A practical clinician guide to video game therapy, including Minecraft, Roblox, telehealth sessions, gaming groups, consent, privacy, and therapeutic intent.
Video games are increasingly entering therapy rooms, telehealth sessions, and group programs — especially in work with autistic, ADHD, and other neurodivergent young people. Used well, they can support rapport, shared attention, emotional regulation, social negotiation, flexibility, communication, and therapeutic observation. Used poorly, they become unstructured play, a reward system, or an ethical headache.
This guide is for clinicians who want to use video games intentionally rather than casually. It covers face-to-face setup, telehealth gaming, browser games, Roblox, Minecraft, therapeutic groups, consent, confidentiality, IT barriers, and the clinical question that matters most: what makes it therapy?
Jump to a section:
I am writing this from the overlap of research, clinical training, and practice development. My own systematic review examined Minecraft-specific therapeutic interventions for autistic young people, and I later used that knowledge, alongside clinical training and community-building skills, to help develop a Minecraft-related social connection group with support from The Kidd Clinic. That experience shaped the central argument of this guide: games are not therapy by themselves, but they can become powerful therapeutic media when held with clinical intent, ethical boundaries, and proper facilitation.[1]
Video games can reduce the social demand of eye contact, provide a shared activity that anchors attention without forcing it, and create real-time opportunities for decision-making, frustration tolerance, collaboration, boundary-setting, flexibility, and repair. These are not abstract therapeutic targets. They happen inside the play itself: when a young person loses, restarts, builds, negotiates, destroys, repairs, avoids, insists, collaborates, or asks for help.
For autistic and ADHD young people in particular, game-based work may lower the demand of traditional face-to-face conversation, support multiple communication modes, and offer a sense of agency that talk-based therapy can struggle to provide.[2][3][4] Voice, text, movement, shared attention, avatar behaviour, world-building, and parallel play can all become clinically meaningful channels of communication.
Research in this area is still developing, but it is promising. Zayeni, Raynaud and Revet's systematic review found that both commercial and purpose-built video games have been used as therapeutic or preventive tools in child and adolescent psychiatry, with evidence across a range of presentations, while also emphasising the need for stronger study designs.[2] Jiménez-Muñoz and colleagues reviewed video game-based interventions for autistic people and reported improvements across cognitive, social, emotional-recognition, and motor domains, alongside high engagement and completion rates in many studies.[3] Therapeutically applied Minecraft groups have also been described as a way to support social engagement, confidence, and communication for neurodivergent young people.[4][5]
The mental health context matters. Autistic young people experience high rates of co-occurring anxiety and other mental health difficulties. Reviews report particularly high rates of specific phobia, social anxiety, generalised anxiety, and separation anxiety among autistic children and adolescents.[6][7] For some young people, a game-based medium may offer a more tolerable entry point into therapeutic work than sitting opposite an adult and being expected to "talk about feelings" on demand.
None of this means every session should be a gaming session, or that games replace therapeutic skill. The game is not a shortcut around clinical formulation. But when a young person walks into your room and the thing they most want to do is play Minecraft with you, that is not automatically avoidance. It may be the most direct route into the work.
This guide is written for psychologists, counsellors, occupational therapists, social workers, school clinicians, and allied health professionals who are considering video game-assisted therapy with children, adolescents, or neurodivergent clients.
It is not a claim that video games are a standalone treatment, or that every clinician should use gaming in sessions. The aim is more practical: if games are already part of a young person's world, how can clinicians think about them ethically, clinically, and intentionally?
The game is not the therapy. The game is the medium.
The therapy happens in the clinical noticing:
"I noticed you gave that player all your resources. What was happening there?"
It happens in modelling:
"I'm getting frustrated that I keep dying here. I'm going to pause, breathe, and try again."
It happens in co-regulation: sitting beside the young person while they navigate something difficult without immediately rescuing or correcting them.
It happens in reflection:
"What was it like building that together? Did it feel different from building alone?"
It happens in formulation:
"When the rules changed suddenly, your whole body seemed to go into threat mode. I wonder whether that happens outside Minecraft too."
It also happens in the choice of game, the boundaries around the platform, the structure of the session, the consent process, and the conversations with families about what gaming is for and what it is not.
Video game-assisted therapy is not babysitting. It is not "letting them play because they will not talk". It is not a reward for compliance. It is a legitimate therapeutic medium when used with clinical intention, ethical care, and appropriate formulation.[2][8][9]
It just also requires knowing what Bedrock Edition is.
A video game-assisted session does not need to be complicated. A simple structure might look like this:
The aim is not to turn every moment of play into therapy-speak. The aim is to stay clinically awake while the young person is engaged in something meaningful.
The simplest version is a Nintendo Switch connected to a television in the therapy room. A big screen matters. It makes the game visible, shared, and part of the relational space rather than something private happening on a small device. The therapist can sit alongside the young person, controller in hand, and play together.
In that setup, you are not "just playing". You are co-regulating, narrating, observing, modelling, and joining. You might notice how the young person responds to losing, how they invite or reject help, how they manage turn-taking, whether they tolerate ambiguity, how they communicate frustration, or what kind of world they build when they are given control.
iPads can also work well, especially for Minecraft therapy. Bedrock Edition runs natively on iPad and supports multiplayer options that can work well in clinics, schools, and group settings when the network allows it.[10] Two iPads side by side can sometimes be enough: one device hosts, the other joins. It is portable, familiar, and does not require a television.
The key decision is what hardware your practice can realistically support. Consider whether you have:
A practical starter kit is a Nintendo Switch with a dock and HDMI cable, two controllers, and a copy of Minecraft. That gives you face-to-face co-play on a large screen with minimal IT overhead.
Online video game therapy expanded during COVID-19, and many clinicians found it worked better than expected — particularly for young people who were already more comfortable online than in a waiting room.[8]
The setup is usually split-attention rather than truly split-screen: the video call remains open while the client and clinician join the same game, server, world, or browser-based activity. The young person plays on their own device, the clinician plays on theirs, and the therapeutic conversation occurs around what is happening in the shared activity.
This is where platform choice becomes a clinical and ethical decision.
A game is not just a game. It may involve accounts, usernames, friend lists, chat functions, public profiles, private servers, moderation systems, age settings, screen recordings, direct messages, and data collection. Those features can have confidentiality and boundary implications.[11][12]
Browser games are useful because many require no accounts, no downloads, no installations, and no friend lists. You open a browser, choose a game, and play. For a telehealth session where you want a quick shared activity — a racing game, puzzle game, drawing game, or cooperative challenge — browser-based platforms are often the lowest-friction option.
The therapeutic value is not usually in the game itself. It is in the shared experience: the laughter, the narration, the competitive moment, the frustration, the repair after someone loses, or the way a young person explains the rules to you.
Browser games are especially useful for:
The trade-off is depth. Browser games tend to be simple and disposable. They rarely offer the sustained identity, world-building, collaboration, and project-based work that games like Minecraft or Roblox can provide.
Roblox is where many young people already spend time. It is a large platform of user-created games, social spaces, avatar customisation, group identity, and peer culture. For some children, asking them to show you Roblox is like asking them to show you their neighbourhood. It can reveal what they value, where they feel competent, how they relate to peers, and what social rules they are navigating.
But Roblox requires careful ethical handling.
Roblox accounts can involve usernames, profiles, connections, chat settings, private servers, content maturity settings, spending controls, and visibility settings.[11][12] If a therapist creates a Roblox account and connects with a client, the therapeutic relationship may become partly visible through platform infrastructure. Depending on settings, a client may see the therapist's username, profile, avatar, connections, or activity. Other clients could potentially identify each other if the therapist uses the same professional account across multiple clients.
There are practical mitigations, but none remove the issue entirely.
A safer approach is to use a dedicated professional account that is separate from any personal gaming. The account should not contain personal information, should not use the therapist's private username, and should have privacy settings reviewed regularly. Clinicians should also consider whether a private server, supervised in-session play, or family-managed account connection is more appropriate than a general friend connection.[11][12]
The friend list is where the ethical complications stop being abstract.
On Roblox, a friend connection is not a private clinical arrangement. Depending on each user's privacy settings, friends can see who else is on your friend list, what experiences you have recently played, when you are online, and sometimes which server you are currently in. If a clinician uses one professional account across multiple young clients and accepts friend requests from each of them, those clients may be able to see each other's usernames on the clinician's friend list. That is a confidentiality breach the clinician may not even realise has occurred.
It gets more uncomfortable. A client could send the clinician a friend request from a personal account at 9pm on a Sunday. Another client could try to join the clinician's server outside session. A sibling or parent could log in under the young person's account and see the connection. A former client may remain on the friend list for years unless actively removed. None of these are theoretical.
Practical safeguards include:
The principle is the one you would apply to any other platform: if you would not let two of your clients see each other's names in a waiting room, you should not let them see each other's usernames on your friend list.
Roblox can be clinically rich precisely because it is socially real — but that also means it carries more boundary, privacy, and peer-exposure risk than a closed, clinic-controlled activity.
Before using Roblox in video game-assisted therapy, families should understand:
This should be documented in the consent process. The ethical principles are not new: informed consent, transparency, confidentiality, boundaries, scope of practice, and clear management of between-session contact. The platform is new. The principles are not.
For a broader look at how modern gaming platforms are designed to influence behaviour, including the monetisation and social dynamics young people encounter, see Gaming Microtransactions, Loot Boxes, and Battle Passes in Australia.
Minecraft therapy is probably the most therapeutically versatile application of a commercial game currently available for clinical use. It supports open-ended creative play, collaborative building, survival challenges, problem-solving, planning, emotional regulation, social negotiation, and family work. It can be used in individual therapy, dyadic work, parent-child sessions, sibling sessions, family therapy, and small groups.
For neurodivergent young people, Minecraft can offer a lower-demand social environment with multiple communication modes and a high degree of personal control.[4][5] It also has cultural reach. Minecraft has been one of the most widely played games in the world, with Microsoft reporting 126 million monthly active players in 2020.[13]
The evidence base is still small but growing. Minecraft-specific therapeutic literature includes group-based interventions, parent and clinician perspectives, family therapy applications, and exploratory uses of Minecraft for social communication, anxiety exposure, emotion recognition, and trauma-related work.[4][5][14][15][16]
My own systematic review of Minecraft-specific therapeutic interventions for autistic young people found only a small number of eligible studies from a much larger initial pool.[1] That pattern is important: Minecraft is clinically promising, but the research base remains early-stage. Most studies are small, pilot-based, qualitative, uncontrolled, or proof-of-concept. The strongest claim we can make is not that Minecraft therapy "works" in a generalised evidence-based-treatment sense. It is that Minecraft appears to be a clinically useful, engaging, and feasible medium that warrants more rigorous study.
One study often worth highlighting is Gerhardt and Smith's case study, which explored how the narrative component of trauma-focused CBT was adapted using Minecraft for an 11-year-old child with autism spectrum disorder.[15] This is clinically important not because it proves Minecraft is a trauma treatment, but because it shows how an established intervention can be adapted through a medium that makes narrative, distance, control, and engagement more accessible for some autistic children.
One consistent finding across the literature is engagement. Young people often want to attend, participate, and return.[3][4][5] That matters clinically, especially for populations who may have had repeated experiences of therapy as socially demanding, adult-led, compliance-focused, or irrelevant.
But Minecraft also has a version problem every therapist needs to understand.
There are two main editions: Java and Bedrock. Java runs primarily on PC and Mac. Bedrock runs on consoles, iPads, phones, and Windows devices. They are different editions with different multiplayer ecosystems. Minecraft's own guidance makes clear that players need to launch the edition their friends are using; Java and Bedrock are not simply interchangeable in the same multiplayer space.[17]
For individual sessions, this is manageable. You match your setup to the client's platform.
For groups, it becomes a real barrier.
If one participant has Java and another has Bedrock, they generally cannot simply join the same LAN world. There are technical workarounds, such as GeyserMC, which allows Bedrock clients to connect to Java servers, but this requires server infrastructure, configuration, maintenance, and technical confidence that many therapy practices do not have.[18]
This is why platform standardisation matters. If you are running Minecraft therapy groups, decide early whether you are a Java practice or a Bedrock practice, and communicate that to families before enrolment. For face-to-face groups, iPads running Bedrock are often the simplest option: they are portable, familiar, and easier to standardise across participants.
Therapeutic gaming groups can be incredibly rewarding. They are also logistically brutal.
The technical layer alone is significant. You need compatible devices, a functioning network, a game that supports the right number of simultaneous players, and a plan for when something goes wrong. Something will go wrong. A device will crash. Someone will lose connection. A world will fail to load. One player will update and another will not. A controller will stop pairing. A child will forget their login. The Wi-Fi will block multiplayer.
The social layer is harder.
Games have rules, and young people have strong preferences about those rules. In Minecraft, conflict may emerge around Creative versus Survival mode, whether PvP is allowed, whether someone can break another person's build, how resources are shared, who gets to lead, who is "cheating", and what counts as fair. In Roblox, it may be which experience to play, who chooses, who follows, who leaves, and whether the group stays together.
These moments are therapeutically rich. They contain negotiation, compromise, autonomy, rigidity, fairness, frustration, repair, leadership, rejection, and belonging. But they need active facilitation. Without a clinician who understands both the game mechanics and the group dynamics, the game can amplify dysregulation as easily as it supports connection.
Practical recommendations for therapeutic gaming groups:
The game should not be treated as a reward at the end of group. It is the group medium. That means it needs structure, containment, boundaries, and clinical purpose.
A lot of clinicians considering this work are genuinely uneasy about it, and they are right to take that seriously.
If you do not have gaming as a hobby, the prospect of running a session inside Minecraft, Roblox, or any other unfamiliar platform can feel exposing. There is the fear of looking incompetent in front of a young person who is fluent in something you are not. There is the fear of getting platform settings wrong and creating a privacy problem. There is the fear of being judged by colleagues for "just playing games" in session. And there is the older, quieter fear that gaming is somehow not serious clinical work — a fear that often comes from clinicians' own upbringing rather than from the evidence.
Naming this matters because the fear changes the work. A clinician who is anxious about the medium tends to over-direct, over-talk, fill silences, and miss what the young person is actually showing them. The fix is not to fake fluency. The fix is to be honest about where you are and to design your approach around it.
Start small and ramp up. You do not need to begin with a complex sandbox multiplayer build. Start with a browser-based co-op puzzle game where the stakes are low and the rules take ninety seconds to learn. Move to a single-player game you and the young person take turns with. Move to Minecraft Creative mode in a private world with no survival pressure. Move to Minecraft Survival, then to a shared multiplayer build, then to a group format if that is where the work is going. Each step gives you more familiarity with the medium and more clinical confidence in what to notice.
Take a back seat sometimes. You do not always have to play. One of the most underrated formats in this work is the clinician who watches, asks, and learns. A young person showing you their favourite Roblox experience, walking you through their Minecraft world, or narrating a game they are playing solo is doing something clinically rich. You are getting access to their inner world: what they value, what they build, who they imagine themselves to be, what makes them proud, what makes them rage-quit, who they invite in, who they exclude.
Use the time for structured observation and brief measures. While the young person plays, you have a rare clinical window. They are regulated enough to engage, distracted enough to lower their guard, and showing you behaviour you would never see in a sit-down interview. This is also a practical opportunity to complete brief self-report measures together. A young person who finds questionnaires aversive in a formal assessment may be willing to work through the SCAS (Spence Children's Anxiety Scale) or RCADS (Revised Child Anxiety and Depression Scale) while their world loads, while they wait for a respawn, or as a structured pause between builds. The play becomes the regulating frame for the assessment, not a distraction from it.
Be honest about not knowing. "I have never played this — show me how it works" is not a clinical weakness. It is a relational invitation. Young people are often given very few opportunities to be the expert in a room with an adult. Done genuinely, this can shift the entire power dynamic of the session in a useful way.
Get your own reps in. If you are going to do this work seriously, spend some time with the games outside session. Not to become a gamer, but to know enough about the platform that you can hold the clinical frame when things go sideways.
The skill is not gaming. The skill is clinical attention inside an unfamiliar medium. Most clinicians already have the second one. The first one is just practice.
Video game-assisted therapy is not appropriate for every client, every goal, or every session. Clinicians should be cautious when gaming in therapy is likely to intensify dysregulation, reinforce avoidance, create conflict with caregivers, blur online boundaries, or become the only tolerated form of engagement.
It may be worth pausing or modifying the approach when:
The question is not "does this young person like games?" The question is whether this game, in this format, with this clinician, is serving the formulation.
Some of the most predictable problems in game-based work are not about whether the young person likes games. They are about the clinician underestimating the medium.
When these mistakes happen, the problem is usually not the game. It is a mismatch between clinical intention, platform knowledge, and the practical setup needed to hold the work safely. Clinicians using video games in therapy are accountable for both the clinical frame and the platform they have chosen to work inside.
If you work in a larger practice, school, university clinic, or health service, your network may not support therapeutic gaming without IT involvement.
Many managed Wi-Fi networks restrict device-to-device communication. Some block game traffic. Some prevent personal devices from joining the same network as staff devices. Some restrict app installation. Multiplayer may require bandwidth, compatible versions, firewall permissions, and sometimes specific network settings.[10][19]
Before you plan a gaming therapy program, check:
This is boring, but clinically important. Getting it wrong means discovering in front of a client that the game will not connect. That can be frustrating, dysregulating, and avoidable.
Therapy rooms are not gaming lounges, and the practical mismatch matters more than most clinicians anticipate.
Noise. Games are loud. Music, sound effects, voice chat, frustrated vocalisations from young people who have just lost a build, and the clinician's own commentary all carry through standard plasterboard walls. If the room next door is doing trauma processing, couples work, or a sensitive assessment, gaming sessions can be genuinely disruptive. Solutions are simple but need to be planned: a quiet room, in-ear headphones for the young person with a splitter so the clinician can hear too, sound dampening, scheduling gaming sessions when neighbouring rooms are empty, or running gaming work from a dedicated space.
Cost. A practical starter kit is not free. A Nintendo Switch with a second controller, a TV or large monitor, HDMI cabling, and a copy of Minecraft sits in the hundreds rather than the thousands of dollars, but iPads, Roblox spending controls, additional accounts, and any future hardware refreshes add up. Group setups multiply this. Practices need to decide whether gaming infrastructure is a clinical asset they invest in deliberately, or an ad-hoc cost that gets quietly absorbed by individual clinicians.
Lag. Telehealth gaming lives and dies on connection quality. Even modest lag can turn a regulated session into a frustrating one, particularly for young people whose distress tolerance for delay is already low. Lag in Minecraft means a block placed twice, a structure half-collapsed, or a player teleporting backwards. Lag in Roblox means dropped voice chat, frozen avatars, and the kind of micro-failure that escalates quickly. Before running telehealth gaming sessions, test the connection at the actual times of day you will be using it, on the actual devices, with the actual game. Wi-Fi performance at 2pm and 7pm can be very different things.
Updates and version drift. Games update. Sometimes mid-session. A young person who joined yesterday on version X may find their world will not load today on version Y. Group sessions are especially vulnerable: if one participant auto-updates and another does not, multiplayer breaks. Building a habit of checking versions at the start of every session is worth the thirty seconds it costs.
The benefits that survive all of this. When the setup works, the trade-offs are worth it. Sessions become genuinely shared. Young people who freeze in talk-based work become animated, expressive, and collaborative. Parents who have been told their child "doesn't engage" see them lead a clinician through a build for forty minutes. Group members who have never had a friendship outside school find themselves planning a co-op project for next week. None of this is sentimental. It is what good clinical engagement looks like when the medium matches the young person.
The evidence base for video game therapy — and Minecraft specifically — is promising but limited. Most studies use small samples, vary widely in intervention design, and lack large randomised controlled trials.[2][3][1]
The field needs:
It also needs neurodivergent leadership. "Nothing about us without us" applies here as much as anywhere else. Autistic and ADHD people should be involved not only as participants, but as collaborators shaping the research questions, intervention design, acceptability measures, and definitions of success.
The question is not simply whether games can make therapy more engaging. The deeper question is whether games can help therapy become more accessible, relational, flexible, and respectful for young people whose communication, regulation, and social participation do not always fit traditional clinical formats. For more on how neurodivergent practitioners and clients navigate systems not built for them, see Neurodivergent Clinicians: Navigating a Profession Not Built for Us.
If you are developing neurodiversity-affirming therapy groups, gaming-based social connection programs, or child and adolescent therapy resources, PsychVault is being built as a place to share practical tools that clinicians can actually use. Browse the resource library, or create a store if you have your own worksheets, group plans, psychoeducation handouts, or clinical templates to share.
For visual psychoeducation on topics covered in this post, I also share neurodiversity-affirming resources on TikTok: @ethonsmoth.
Share your thoughts and experiences with this resource.
Sign in to leave a comment
Move from strategy into implementation with templates, handouts, and psychoeducation tools already live on the marketplace.
Publish clinician-grade templates, build trust signals, and start growing an evergreen library under your own brand.
A practical guide to finding psychology resources that actually work in Australian practice — covering templates, handouts, NDIS tools, and what makes a clinical resource worth using.
A practical and honest guide for neurodivergent psychologists and allied health clinicians navigating systemic ableism, masking, burnout, and finding environments where they actually thrive.
Most provisional psychologists can tick the insurance declaration. Fewer can explain exactly what is covered, when to notify, and what happens across placements.
