Call of Duty discusses matchmaking process for multiplayer games

Activision said it will actively discuss its multiplayer matchmaking process with Call of Duty fans in an effort to be more collaborative.

The company talked about its current process for multiplayer matchmaking in a blog post today, but it also said that it will discuss other matchmaking processes, such as for Warzone and Ranked Play, in the future. Activision will be testing refinements to matchmaking in the days to come.

The community refers to the matchmaking process as skill-based matchmaking, but while Call of Duty considers skill (or more specifically player performance) as a component as do most in the industry, the truth is that developers consider and prioritize several other factors when bringing players into lobbies.

Activision said it considers these factors in matchmaking right now:

• Connection — As the community will attest, ping, or response time, is king in Call of Duty matches. Connection is the most critical and heavily weighted factor in the matchmaking process.
• Time to match — This factor is the second most critical to the matchmaking process. All players want to spend time playing the game rather than waiting for matches to start, the devs said.

Other factors that matter are:

• Playlist diversity — The number of playlists available for players to choose from.
• Recent maps/modes — Considering maps players have recently played on as well as player mode preferences, editable in Quick Play settings.
• Skill/performance — This is used to give players –- a global community with a wide skill range -– the opportunity to have an impact in every match.
• Input device — Controller or mouse and keyboard.
• Platform — The device, PC or Console, that the player is playing on.
• Voice chat — Enabled or disabled.

Game data indicates that having some limitations on the disparity of skill across the players in a match makes for a healthier ecosystem. The devs also understand that many high skill players want more variety of experience, but often feel like they only get the “sweatiest” of lobbies. The devs have heard this feedback clearly and will continue to test and actively explore ways to mitigate this concern.

In addition to today’s blog, the technology team is developing a ping and matchmaking white paper to provide more granular information about Call of Duty matchmaking. Call of Duty previously talked about matchmaking for the launch of Modern Warfare III Season One late last year.

More details

Connection speed: Whether you’re playing for fun with friends or looking to climb the leaderboards, connection is the most important part of the online Call of Duty experience. Connection dictates the speed at which the game can transfer information from every player to and from the game’s servers.  

Call of Duty’s matchmaking process evaluates a metric called “Delta Ping,” which is the difference in round trip time of the data between the player’s best data center (almost always the one closest to the player) and the data center onto which the player’s lobby has been placed (based on all players in a lobby).

To reiterate, Activision always tries to maximize the times it places players in data centers that are closest to them.

Call of Duty uses a client-server model to host matches, where the time it takes to share information between the player (client) and the data center (dedicated server) has an impact on the overall feel of a match.

The Call of Duty netcode, which devs discussed in the past, works to reduce the effect of latency, but cannot completely eliminate it. The matchmaking process seeks to reduce the overall amount of latency by prioritizing stable connections or low ping – with a shortened wait time in mind.

Time to matchmaking: Any form of matchmaking takes time. If the wait time in a lobby is excessively long, players typically recycle the process by canceling out of matchmaking search and restarting it, or even quitting. This does not quicken the matchmaking process and in fact can even be detrimental.

For example, in the popular Modern Warfare III “Rustment” playlist (consisting of Rust and Shipment in rotation) – players often leave lobbies and/or matches early on, hoping to requeue into Shipment instead. This creates a vacant spot on a team during an early stage of the match. As the matchmaking process may prioritize backfilling that spot, this could result in players perceiving that Rust is disproportionately selected over Shipment. TL;DR -– trying to cherry-pick maps may have an unexpected result.

The team’s goal is to ensure that players spend more time playing matches rather than waiting for them.

Measuring skill for matchmaking: While skill is one of several factors in Call of Duty multiplayer matchmaking, the devs know the community wants more information about how it fits into the process.

Skill is determined based on a player’s overall performance: kills, deaths, wins, losses and more, including mode selection, and recent matches as an overall metric across all multiplayer experiences. This is a fluid measurement that’s consistently updating and reacting to your gameplay. Skill is not only a factor in matchmaking players against appropriate enemies, but also when finding teammates.

Call of Duty has historically considered player performance among other factors as part of its matchmaking process. The development work in this area dates back as early as Call of Duty 4: Modern Warfare (2007). Skill is implemented across the video game industry, and the devs recognize that continuous refinement is required to deliver the best possible experience for players.

Activision uses player performance to ensure that the disparity between the most and least skilled players in the lobby isn’t so vast that players feel their match is a waste of time.

The team’s data on player outcomes clearly indicates that the inclusion of skill in Call of Duty’s multiplayer matchmaking process (as it currently stands) increases the variety of outcomes experienced by players of all skill levels. In other words, all players (regardless of skill level) are more likely to experience wins and losses more proportionately.

The data shows that when lower skill players are consistently on the losing end, they are more likely to quit mid-match or stop playing altogether. This has an effect on the player pool. A smaller player pool means increased wait times and a higher likelihood of connection issues. This can compound over time to create a spiral effect. Eventually, when only high-skilled players remain because lower skilled players have quit out of frustration, the result is an overall worse ecosystem for everyone.

Game data indicates that having some limitations on the disparity of skill across the players in a match makes for a healthier ecosystem. The devs also understand that many high skill players want more variety of experience, but often feel like they only get the “sweatiest” of lobbies. The devs have heard this feedback clearly and will continue to test and actively explore ways to mitigate this concern.

Does Call of Duty consider player engagement (time played) as a factor in matchmaking?

The devs do not consider how often, or how much, you play when determining matchmaking.

Does the Call of Duty matchmaking process impact any in-game elements such as hit registration, player visibility, aim assist, damage, etc.?

No. Our matchmaking process does not impact gameplay elements.

Does spending money on Call of Duty content (such as bundles, Battle Pass, or BlackCell) change how players are matched?

Money spent does not in any way, shape or form, factor into matchmaking.

Does Call of Duty use bots in Multiplayer matchmaking?

Call of Duty Multiplayer does not use bots as part of the general matchmaking process. If this changes in the future, we will inform the community.

Do partners or content creators get special consideration in general matchmaking?

No. We do not change the matchmaking process based on who owns the account. In specific cases, such as for events like Call of Duty Next, we may be required to customize how lobbies are formed; however, these events usually take place in private matches and do not impact general matchmaking.

Have you ever considered an opt-in/opt-out system for the matchmaking algorithm?

Our data suggests that splitting the player base with an opt-in / opt-out matchmaking system will have negative consequences on the overall player pool. That means, potentially, longer wait times based on the type of matchmaking selected (plus add into that playlist, map and mode history, platform, and more) and matches with poor connections.

Have you ever tested removing skill as a consideration from matchmaking?

We have run tests over the years to determine if removing skill as a consideration from matchmaking makes sense. We will continue to launch these tests periodically. To date, the data remains consistent with what we detailed above — players tend to quit matches or stop playing if they’re getting blown out, resulting in a negative overall experience for all players in the lobby and the general player population. We purposefully do not disclose when these tests occur because it may impact feedback or the data
we see during these tests.

Have you considered removing skill from matchmaking in specific general multiplayer game modes?

We have considered this in the past and we will continue to examine if this idea makes sense as part of an experimental playlist or in specific modes. We have nothing to announce on that front today.