set the parameters for the design will set this as a design goal.
- CCP has shown a willingness to work through the aforementioned hiccups.
- With these things in mind, I further submit that CCP’s engineers will make it work, even if it turns out to work less than flawlessly at launch.
Now, one can make the argument (as Scott did,) that EVE gets away with a single-server setup because its universe is three-dimensional and has lots of empty space. My response: So? Why can’t the WoDO be designed in a similar way? If we regard each system in EVE as a discrete location that multiple players can occupy, then a 10x10x10 3d starcuster contains an order of magnitude fewer locations than a 100x100x1 ‘sheet’ of urban sprawl. The systems in EVE were procedurally generated in the first place; I see no compelling reason why a algorithm for procedurally generating city blocks couldn’t be written. Granted that the algorithm would have to be substantially more complex, and you’d have to put in more stuff by hand to avoid the samieness that you see in EVE (where that actually works,) but I see it as achievable. You’d only have to spread the space out (mostly) across a flat landscape instead of into a third dimension.
At that point, you have to manage how players congregate in the locations in which activities take place. EVE handles this largely via the economic engine that drives the whole game (as well as by the lack of instantaneous travel,) and it’s far from clear that WoDO will have a similarly open economy. But again, this is a mere engineering hurdle. If EVE can handle fleets of thousands fighting in one system and keep client performance acceptable (not good, but acceptable,) I can’t see a reason why a game with what seems to me to be a lower emphasis on huge battles couldn’t keep clients and servers running.
Note that I’m not saying that such a thing isn’t difficult. There’s a reason why EVE remains the only single-shard game with a significant population – Champions Online had to resort to a set of horrendous kludges to make something that’s only technically a single server, with (presumably) a much smaller number of players and a lower concurrency rate. While this may not be how the game turns out, I and a lot of other people would be disappointed by a traditional multiple-server approach – and also that if anybody has proven the conventional wisdom on the limits of MMO technology wrong in the past, it’s CCP. If anyone can do it, they can.