A pioneering New England developer and operator of online entertainment, with successful long-running titles based on venerable classics from both the game world and fantasy literature, prides itself on the strength of its teamwork, its loyal community of fans, and its focus on excellence.
By using IWL's Maxwell Network Emulator for pre-deployment testing, the New England Developer could tune its netcode to compensate for poor network performance and deliver smooth gameplay for its fans.
The New England developer's online, interactive, multiplayer games must deliver:
All online, interactive, multiplayer games face this challenge. For First Person Shooter games and Multiplayer Online Battle Arena games, meeting the challenge defines success.
In addition, a highly marketed game launch presents a newer and greater challenge – scalability and performance
The challenge becomes “will the game adequately scale to meet the demand of an unknown number of simultaneous interactive players?” and “is the netcode adequately tested?”
The authors (or the publishers) of online, interactive, multiplayer games want to drive enthusiasm, passion, and anticipation for the new game, however, they cannot reliably predict the number of simultaneous players at the launch or the netcode performance at scale.
Often, at the start of the product launch, the number of players exceeds game server capacity and performance, making game play unacceptable.1 Fans with high expectations complain bitterly, leading to negative publicity that compromises the success of the game's production release.
Game developers and publishers must regroup and rethink the go-to-market strategy.
The question becomes “how can we test online, interactive, multiplayer games under a heavy load with adverse network conditions prior to product launch? then based on what we have learned, how we can improve our code to compensate for adverse network conditions?"
How did the New England Developer's team conquer these challenges?
Game players do not want to wait for 20 to 30 minutes while a game downloads and installs on the local device. Players are impatient. The New England Developer found that by using a combination of data streaming solutions from the Happy Cloud and Akamai, they could modularize the game so that enough of it installed within the first five minutes to allow the player to begin, while the balance of the game continued to install in the background. This addressed the "instant gratification" requirement of the gamers.
The New England Developer's team created an internal load testing harness to simulate thousands of players and measure performance degradation as the number of players increased. When a performance limit was reached, they tuned the game application and/or the netcode to eliminate bottlenecks.
In addition, the New England developer's team found a content delivery network that could scale up to provide the necessary resources to accommodate spikes in demand from game players.
The New England Developer's team used the Maxwell Network Emulator to emulate bandwidth limited connections. With Maxwell, the team could emulate a game player on a limited bandwidth connection type (e.g. satellite) in the lab, and examine the game experience for that player on that type of connection.
By systematically stepping through each player's Internet connection scenario (e.g. DSL, satellite), the team can examine the player's experience with the game and adjust, tune, and improve the netcode to meet the requirements of the game for that connection type.
In addition to the bandwidth limited connections, many other conditions affect game play, such as network congestion, resulting in delayed or dropped packets and requiring entity interpolation and latency/lag compensation techniques by the game application. Using Maxwell, the New England Developer's team could introduce the most common network anomalies to emulate a degraded network in the lab. With this degraded network, the team could evaluate the impact on the game application and adjust, correct, modify, or improve the interpolation and latency/lag algorithms appropriately.
“Using the Maxwell Network Emulator made a big difference to the quality of our gaming applications,” reported the Technical Services Manager. “We could precisely emulate real-world game play from the perspective of the players to assure fair access to all players and smooth gameplay."
The New England developer now incorporates the Maxwell Network Emulator in the quality assurance life cycle for all of its game development efforts.
More Information on IWL and Maxwell Network Emulators:
PO Box 66190
Scotts Valley, CA 95067
© 2014-2018 InterWorking Labs, Inc. dba IWL. ALL RIGHTS RESERVED.