Collaboration: The skill every organization wants, but almost none will invest in
- Justin Matheson
- Jan 12
- 5 min read
Collaboration is the skill every org claims to value, and then quietly punishes when deadlines hit.
A launch slips, a client escalates, someone “just takes it on” to get it done, and the team calls it efficiency.
Nope.
It is a coordination failure that got temporarily masked by one competent person burning extra fuel.
In my experience, most collaboration problems are not personality problems.
They are practice problems.
We put people in meetings to talk about teamwork, then we send them back into real work where the stakes are high, the feedback is vague, and the incentives reward heroic solo saves.
So nothing changes.
Today I want to make a simple argument.
If you want collaboration to improve, you need a practice environment that forces shared goals, interdependence, fast feedback, and low-stakes iteration.
Commercial video games (COTS) do that unusually well.
Not because they are “fun.”
Because they are full of constraints.
What collaboration skill development actually means
Collaboration is not just “being nice” or “communicating more.”
In practice, teams collaborate well when they can:
Build a shared picture of the problem
Coordinate roles and handoffs without constant supervision
Share information early, not after it is too late
Adapt when the plan breaks
Repair mistakes without blame spirals
If you read that list and think “yep, that is exactly where we fall apart,” you are not alone.
There is a growing research base showing game-based interventions can support communication and collaboration outcomes in team contexts. (JMIR Serious Games, 2021). (PLOS ONE, 2025).
I am not claiming games are a magic wand.
I am saying they are a clean way to get reps.
And reps are what we need.
Why traditional collaboration training often fails (even when it is well-designed)
Heads up, this part is not a dunk on facilitators.
I am a facilitator.
I have used most of these approaches.
They are helpful.
They are just structurally limited.
1. The feedback is slow and social
In real work, the consequences of poor collaboration show up late.
Sometimes weeks late.
And when they do show up, they are political.
People do not get clear signals like “you failed because your information sharing collapsed.”
They get blame, stories, and messy attribution.
2. The stakes are too high for experimentation
If a project is already on fire, nobody wants to test a new coordination strategy.
Teams default to what feels safe.
That usually means:
One person over-functions
Someone else disengages
The group stops telling the truth about capacity
3. Most training is explanation-heavy and rep-light
You can explain teamwork all day.
But collaboration is a behavioural skill.
It improves through doing.
Repeatedly.
Under constraints.
With reflection.
This is the same logic as sports practice.
You do not teach passing under pressure by lecturing about passing.
You create a drill with time pressure, limited options, and immediate feedback.
Games are drills with better constraints.
Why games are weirdly good collaboration practice environments
When I talk about “game-based learning,” I care less that it is a video game, and more that it creates the conditions most collaboration training struggles to reproduce:
Shared goals
Interdependence
Fast feedback loops
Low-stakes iteration
That is the backbone of why game-based learning can support soft-skill practice, transfer, spaced practice, feedback, and safe failure.
And this is the Rift point:
The game creates the experience. The debrief creates the learning.
Collaboration in action, 3 COTS games worth stealing from
I am going to keep this practical.
If you want collaboration to improve, you need games where players cannot succeed by parallel solo play.
You want real coordination pressure.
1. Overcooked, coordination under time pressure
What it trains: role fluctuation, proactive communication, adaptive handoffs
Overcooked forces teams to build a rhythm.
Not “talk more.”
Who chops, who cooks, who plates, who cleans.
Who calls audibles when the kitchen layout changes.
In the workplace, that maps directly to:
Managing cross-functional handoffs
Running operations during peak load
Keeping status updates short and useful
2. Portal 2 Co-op, collaborative problem-solving with asymmetric information
What it trains: explaining thinking, coordinating timing, learning from failure together
Portal 2 co-op is not about having the right answer.
It is about building a shared mental model.
One person sees a possibility.
The other person sees a constraint.
3. Minecraft (or other cooperative sandbox survival), planning and coordination over time
What it trains: division of labour, resource sharing, long-horizon coordination, trust
Minecraft is collaboration with delayed payoffs.
If teams do not coordinate early, they feel it later.
If they do coordinate, they build momentum.
That is why it is great for practicing:
Role negotiation
Shared priorities
“How do we know we are on the same page?”
What this means for L&D teams building collaboration training
If you want to pilot this without turning your org into an eSports league, here is a simple, evidence-informed approach.
1. Pick a collaboration behaviour, not a game
Behaviour: “Share key info early” or “Make clean handoffs” or “Resolve disagreement fast.”
Success indicator: something observable, like fewer duplicated tasks, faster alignment decisions, or fewer last-minute escalations.
2. Choose games that force the behaviour
You are not choosing “fun.”
You are choosing constraints.
Time pressure and tight coordination, use Overcooked.
Shared problem-solving with timing precision, use Portal 2 co-op.
Longer planning and role division, use Minecraft or another co-op sandbox.
3. Debrief like you mean it
A debrief is not “so, did you enjoy it?”
A debrief is the transfer bridge.
Use a simple structure:
What happened? (facts)
What patterns did you notice? (behaviours)
Where does this show up at work? (transfer)
What will you do differently next time? (commitment)
4. Measure lightly, but measure something
You do not need a 40-item rubric.
You need a signal.
Try:
Pre/post team self-rating on one behaviour (1–10)
Observer checklist with 3 items
A short reflection prompt collected after the session
Games can support collaboration outcomes, but the design of the intervention matters, which is why I treat the research as supportive evidence, not a guarantee. (JMIR Serious Games, 2021).
The bigger question this raises
If we agree collaboration improves through practice, why do we keep designing collaboration training that minimizes reps?
Games are not the point.
Practice conditions are the point.
Games just happen to be one of the best off-the-shelf ways to create those conditions.
What to do next
If you want to test this, run a 60–90 minute pilot with a single team.
Pick one behaviour.
Pick one game.
Debrief properly.
Then watch what changes in how the team coordinates in the following two weeks.
If you want help selecting a game that fits your context (team size, tech constraints, and target behaviour), that is literally what Rift does.
Comments