Scott Hartsman - Off the Record

A comment was raised on f13.net yesterday that I see a lot every time an MMO doesn’t make it all the way out the door. Emphasis is mine on the parts that caught my attention:

True, but I have to think that someone has managed to get people to collaborate in other venues… and so we are not talking about creating something completely new here. I am just having a hard time figuring out why skilled people (I assume some of the people making MMOs actually have the skill to work on other types of projects and just Chose an MMO) given a LOT of money (yea I still see 10’s of millions as a lot) cannot get through a successful design/production phase. It seems that there is something inherent in the MMO beast itself that kills the process.

There definitely is. And it’s a lot more than a single “something.” Some of the issues have to do with MMOs in particular, and others are compounded by the types of people who are most likely to attempt to develop them. Generally very sharp and motivated people.

My reply clipped from the same thread:

(Disclaimers: Personal opinions here only, unrelated to SOE. I haven’t even remotely been involved with G&H or Perpetual in any capacity and don’t know a thing about their game. My comments are speaking entirely in generalities. Dealerships negotiate their own prices. Beware of falling rock.)

A few observations from past MMOs:

#1: MMOs are still really young. To a lot of the people working on them, it very much is creating something entirely new. Compare to movies or single player games, for instance. It’s less of a challenge to staff those types of projects up with people who’ve worked on them before, in all of the right positions. Doing the same on a high-budget MMO remains next to impossible.

I don’t mean “key management” or “leads” like you see in studio announcements and press releases all the time. I mean everyone other than a small number of entry-level folks. Until you’ve done it once, you have no idea what you’re getting yourself into.

I don’t know of a single high-budget MMO that’s been staffed with that kind of experience throughout, simply because those people just plain don’t exist yet in sufficient numbers. We’re just now at the point where it’s starting to become possible to build teams like that.

Just a guess, but I’m betting that you don’t hear from the $100m movie set: “Yeah, Bob the Key Grip has done this once before, and he picked out some really sharp guys from a construction site downtown to do the rest. He’ll teach ‘em what to do.”

Leading to…

#2: The things that make for a great demo and pitch that get you funding, publishing deals, et al, are a much smaller part of making a great MMO than they are of making any other kind of game, and it’s easy to lose sight of that.

This is painful for MMOs in particular because of the unique (huge) number of critical, non-sexy things that you have to succeed at, where failing at any one of them can entirely sink your game:

- Pipelines
- Tools
- Infrastructure
- Stability (again, doubling the work - the client and all the servers)
- Scalability
- Stability
- Security (added this in for the blog post - Can’t trust that client)
- Performance (optimize both that client and all those server processes)
- Oh, and..Stability

In any development effort that has a finite set of resources ($$$ + time), the more you invest in the flash elements, the less you can invest in the far less sexy parts. (Core files aren’t sexy.)

Which, in turn, leads to…

#3: Wild misscoping. It’s a common newbie (and overly-optimistic-veteran) mistake to scope far too optimistically, as the schedules end up based mostly on the flash elements and end user features.

If a person is new at making one of these (especially noted with people from non-MMO games backgrounds), they tend to be more likely to focus on scoping dev time out with more of an emphasis on the visible features than the budget will end up allowing, and not enough on the critical, non-visible features. Those, coincidentally, end up taking far longer than anyone ever predicts.

The team who scopes 80% of their time on the visible features and 20% on the rest is going to make a far different game than the one who scopes 25% features, 25% tools/pipelines, and 50% stability/scalability/infrastructure.

If your timeline has some elasticity, you can make up for misscoping by stretching the schedule, and still go on to make a great game. If you can’t, Bad Things happen.

There are plenty more things that go wrong, and from all different angles, but from the production “why can’t people seem to get these out the door?” angle, these are the ones that’ve been the first to jump out at me.

As for the things that go wrong from the other angles? Now that’s a subject for another post entirely.