Skill-based matchmaking was supposed to solve one of online gaming’s oldest problems. New players would avoid being crushed by veterans, average competitors would receive reasonably balanced matches, and high-level players would face opponents capable of testing them. That sounds fair. On paper.
In practice, SBMM has become one of the most distrusted systems in modern multiplayer gaming. Players frequently accuse it of manufacturing streaks, punishing improvement, breaking up social groups, favoring retention metrics, and turning casual playlists into ranked competition without visible ranks.
Some accusations are supported by recognizable patterns. Others are difficult to prove because publishers rarely reveal enough information for anyone outside the company to audit what is happening. That information gap is the real crisis.
SBMM Is No Longer One Simple System
The term skill-based matchmaking creates a misleading picture. It suggests that every player receives a rating, the system finds nine or eleven people with similar ratings, and the match begins. Modern matchmaking is rarely that simple.
A matchmaking system may consider connection quality, physical location, queue time, input method, platform, party size, playlist population, account history, recent performance, estimated skill, team composition, rank, and available data centers. Some factors may act as hard restrictions. Others may be weighted preferences that weaken as the queue gets longer.
Two games can both claim to use SBMM while operating in completely different ways.
One title might measure skill across hundreds of matches. Another might react aggressively to recent performance. One system may form a lobby based on average team strength. Another may select players first and then divide them into teams. Those approaches can produce radically different experiences even when the predicted win probability appears balanced. The label has become almost meaningless without supporting details.
Hidden Ratings Create Visible Suspicion
Competitive ladders traditionally show players where they stand. Elo ratings, divisions, league points, win records, and promotion requirements give competitors something they can inspect. Public matchmaking often does the opposite.
A game may estimate a player’s skill internally while showing no rating, no acceptable skill range, no match-quality estimate, and no explanation for why certain opponents were selected. Players are expected to trust an invisible calculation that can alter nearly every session they play. That trust collapses quickly when the result feels inconsistent.
A strong player may dominate one lobby and then spend several matches against coordinated parties. A developing player may have a good evening, return the next day, and feel as though every opponent suddenly has tournament-level movement. Friends with different ability levels may discover that playing together creates miserable matches for one side of the group.
Without visible data, players build their own explanations. Some call it SBMM. Others call it engagement-optimized matchmaking, loser queues, winner queues, forced streaks, or punishment lobbies.
Publishers often reject those claims, but rejection alone does not settle the argument. A black box defended by another black box remains a black box.
Call of Duty Forced the Debate Into Public View
Few franchises are tied more closely to the SBMM argument than Call of Duty. For years, players complained that public matches felt aggressively competitive while Activision provided little detailed information about the system.
That began changing in 2024 when Activision released a series of matchmaking papers. The company described connection quality and time-to-match as major priorities, while also explaining how skill was included among several matchmaking factors. Activision also disclosed a Modern Warfare III experiment in which skill restrictions were loosened, though not removed.
According to the paper, the test produced worse outcomes for parts of the player population, including higher quitting and lower return rates. The disclosure was valuable. It was also incomplete.
Players could see the company’s conclusions, but they could not inspect the full testing framework, raw datasets, exact skill calculations, lobby distributions, or how different player groups were affected. A broad retention gain might hide a worse experience for veteran players. An improved average quit rate might come with weaker connection quality in certain regions. Aggregate results rarely answer individual complaints.
The argument changed again with Black Ops 7. Following beta feedback, Activision announced that open matchmaking with minimal skill consideration would become the default for multiplayer. Later patch notes confirmed that open matchmaking was being applied across multiplayer playlists, and Warzone adopted a similar approach for non-casual modes.
That shift matters. After years of defending skill consideration as an important part of match quality, one of the largest shooter franchises moved toward more varied public lobbies.
The decision does not prove that SBMM was inherently bad. It proves that players had enough influence to force a major publisher to reconsider how heavily skill should shape casual play.
Apex Legends Is Testing Transparency in Public
Respawn has taken a different approach with Apex Legends. Rather than presenting matchmaking as a settled system, the studio has increasingly discussed it as a collection of active experiments.
Apex introduced a pre-match skill display in Ranked that shows the tiers represented in the lobby. Respawn said the display was intended to build trust and give players a visible way to document matches that appeared poorly balanced. That is a small feature with a large effect.
A player who sees several Masters competitors in a Platinum lobby no longer has to rely entirely on instinct. The mismatch can be captured, shared, and discussed using information supplied by the game itself.
Respawn has also published updates covering queue-time tests, premade restrictions, rank-width targets, solo queue rules, and changes intended to reduce extreme skill gaps. In June 2026, the studio reported that longer maximum queue searches reduced large rank differences, especially among stronger players, without significantly increasing wait times during the test.
The studio still keeps many internal calculations private. Players cannot fully inspect their matchmaking profiles or reconstruct every lobby decision. Yet Apex demonstrates that developers can reveal meaningful information without publishing exploitable source code. Visible rank ranges, test goals, measured results, and known trade-offs give the community something concrete to judge.
Casual Play Has Lost Its Definition
Much of the anger surrounding SBMM comes from a basic identity problem. Players no longer agree on what a casual playlist is supposed to provide. For some, casual means low stakes. There is no visible rank to lose, so the match should be relaxed.
For others, casual means balanced competition without ladder pressure. The absence of rank does not mean weak opponents should be thrown into the same lobby as experts.
Veteran players often remember public servers, persistent lobbies, community-hosted matches, and mixed-skill rooms where the same names appeared night after night. Those spaces were not always fair. They were social, unpredictable, and memorable.
Modern matchmaking trades much of that identity for speed and statistical balance. Players are placed into temporary groups, the match ends, and the lobby dissolves. Rivalries rarely develop. Strong players are separated. Weak players are protected. The system quietly rebuilds the next match from scratch.
The result may be mathematically cleaner while feeling emotionally flat. A balanced match is not automatically a fun match.
Team Balancing Is Often Mistaken for SBMM
Players also tend to blame every bad lobby on skill-based matchmaking, even when team balancing may be the larger problem. Matchmaking determines who enters the lobby. Team balancing determines which side each person joins.
A system might gather twelve players with a broad skill range, then divide them into two teams with similar predicted win chances. One team could receive the strongest player and several weak teammates. The other could receive a more consistent group of above-average players. The predicted outcome may be close to 50 percent. The experience for the strongest player may still be awful.
That player feels responsible for carrying every objective, winning every duel, and compensating for teammates who cannot keep pace. The opposing team experiences steady pressure without relying on one standout performer. On a spreadsheet, the match is balanced. In motion, it feels rigged.
Publishers rarely explain this distinction clearly. As a result, public discussion compresses skill estimation, lobby creation, team assignment, spawn logic, party matching, latency, and streaks into one angry acronym. SBMM becomes the name for everything players dislike.
The Engagement Question Will Not Disappear
The deepest suspicion is that matchmaking is designed not merely to create fair games, but to influence how long people continue playing. That fear is not irrational. Live-service games measure retention, session length, churn, purchasing behavior, queue abandonment, and countless other signals. Matchmaking changes are commonly evaluated through player behavior, not simply match scores.
Activision’s own papers discussed return rates and quitting as key results. Respawn evaluates queue times, match quality, tier differences, and player responses during tests. These are reasonable measurements for developers operating games with millions of users. The problem appears when players cannot tell where fair competition ends and behavioral management begins.
A system designed to protect new players may also improve retention. A system designed to reduce quitting may create tighter matches. A system designed to increase session length could potentially produce similar statistical patterns. Intent cannot be determined from the player’s side.
Publishers do not need to disclose every commercial metric, but they should state whether spending history, cosmetic ownership, recent purchases, battle-pass activity, session length, or predicted churn have any role in matchmaking. A direct denial would carry more weight than vague statements about fairness.
XDefiant Proved That “No SBMM” Is Not Enough
Ubisoft marketed XDefiant partly as an alternative to heavily managed public matchmaking. Its unranked playlists avoided SBMM, while Ranked used skill-based matching. The game was presented as a return to an older arcade-shooter style with more varied public lobbies. That position attracted players who were tired of tightly controlled matches. It did not guarantee long-term success.
Ubisoft ended development after the game failed to retain a large enough audience, and the servers closed on June 3, 2025. The shutdown showed that removing SBMM cannot compensate for every other challenge facing a live-service shooter. Content output, technical quality, progression, identity, competition, and player retention still matter.
XDefiant should not be treated as proof that open matchmaking fails. It should be treated as proof that matchmaking policy is only one part of a multiplayer game’s survival. Players may dislike SBMM without automatically committing years to every game that removes it.
Full Disclosure Could Damage the System
Complete transparency has real drawbacks. Publishing the exact formula behind a skill rating could encourage reverse engineering. Players might discover which actions produce the largest rating gains and change their behavior around the model rather than the objective. Smurfs could manipulate placement more efficiently. Boosting services could identify the fastest methods for distorting accounts.
Developers also adjust matchmaking constantly. A public formula could become outdated after one update, creating more confusion rather than less. The answer is not to publish every coefficient.
Publishers can reveal the categories being measured, the relative priorities of major factors, the acceptable connection thresholds, the widening process during long queues, the treatment of premade groups, and whether recent performance receives extra weight.
They can show lobby skill ranges after matches. They can publish predicted win probabilities. They can separate ranked ratings from casual protection systems. They can explain whether teams were formed before or after the lobby was assembled. That level of disclosure would still protect the system from easy manipulation.
Players Need Receipts, Not Reassurance
The standard public-relations response to matchmaking complaints is no longer enough. Players have heard that connection remains a priority. They have heard that matches are intended to be fair. They have heard that developers are listening.
Those statements do not explain why a solo player faced a coordinated squad, why a Gold competitor entered a lobby filled with Diamond players, or why an evening of strong matches appeared to trigger a dramatic jump in opponent quality. Useful transparency requires visible evidence.
A post-match screen could show the estimated skill range, party composition, average latency, search duration, and which matchmaking restrictions expanded while the player waited. Ranked modes could show the rating spread before the match begins. Casual modes could identify whether they use open, protected, or tightly skill-matched pools.
Players do not need the source code. They need enough information to compare the system’s stated purpose with what actually happened in their match. Without that evidence, every losing streak feels suspicious, every dominant victory feels artificial, and every publisher explanation arrives after trust has already been spent.