Network Effects and PvP
Dec 11th, 2008 by Billiard
Lately I have been observing some interesting trends concerning a couple of the PvP guilds I help manage and mentor as part of the Team Love [kiSu] project. Just a few months ago it seemed that any time our guilds were looking to participate in Guild vs Guild (GvG) battles, we’d see the same teams over and over again, and on any given night there would be a long wait before getting matched up with an opponent, especially as we played later into the night West Coast time. While things are not a good as they were in the heyday of Guild Wars back in 2005-2006, it at least seems better than it earlier this year. I think this was due in part to the release of Age of Conan and Warhammer Online: Age of Reckoning. Both of these games seemed to draw Guild Wars PvP players away - at least temporarily. Over the months though I have seen a lot of old players slowly trickle back into the PvP scene, and things seem somewhat improved because of that. Left 4 Dead is again drawing some players away now from Guild Wars PvP, but many seem to be play L4D in addition to GW rather than instead of, which is similar to the effect that Team Fortress 2 had in the past.
What all of these games have in common is that they involve some sort of co-op PvP play. Looking deeper into the games, you can see that Guild Wars, Team Fortress 2, and Left 4 Dead all allow players to jump into co-op PvP right away and compete on an even playing field. Age of Conan and WAR though require significant leveling and grinding of equipment. In addition, Age of Conan and WAR both require players to commit to a specific server, from which they cannot easily switch. I think it was these two aspects, in addition to a very buggy release, that led to the relatively fast demise of Age of Conan. I also think that even though Mythic is working hard on patches and eventual expansions for WAR, it too will suffer the same demise. The reason why I think Guild Wars PvP has remained attractive for players even as it approaches four years since release is that it doesn’t have these limitations, though there have been and will always be other issues concerning balance and such. This is where network effects come in.
According to Metcalfe’s Law, the value of a telecommunications network is proportional to the square of the number of connected users to the system. The reason why this is the case is because the more people on the network increases the likelihood that there is someone any given individual on the network would like to communicate with. This is one reason why cell phone companies typically offer free in-network calling while also continually adding on and growing their networks. This relates to PvP in that the larger the pool of PvP players, the more likely PvP players will find others to play with, and quickly. So when games like AoC and WAR spread out their PvP player base across multiple servers in order to help reduce lag caused by large scale PvP battles, the usefulness of their “PvP Network” is exponentially decreased. This is one reason why during the early stages of each game things seems to be fun and going great - but suddenly a month or two later activity can seem to have dropped to nothing. At game release, you are seeing exponential gains to network value with the addition of each player to a particular server. But as the initial interest wanes, you start seeing exponential losses. Players complain about nothing to do or lack of end game content, PvP scenarios take increasingly longer times to pop, and previously busy instances get deserted. On top of the server separation problems, requiring players to grind for hours in order to earn levels or equipment as causes the player base to become widely dispersed within a given server, further contributing to decrease value of the PvP network on the particular server. This secondary network effect of grinding should not be as pronounced if players can reach suitable character and equipment levels during the initial euphoria games often experience just after launch.
These notions suggest that any new co-op game that is primarily PvP based should not restrict users to a single server, nor should it require more than say a week of grinding before players are able to participate in high end/end game PvP. Enabling players to reach the “best” PvP aspects of the game relatively quickly will help you maximize the size of the PvP Network, and thereby the value of the network to the players. If some players reach the higher levels but have no competition, or if they are stuck on a server that has few opponents, eventually they will see little PvP value to the game and look to move onto something new, or even back to something old that offers that value.
An interesting thought, and rather intuitive since even players who think nothing of networks would say they hate server separation. World of Warcraft, where the grind is not just a necessary part of the game but is instead the actual content, seem to be uniquely exempt from this problem, since PvP has been growing into a legitimate exercise there, complete with tournaments and everything. Perhaps, though, that is simply because Blizzard has SO MANY CUSTOMERS that can afford such exponential losses?
Does this mean that ANet flat out did it wrong with skill and item unlocks? I had heard they’re abandoning the need for unlocks in Guild Wars 2’s PvP, so perhaps…
One question that springs to mind, then: is there a critical mass for successful PvP games? As you wrote, companies will often separate servers to reduce large-scale lag. However, even if they were convinced that unified servers and the networks they facilitate were superior, the point is moot if they can’t afford the infrastructure (i.e. large enough servers) to support the load. If we think that, like AoC and WAR, games that violate this network concern are doomed to failure… does this mean that only companies able to support large enough, lag-free, unified servers should consider designing games using PvP as a significant selling point?
I would suggest that for games with a lot of PvE related emphasis and grind such as WoW, network effects are less problematic. Players may solo or party regularly with a small group of people in order to grind away - they lack of 100 or 1000 other players in their particular network has much less impact that it would for a PvP related game where players rely on others to be around to combat. I think with PvP growing though in WoW you may still see these sorts of issues - at least on the PvP side.
Also, I think that there is a critical mass required for PvP games, and that companies are aware of this. What is at odds though is their ability to handle everyone on one or few servers (Battle.net’s USWest and USEast for example). I mean - how can you build an endgame in AoC and WAR that requires so many players to be in the same server/instance at once, yet not have servers designed to handle that at a playable level?