(Continued from Making PvP Scale.)
When I wrapped up the first part of this discussion, I had just mentioned the core issue with designing scalable PvP: finding some way to maintain a rough balance between technique and strategy regardless of the numbers of players involved in the conflict. Before I go any further, I’m going to admit that I don’t think there’s any perfect way to go about this - whenever you have a truly large number of players in a battle, the drive to “follow the crowd” can sometimes displace all reason and the resultant zerg is inevitable. Instead, the goal we should shoot for is to minimize these occurrences as much as we can. There are some design considerations that, if carefully considered in the process of creating a game, can heavily influence whether or not the resultant PvP system favors zergs or more interesting combat.
Any time you and 99 friends are charging into a group of 100 enemies, technique is pretty much going to go out the window. That much is obvious, and any attempts to design a technique-based game with the vision of giant armies colliding in close quarters are probably in vain. Rather than attempting to do the impossible, a different approach is far more likely to succeed: motivating players to split up. A number of games already do this to some extent - usually, the manner in which rewards are split from vanquished foes scales in such a way that the best rewards (on a per-kill basis) come from small-group battles. Unfortunately, such a measure tends to overlook the relative effort/reward ratio involved. Small group battles require a much larger investment of personal effort to achieve individual kills, while zerging tends to rake in a fairly significant number of kills overall with little personal investment, even if the share of the spoils is much smaller. Thus individual players often gravitate to the zerg if they feel that it’s a more consistent source of rewards.
It can be hard to make rewards for small fights attractive without at the same time encouraging abuse of the system. There are many examples from a variety of games with PvP-based rewards where players attempt to game the mechanic, trading kills to acquire the benefits at a much greater rate (and via a much simpler process) than was intended. The core reason for this is that the measures designed to convince players to split up are all reward-centric. They try to change players without changing the actual gameplay - only the rewards obtained from it. It follows logically then that the same players who are willing to be affected by these incentives are the ones that are most likely to abuse gameplay in order to obtain rewards.
Better would be to design reasons to split up into the underlying gameplay itself. Instead of having a major objective that everyone focuses on, decentralize the PvP objectives. Instead of having 3-4 objectives in a zone, have 10-20: the more objectives you have, the harder it is for a zerg to dominate them, and the more effective hit-and-run attacks become. Instead of having movement based completely on simple overland travel, which allows zergs to move just as fast as small groups, provide mobility advantages to players that are split up.
Perhaps your world has a teleport grid that links all of the battle objectives, but that grid has limited throughput: only a certain number of players can teleport into a given location each minute. This hinders the movement of zergs while allowing small groups to roam freely, thus discouraging large groups from chasing smaller ones while still allowing small groups to chase other small groups. Contrast this to the party speed buffs employed by some MMOs, which, while sometimes helping small organized groups evade less organized zergs, also allowed them to evade other small groups if they didn’t wish to fight. (In addition, there is no actual restriction against zergs using such speed buffs; it just isn’t as common due to the disorganization of a typical zerg.)
With such a system, you get a result that can function well regardless of the number of players active in the zone: with a small population, conflicts occur at a few of the nodes, with the quick-travel system allowing a small set of defenders to quickly arrive on the scene where another small set of attackers strikes. With a larger population, the same still occurs, but it’s now possible (an encouraged) for the attackers to strike at many nodes at once, and also possible (and required) for the defenders to allocate their forces to the defense of these nodes. The strategic element remains in both cases (in choosing where to engage) but the tactical element is also preserved (due to the soft limitations imposed by the quick-travel system, it’s much harder to get a large force into position to respond to the strike of a smaller attacking force before they disappear again). You also have the occasional strategic surprise when one group or the other decides to make the sacrifice of forgoing quick-travel to organize an attack in force on a particular point.
This is certainly not the only approach one might take to designing scalable PvP, but it is one that I think has merit. I’d be curious to see what other ideas people might have when it comes to PvP systems that can scale.