As with every single competitive game out there, the developers of Apex Legends have to deal with player perception of what's wrong alongside the realities they see behind the scenes. A persistent thread in the Apex community's grousing has been about the game's matchmaking system, and specifically how it measures skill and forms squads and matches around it.
To enumerate the various issues players have raised would take forever, but one regular example is that skilled players sometimes feel that they're being deliberately matched with much less skilled players, and vice-versa, in a doomed attempt to create balanced matches that only results in blowouts.
This has been going on for basically as long as the game's been out, but developer Respawn has now posted what I can only describe as a comprehensive response to the community. This thing is absolutely enormous and goes into detail on how the game's matchmaking works, recent changes made, the problems that the developers acknowledge, and what's being done to fix them.
If you're an Apex head it's probably best to read the full thing (get a cup of tea first), but Respawn itself lists the takeaways thus:
- Apex Legends currently makes matches using your pre-made party’s best player’s skill rating, and our system does consider your pre-made squad’s size when matching you with opponents.
- We are working toward creating fairer matches compared to today as you play with friends of different skill levels.
- We are in the process of retiring our old Skill Based Matchmaking (SBMM) system for a new one that more accurately groups our players based on skill, and thus lets our matchmaking algorithm make better decisions when forming groups. The end goal is to create fairer matches and experiences that are more fun for all.
- We are continuously testing and iterating on our matchmaking systems in the live game to figure out what works best. In fact, many changes have already been tested and rolled out in various regions for Ranked and regular Pub games. And, we’re not done—you can expect more matchmaking refinements to come in the near future.
Buckets of data
Respawn says there are three elements it considers in matchmaking: the progression system that players see; a hidden skill rating; then the matching system. The progression system might seem a slightly odd fit there but Respawn says it affects player behaviour as well as player engagement and can "strongly affect their perception of how fair a match is". The studio also observes that account level is not correlated to skill and that this is another factor: "A low account player can be highly skilled, and (like most of us here…) a high level player can be a potato".
The skill rating is a simple numerical value, hidden from the player, that represents their skill based on "a number of factors". We then get a chart showing one way of viewing skill level distribution, the main takeaway from which is that the vast majority of players are grouped in the middle, with very few at either extreme. There is also what Respawn calls the "discrete buckets" approach, whereby based on these ratings players are put into five groupings (0-2, 2-4, and so on) which are "pretty wide" in non-ranked matchmaking, with special groupings for low-skilled and new players as they learn the game.
The matching system uses the skill ratings to group players together, obviously based on who's searching for a match at that particular time, and tries to strike a balance between forming a match quickly and taking time to more closely match players. Once a group is chosen comes another problem: how it forms squads and pairs them up.
"The new system is more granular, and has many more buckets".
Respawn
Respawn goes into detail about how the algorithm would try to form fair teams on this basis, and here's where we start to get some real meat. In non-ranked matches the game will form teams with the closest average skill, but in ranked matches it does something different:
"In a fully competitive environment, like Ranked, we value having teammates of equal skill, and a lower burden of “carry”, over a perfectly fair match. Note: while we are effectively stacking teams here, the set of players we are picking from is limited—we only pick from the players that made it into the match and are therefore similarly skilled. Because of the inherent randomness and chaos of a battle royale, we think this is the right tradeoff".
There's more on how the game computes a team rating, but the above outcome is what matters: and no doubt some players will feel that this acknowledgement shows they were right all along. Respawn further says that mixing premade squad sizes and the squads the game forms from randoms is "inevitable" simply because you're never going to get 60 players of equal skill in the same size of squad looking for a match.
The solution? Respawn has a new algorithm that it's going to implement in early 2023. The new system "will do a much better job at compensating for the competitive advantage that premade duos and trios have. We’ve done the science to quantify this advantage, and we’ve been experimenting with a new system that can continuously learn and update over time". The core is that this new method acknowledges the small advantages of being in premade duos or trios, and adds this to the skill measurement of a given team.
This new system will also replace the various different ways Apex Legends currently matchmakers across modes. Once it's rolled-out, "all matchmaking will use the same skill rating technology, but each mode will use different tuning values based on what works best for that mode".
And remember the skill buckets? They're not gone: We're just gonna have a lot more buckets. "The new system is more granular, and has many more buckets. This allows our matching algorithm to make better decisions when placing players into matches".
Respawn provides a few mathematical examples of how it believes this new system will measure skill in a more "meaningful" way, as well as being smarter about how it's using all those buckets in a different time frame (it will also predict which other buckets are coming and incorporate this in its calculations: quantum bucketry?). The studio says that live tests show that with this system "matches are tighter and wait times remain mostly unchanged" then goes on to prove it with a graph.
The old system will be phased out across regions to allow Respawn to monitor the new one's launch and deal with any unforeseen problems, and once the replacement is complete the matchmaking team's going to start looking even more closely at "additional corrections" to further account for the solo/duo/trio nature of premades.
Big news for the Apex Legends lovers, but as well as detailing the new system Respawn addresses some of the more regular community questions. On the topic of why higher-skilled players end up in some of your matches, it says this is down to dynamic adjustments in skill rating. This works both ways, so if you're on a losing streak your matches should gradually get "easier". Opponents "could happen to be having a bad day and on a loss-streak" if they're much higher ranked. Other factors are the skill distribution in a match (which the future changes will apparently "shrink") and the fact that more skilful players can bring weaker players into matches.
Does the game give good players bad teammates?
"No, we do not intentionally give good players bad teammates," says Respawn, before going on to explain that it kinda does. "If you are one of the stronger players in a particular match, you are indeed more likely to be paired up with players weaker than you [...] If you’re at the top end of skill distribution, no matter what players you end up being matched with, your teammates will almost always be worse than you. This is because the tail ends of the distributions are poorly populated and matchmaking in this zone requires incredibly long matchmaking times, incredibly large skill gaps, or some intermediate mix of the two". It further notes that "Higher skilled players tend to squad up, which worsens the issues".
Respawn acknowledges that this may feel like “big brother matchmaking” in effect, and the new system aims to lessen the impact with "tighter matches that will lessen these issues".
Finally, it addresses the question of whether the matchmaking system is built with the goal of optimising retention and engagement: That is, built to keep players playing Apex Legends for as long as possible. This unsurprisingly gets a firm "no" though the studio points out that one big difficulty it faces is in trying to make "the most fun matches" when "you can’t actually measure fun".
Respawn uses retention as a yardstick, reasoning that "players are more likely to stick around if they’re having fun". However it clarifies that "we never build an algorithm that is directly optimizing for retention (and definitely not engagement—convincing you to play an extra hour a day when you’d normally do other things isn’t good for us or you)".
The ongoing success of Apex Legends has undoubtedly transformed Respawn as a studio. Where so many others have tried and failed to get live service right, it's managed to attract and retain a huge playerbase with an ever-rolling stream of new content and ways to play. It's also changed how it interacts with the community, with Apex arguably one of the best-in-class examples of how to be as responsive and transparent as a big studio can be. This long and comprehensive account of the game's matchmaking principles, the problems inherent in them, and how the studio's working to improve it is exemplary communication. Even if, every time you lose a match, it'll still feel like the game's fault.
