Critical Mass

So, my goodly friends over at BeyondUnreal are having a bit of a lovefest over Unreal Tournament 2004 being a year old. And let's be honest - it deserves it. UT2004 is hands down the most versatile, action packed, old school online shooter this side of Neptune - and all the Counter-Strike fanboys can bite me if they don't agree. The fact that there's still a franchise out there that respects the value of a good Capture the Flag match earns mucho bonus points in my book.

But my fondness for the Unreal platform comes from it's mod framework - which is bar none the best out there. I'm a big fan of Doom 3's scripting structure, where you can edit the game on the fly without compiling for many objects, but readers of this blog know that I place value on Epic's decision to keep their accessible code open, free of Visual Studio and easy to learn.

Unfortunately, there's this odd problem. There's almost nobody playing them. I've lamented on this before, when I swore I was done with multiplayer modding (I am, of course, working on one now...) ... but this time you can hear it from none other than Steve Polge himself:

It's unfortunate that some of these great mods aren't seeing more players. I think part of the problem is that there are so many online multiplayer FPS games out there that it is much harder to gain a critical mass of players for a mod. Back when counterstrike first became popular, it wasn't competing against many online FPS's in the same genre. Today, a new mod typically competes against many other mods and full retail games if it chooses a conventional genre, such as a tactical shooter or WW2 FPS, while it has to be overcome player inertia to get past its learning curve if it is unconventional (Air Buccaneers).

Both these scenarios require a mod to be very polished to have a chance of gaining a significant player count - players have many other options rather than sticking with a mod that is still very much a work in progress.

I don't mean to suggest that mods can't be successful in today's environment, but rather that mods are likely to need to be further along in terms of polish and completion before they can start attracting a big following. This means more perseverance on the part of the mod team during the early development phase when they aren't getting a lot of positive feedback.

Now Red Orchestra's biggest strength is the quality of it's art production. The screenshots are simply phenomenal. I honestly can't speak very well to it's gameplay - but when it comes to polish and completeness ... RO is definately on the far end of the spectrum. There's simply not a lot of mods, for any engine, with this kind of shine.

So - here's the question. If a mod that is of such near professional quality that it actually beat out several professional studios during a mod contest to win the grand prize can't muster a significant online presence - who can?

I think that answer is - almost nobody. Unless you're modding for Valve, who will shove your game into Steam and force feed it upon a mass of unexpecting gamers ... how will anyone achieve a critical mass which is required to maintain online players?

The thing that Steve isn't mentioning is that online games start life with a serious handicap. You almost have to have human players to get the full experience. While Steve himself is responsible for some of the best bot AI in game history - nothing replaces having a boatload of real humans trying to kick your ass. I found this out in full color when I tested the first round of my mod, Riftwar. The bots were semi-entertaining ... but mostly just target practice compared to having even just a handful of real players on a server. But of course, most people hadn't played it - didn't realize what it was like and didn't know anyone to try it with. So there's this awful chicken/egg problem that any online game (mod or otherwise) suffers from.

But what Steve is saying is - it's not the age of Quake anymore. People aren't willing to try an online game simply because they haven't yet. There's too many to choose from.

Here are three ideas to work around this:

Focus on the offline experience One of Red Orchestra's most noted flaws was an utter lack of bot support (somthing they are just recently addressing). With a strong AI presence, people might be able to at least get the taste of an offline game to whet their appetite to find a server.

Or, of course, just design a single player game completely and forego the multiplayer. Personally, I think coop is the new black. Design a game as a single player with a coop component (which is actually what I'm doing now) ... get a bit of both worlds and see what happens.

Focus on smaller groups Battlefield 1942 has apexed, I think, the "big server" concept of shooters. There's definately a draw to having more than thirty people playing in the same small space. It's also much harder to get a decent game going. On the other hand - if you focused on say, only six people being online ... maybe a 3v3 match, especially with decent bot support ... that's a lot smaller critical mass required.

Learn from Microsoft Ewww, that stung to write. But let's face it - Halo 2's matchmaking concepts are still the talk of the town, the belle of the ball, the ... OK - I'm out of phrases. At any rate, if you can't rely on people to find each other to play your game ... find some way to bring them together. And here's a tip - a simple chat lobby isn't going to cut it. It never really did.

Steve's solution seems to be "get more professional" ... and to be honest, that spooks me. If mod teams get any professional they might as well stop calling themselves mod teams. And by forcing themselves to have the same demands as a pro studio, mod teams will be cutting their ability to innovate - because as Steve mentions with Air Buccs, that will just make their job harder.


What would be awesome is to see some small mod teams focus on the first two suggestions, while some kindly game company focuses on the third.