[Week 9 Discussion] The Effect of Latency and Network Limitations on MMORPGs (EverQuest2 Field Study) and related papers

Overarching theme: Design of Networked Virtual Environments.

Context: In a traditional non-networked virtual environments, 3D graphics have always been driven by the need to attain greater realism, within the constraints of the number crunching power of contemporary times. In such environments, the entire world state is present on the single device delivering the experience. For networked virtual environments, however, this notion of the omniscience of the world state is invalidated. As such, these experiences need to the designed keeping in mind the problems that be, so that one doesn’t attempt the impossible, and an awareness of contemporary techniques for handling some often encountered issues. My main paper and proposed readings are aimed at discussing these issues.

Primary reading: The Effect of Latency and Network Limitations on MMORPGs (A Field Study of Everquest2) – http://dl.acm.org/citation.cfm?id=1103623 – this paper performs a case study of the effect of varying latency to various gameplay elements for the gameplay of an client-server architecture MMORPG(EvertQuest2).

Proposed supplementary readings: Collaborative virtual environments – http://dl.acm.org/citation.cfm?id=379322 – I chose this paper as I think its a nice starting point to begin thinking about the new design and technical challenges that distributed virtual environments open up, and will serve as a good lead-up to paper 1.
Peer-to-peer support for massively multiplayer games – http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1354485 – I chose this paper as I think it presents an important alternative for the design of virtual environments. Distributed virtual environments have also received significant research interest in recent years.
Latency Compensation Methods in Client Server Games – https://developer.valvesoftware.com/wiki/Latency_Compensating_Methods_in_Client/Server_In-game_Protocol_Design_and_Optimization – The primary reading performs a case study of a client-server based game. I feel this reading provides good insight into what is going on under the hood, in a typical client server system.

Discussion Questions: 

  1. Why is there a consistent improvement in Everquest’s behavior at one particular data point (3000s) for movement time for both players? [I don’t know a reason, perhaps you guys have a hypothesis :)]
  2. How do challenges that are faced in creating networked virtual worlds (in particular games) differ with the genre of the experience? (FPS, RPG, Turn based strategy)
  3. Can intelligent game mechanics design lower latency requirements?
  4. How can designers use the knowledge of technical constraints to create great experiences? (http://www.youtube.com/watch?v=BiQCz2NjPR8&t=11m27s)
  5. Can P2P architectures be employed in main-stream games?

Additional Readings if you are interested in this area :
On the impact of delay on real-time multiplayer games – http://dl.acm.org/citation.cfm?id=507674 – very similar to the main reading, makes for an interesting read.
Networked virtual environments: design and implementation – http://www.lavoisier.fr/livre/notice.asp?id=O3OWR2A3AALOWZ – a seminal book on the design of such environments.
Readings that provide greater insight into the internals of such systems:

  • https://developer.valvesoftware.com/wiki/Latency_Compensating_Methods_in_Client/Server_In-game_Protocol_Design_and_Optimization for actual implementation of real-world systems
  • https://developer.valvesoftware.com/wiki/Source_Multiplayer_Networking
  • http://www.gamasutra.com/view/feature/3230/dead_reckoning_latency_hiding_for_.php
  • Issues related to mobile multiplayer real-time games over wireless networks – http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4543937