User:Veron/Boon spread and parties

From Guild Wars 2 Wiki
Jump to navigationJump to search

Author: Veron
Last updated: Oct 11, 2014

Introduction[edit]

In organized world boss raids such as Tequatl and Triple Trouble Wurm, boons such as Might.png Might and Stability.png Stability play a major role. Thus, making sure that a zerg has proper boon coverage is an important part of a fast and smooth run. Two of the most common situations are (1) might stacking via fire fields and blast finishers before burns and (2) stability coverage to counter knockbacks, such as at Tequatl or while waiting to be swallowed at Amber wurm.

Most AoE buffs have a 5 target limit. Prioritization works as follows:

  • Party members (that are in range) first (added in Oct. 15, 2013 update)
  • Then by proximity to centerpoint of skill (usually you)

Indisputably, the absolute optimal strategy is to have all members of the zerg partied, each with a proper party composition. For might stacking, this means having at least 9 blast finishers to get to the 25 might stack cap. For stability, this means having at least one guardian in each party in order to use  "Stand Your Ground!".png "Stand Your Ground!". This guarantees that every player in the zerg gets to 25 might or gets stability. Thus, any sufficiently organized group should elect to use full partying with proper party compositions.

In the absence of perfect organization, other strategies with suboptimal outcomes can be easier to pursue. One popular heuristic strategy that has been popularized by TTS is to completely unparty:

  • "No one should party with others unless they are in the Defense teams." --TTS Tequatl guide
  • "Do not party unless absolutely required. This is to ensure buffs can be spread evenly throughout the zerg." --TTS Great Jungle Wurm guide

How effective is the unparty strategy? How does it compare to the optimal outcome of perfect boon coverage? Intuitively, players who are in the center of a zerg stack would nearest neighbors of more other players than players who are on the edges. How does unpartying compare to other strategies such as full parties with random compositions?

In this study, I will investigate the effectiveness of different strategies using computer simulation in MATLAB.