Natasha Miller, Elaine Gusella, Chelsea Howe, Osama Dorias
What do you do when carefully planned social mechanics fail to land with your player base? What invisible frictions and unmet anxieties cause players to stall out in their social interactions? This paper covers common and subtle barriers in social games and a variety of techniques and tools across genres to address them.
Games are powerful engines to foster social interactions, community building, and player connection.They offer shared spaces and create opportunities for players to build trust, communicate, and support each other. In fact, in a recent Deloitte survey, 40% of Gen Z and Millennials said they socialize more in video games than in the physical world4 representing the degrading social capital of modern day9, and 50% said they believe games are “meaningful replacements” for in-person experiences.4 Yet many players struggle to find and maintain relationships with other players despite the desire to play games socially.2 Specifically, they fail to go beyond the transactional relationship to find patterns of repeated, meaningful interactions with others in games.
This matters because repeated, positive interactions are the foundation of long-lasting communities. When new users have a positive social experience, they are 6x more likely to return3 and half of Gen Zers say they join game sessions simply to hang out, even without playing the main game.5 When barriers impede these connections, players are left feeling isolated and overlooked, putting the the at risk of losing momentum, long-term satisfaction and engagement. Games that fail to support meaningful interactions see more solo play (including fully isolated experiences and ‘parallel play’), leading to increased churn and decreased LTV.1
This correlation between social engagement and key metrics is well proven. Players that feel acknowledged, supported, and safe are more likely to return to the game, play longer, and spend more.1 Social features are often the strongest drivers of engagement and retention across genres, with the top 2 highest grossing mobile games each year for the last 5 years having significant social features.6 Additionally, a 2023 Unity report noted that 34% of players claimed to stop playing due to lack of players/friends playing (i.e. not having a sense of belonging in the game community).7
Reducing social barriers and building stronger social bonds by viewing social game design as building social infrastructure is to everyone’s benefit. By considering both the spaces and dynamics within them, we can create more welcoming social environments that players will want to engage with longer and be more willing to spend money to be a part of the community.8
This paper focuses on lowering barriers to repeat (and advance) interactions in established social games. We encourage readers to investigate these tangential topics: Friendship/Trust Building, Kind Games, Player Impact, Toolkit for Cringe.
It should also be noted that this is not an exhaustive list of barriers, solutions, or examples, but rather those most frequently encountered.
Side note: We also didn’t go directly into mitigating disruptive behaviors in this paper, but this should also be a factor that is thought about in this space. Check out these resources for more information on Addressing Disruptive Behavior, Mitigating Intra-Team Conflict, and Disruption and Harms in Online Gaming Framework.
Social barriers can come from the player (social anxieties, skillset & abilities) or the game (systems, affordances, incentive structures), with some barriers crossing both (mismatched expectations, past experiences).
For each barrier, we’ll cover different contexts and sub-types, then provide solutions with specific examples.
One of the most common reasons players hesitate to reach out to others is fear of rejection. Even if they enjoyed playing with someone and potentially had a positive game outcome, they worry the other player might not feel the same and that their invitation for additional interaction or deepening relationship might be ignored or declined. The possibility alone is enough to disincentivize a large portion of players from taking the next step.
Some games attempted to address this barrier with features like Overwatch 1’s “Stay as Team” functionality. While successful in some cases, players still worried about everyone else saying no. It also created pressure if a player only enjoyed playing with one or some teammates, since it is an all-or-nothing proposition.
A more effective solution could be to create low-stakes, individualized signals. Instead of forcing players into high-commitment choices, we can give them lightweight ways to show interest. This could manifest in multiple ways:
What if someone is specifically looking to advance or deepen their relationship? Going directly from strangers to adding friends is a considerable jump in trust. Adding intermediate steps that are invisible and don’t require direct reciprocation or response can assuage fear of rejection.
For example: after a match or encounter, players can ‘star’ the people they enjoyed playing with. This can weigh chances for future matchmaking and provide additional opportunities for mutual play without the social risk of asking to be friends.
Instead of requiring a player to reach out, surface what other players are doing in case of mutual interests — like Discord presence letting users see which games people in shared servers are playing (Battle.Net has a similar system). This same technique could be used within a game on something like a preferred (starred) player list, where players could hop into matches, battles, or areas of a map where a preferred player currently is.
Some players worry about being teamed up with lower-skill players who might hold them back. This generally stems from not having a chance to gauge the players’ personality or skillset before being thrown into a high-trust environment. Without that baseline knowledge, players fear being dragged down, blamed for mistakes, or being stuck in a frustrating match.
This lack of trust and coordination is often the result of skipping the onramp. Players are made to cooperate in challenging situations before they’ve had a chance to gauge each other’s personality or skill set in low-stake environments. When players get the first impression of their teammates in a high-stress situation—like in the middle of an important raid, ranked match, or in a competitive event—there is little room for learning about each other, forgiveness, or building confidence together.
As a result, many players opt out altogether, choosing to play solo or only with pre-selected groups instead of risking a poor experience with strangers. This limits the potential for organic connections and narrows the pool of players who are willing to engage with the game’s social content.
There are many methods available to make player ability clearer.
It’s important to note that, especially in competitive games, skill alone is rarely a good indicator of team compatibility. Attitude towards the match (rank-focused, meta-driven, casual, etc.) and role on the team (tank, healer, DPS, etc.) matter considerably more and will be discussed under Mismatched Expectations.
For example, in Naraka: Bladepoint, the developers used several metrics to determine player goals and playstyles. They then integrated this information into the matchmaking system so that more like-minded players would be matched together. Across all ranks, players were more likely to stay with their teammates after this implementation.11
Some players have an internal worry that they aren’t good enough and will hold the team back. This is generally caused by a lack of confidence, and it can be just as limiting as fear of rejection, keeping players from reaching out to others, joining groups, or queuing for cooperative modes at all.
In many cases, this fear is unfounded.10 While skill gaps exist and a less experienced player may struggle to keep up with demanding content, it does not have to be a dead end. These situations can become opportunities for growth and connection. A team that chooses to bring in a less experienced player might still succeed, and in doing so, help a player ramp up in a positive and supportive environment. The result is not only skill mastery for the newcomer, but also a stronger sense of camaraderie for the group.
However, without systems that frame these moments of asymmetric skill as positive, the default experience is often shame or frustration. Players who feel they are a burden will avoid social interaction, and groups that encounter struggling teammates may resent the mismatch rather than embrace it. This barrier blocks not only individual confidence but also the chance for teams to build trust and connection alongside training and skill mastery.
There are many ways to signal to a community that skill building is a positive and expected part of the overall social ecosystem, and that low skill will not be judged negatively.
By providing players with ways to safely practice and improve their skills, they can reach a level of mastery that will help decrease fear of judgment.
This can be as explicit as Super Smash Brothers’ training mode—which lets players try out new moves and practice over and over again without any time, resource, or social pressure—or it could be implicit like environmental challenges in loading areas where players can both show off and build skill.
This fear refers to the inability to escape a situation—or the inability to escape it gracefully and in a socially acceptable way. For example, players may worry about starting conversations with players because they don’t know how much or how long the other player will talk, and they have no means to remove themselves from a long conversation without causing offense.
Previously mentioned solutions, like the preferred playlist, can also help with fear of entrapment by giving players lower stakes forms of engagement without commitment. However, we’ve identified a couple other solutions as well.
A ‘Soft No’ is a way to indicate some resistance to a social ask without a hard rejection. It lowers the perceived offense and lets players feel more comfortable moving fluidly out of social situations.
Note that by design soft nos leave flexibility and room for interpretation, and some players may miss these signals entirely. For that reason, it’s often beneficial to include plausible deniability mechanics as well as a failsafe.
Plausible Deniability refers to ways a player can be absolved of the social ramifications of an action. For example, in Genshin Impact players have 10 seconds to accept a join request. Because this is such a short time frame, some players miss the notification, or are in the midst of a battle when it appears. The person who requests to join isn’t told whether they were denied or if the request simply expired. The recipient is absolved of guilt because they may have been busy, or AFK and simply didn’t see the request.
In Sky, walking a certain distance from a group can cause players to switch which instance they’re part of, and players are frequently shuffled between instances by the game itself. This gives fantastic plausible deniability for any player who wants to leave a group, as it’s ambiguous whether they intentionally left or simply got moved to a different instance by the system.
A similar solution could be implemented in higher skill games like WoW that adjust instancing/phases by giving the player the ability to trigger something that looks like a phase shift.
Sometimes players have mismatched expectations which leads to lack of social interaction or negative social interactions. These mismatched expectations can take the form of a mismatch on how intense or casually players want to play, if they want to play to the meta or not, which communication methods they feel comfortable using, how much experience they expect the other person to have, and more.
While some solutions to Fear of Being Held Back apply, there are many things a game can proactively do to address this specific barrier.
The best option to resolve mismatched expectations is to make expectations more obvious or explicit.
Prior solutions relating to soft nos and plausible deniability also help with exit strategies if mismatched expectations still occur.
Some players don’t interact with each other because there isn’t a need or push to interact. While social incentives could be a paper in and of itself, we wanted to cover a few common categories of incentives and associated, desirable social interactions.
Information Sharing as Altruism
Though counterintuitive, making information less explicit or clear in game gives experienced players a benefit of having accumulated knowledge or share-worthy knowledge.
This happens unintentionally in many games, where the wiki or Discord community become a thriving locus of activity due to lack of in-game clarity. However, it can also be an intentional design choice, like in Sky, to create value in tribal knowledge and meaning in exchanges of that knowledge.
All of these experiences are united not by explicit incentivization but by having opportunities for altruism that players find intrinsically incentivizing.
Information Sharing as Survival
In higher skill or higher trust games, sharing information and coordinating actions becomes a matter of survival or success.
There are many ways to provide progression feedback for social interactions.
Many games incentivize social interaction passively—like games that provide resource or experience buffs for being near other players, e.g. Glitch, Diablo IV, but remember that active social interaction and meaningful exchanges have bigger impacts on player experience and game metrics.
To that end, how could games ensure there is some form of interaction beyond parallel play and co-presence?
Avoiding Dark Patterns: some solo-focused players will find any social interaction incentives offensive. These techniques work best in established social games, or as elements that are only surfaced if players have explicitly opted into social already.
When players don’t understand what’s currently going on in a situation or how they fit into it, they are less inclined to engage, and this applies to any scale of social interaction.
This can happen, for example, whether a player is unsure if another player received their request, what the socially acceptable next steps are, or if a player’s guild is active / what their guild’s goal is.
If certain states are unclear, players experience social ambiguity, which suppresses participation. Think of social interactions as a state machine. These states can either be clearly defined, implied, or left up to interpretation. In social situations, players don’t just seek to understand what they can do, they want to understand what’s socially allowed.
If these states remain unclear, social pressure may lead players to lurk indefinitely, exit the system entirely, or limit themselves to interactions they perceive as “safe” (ex: only interacting with people they know or know well.)
By making the status of the player and others more visible, it can give them the confidence or foundation to interact.
Note that tagging systems mentioned under Mismatched Expectations are also a form of status visibility when used transiently.
Some players don’t know what the best next step is. By having contextual, relevant recommendations in the moment, you can both help shape player norms and encourage specific social interaction paths.
By making social structures more visible in game, players can understand how they fit both into the game world and with the other players they have chosen to have in their social circle.
For example, Helldivers 2 allows players to understand how their contributions relate to the bigger picture by giving visibility into the ‘war effort’ via the galactic map. Players can see how doing missions in certain zones would push up the percentage for that zone globally. In this way, players can choose missions at their level in zones that will have the biggest global percentage impact for their team. This allows the player to see how working in small groups contributes to the greater good, giving small social groups purpose and recognition.
For Honor had three factions (samurai, knight, and viking) that players could join to help the group earn rewards, but players would jump around frequently as there weren’t social structures in place to allow players to feel tribalism with their chosen faction. As a result, factions felt abstract and interchangeable.
When rewards are the primary motivation for a decision, players will make the decision that will bring them the better reward. That’s why players followed whichever faction was winning that season.
Players who belong to clans that are themed after one of the factions are far more likely to remain with that faction regardless of the outcome. Once a player is part of a Samurai-themed clan, “being Samurai” becomes part of their social identity. Leaving the faction is no longer a mechanical choice, but an emotional one as well.
This demonstrates a key principle: Players are more likely to stay loyal to other people, not abstract banners. Emergent clans reveal the missing social layer that the system needed to work as intended.
Sometimes, players simply don’t have enough ways to interact with each other in a manner that they are comfortable with using.
While simply adding features is always a delicate balance, many games have had success targeting multiple play styles, whether that means the number of people players want to engage with, the means of communication they want to use, the time or commitment level required, or the level of effort and skill required.
We recommend analyzing the systems or features in your game meant to drive community and belonging and broadening it as much as is feasible. While this will change both from game to game and for the stage a player is at in your game, e.g. early gamers will be looking more for guidance and later stage gamers will be looking for respect, we recommend starting with the communication styles available as this is often one of the main ways players are able to socialize.13
Apex Legends has multiple ways for players to communicate (text, voice, ping systems) that are all acceptable by the community therefore allowing players to choose the one they are most comfortable with.This not only improved the social capabilities of players needing more ways to feel socially comfortable interacting, but also improved accessibility for players that needed to use different communication styles due to physical limitations.
Sky’s Meditation area vs Festival area12 gave players multiple ways of enjoying a festival in game from exciting celebratory environments to meditation areas to take a mental break.
The opposite of social incentives, it’s equally important to ensure you are not unintentionally disincentivizing social interaction. Sometimes games can implicitly punish players for playing together by requiring them to compete for loot, gear, objectives, and so on, or by not accounting for team dynamics that require moments for discussion, strategy, and norm forming.
D3 drops personal loot for each player in the rift that only they can see and pick up so players aren’t competing for the loot even if they are the same class or gear spec. However, they do allow trading between players so that they can still cooperate and share if wanted.
On the casual spectrum, key items picked up in a host’s island by a client character in HKIA will still go to the host, but the client can have the fun of running around and helping gather.
Design spaces and UX flows to funnel players into social areas. For example, in Helldivers, rather than having multiple objectives on the map that players have free will to choose to go to, they funnel players to specific areas and objectives to complete together.
Designing for optimal social interactions is a recipe for players to fail or miss expectations. Instead, when balancing and tuning, aim for imperfect coordination and skill matches.
For example, Destiny balances 6 player raids with its “success” at 4 active member participation to allow for some struggle in the team dynamics and opportunities for norm formation. In LoL Summoner’s Rift, 3 or 4 players anchor the experience despite it being a 5 person team. ARAM (single lane gameplay) took precedence because it has more flexibility for team dynamics leading to more success over time.
While barriers may manifest in different ways, the vast majority of solutions fall into the following overarching categories: adding ‘gray’ areas, adding signaling, and adding triggers.
When we talk about adding gray areas, we mean adding intermediary, lower risk, and transitional stages between existing interactions. Essentially, any time there is a large gap in trust, persistence, or commitment between two social states (for instance, going from stranger to friends list), try adding an in between step (for instance, the ‘preferred’ player list).
This can also apply to states of interaction where players desire ambiguity or plausible deniability for a potential social slight (think of texting a friend to call you so you can get out of a conversation without it being ‘your fault’). These gray areas and in betweens resolve a variety of social anxieties.
Finally, it can also refer to smoothing ramps in difficulty more generally, such as having mentorship, training, and onboarding mechanics that help players embed more fully into a social ecosystem.
Signaling refers to mechanics that make the player state clear and eliminate ambiguity around player needs. Tagging systems that let you showcase, for example, whether you want to play to the meta, be in voice chat, or casually grind help connect players with like interest and prevent connecting with mismatched expectations. This can also apply to things like Discord’s notes section, where you can leave annotations about your experience with other community members only for yourself.
The style of signal is important as well, and works best when there are gentler signals (headphones to indicate busy, a ‘later’ option to a match request) vs hard rejections.
Triggers refer to anything that affords, implies, or initiates social interaction. This can be as subtle as a timed resource spawn point, which will de facto bring players together at a certain time, or as explicit as a global event like an in-game concert. It can be player initiated (world pings, party invites) or game initiated (events and resources).
It also refers to the accessibility of social actions, like surfacing contextual social animations in response to a given situation, to make social interaction as easy, seamless, and main path as possible. Adding a prompt to thank someone in response to a gift or match is a guaranteed method of improving adoption.
None of these are perfect solutions on their own, there will likely need to be a combination of them or even ones we haven’t thought of for each individual game. However, the goal is to have provided some insights into the barriers your players may be facing preventing them from experiencing a sense of belonging, being social, and finding community in your game. If you are able to alleviate some of these barriers so that players are free to socialize, make friends, and gain respect, your game will be able to build a long-lasting community.13