Serious games in learning and development

The term is older than the software. Clark Abt published a book called Serious Games in 1970, when the games in question were played with cards and a room full of people, and the modern revival dates from the Serious Games Initiative set up at the Woodrow Wilson Center in 2002. Nothing about the idea requires a screen.

The label is not the useful part anyway. What matters when you are choosing one is where the learning sits. In gamification, the content stays exactly as it was and a game is bolted on as reward: points for finishing the module, a badge for the streak. In a serious game the game is the content. What you want people to learn is inside the rules, and the only way to play well is to do the thing well.

Serious game, gamification, simulation: what you are actually buying

Most disappointment with this category comes from buying one and expecting another. A platform that adds points to your existing e-learning will lift completion and change little else, which is fine if completion was the problem. It is not fine if the problem is that people escalate too late.

What a serious game is made of

Strip the theme off any serious game worth running and the same five parts are underneath. Design them in this order. The order is most of the discipline.

1. The behaviour. Finish this sentence before anything else: after this, people will ___ instead of ___ when ___. If you cannot finish it, you are not ready to design a game. You are ready to write a document, which is cheaper.
2. The model. The rules that decide what happens after a choice. This is the real curriculum. Whatever the model rewards is what gets learned, whatever the objectives slide says.
3. The decision. One or two per round, with a genuine trade-off and never quite enough information. A choice with an obviously correct answer is a quiz in a costume.
4. The consequence. What the choice produces, and crucially when. Feedback that lands instantly teaches pattern-matching. Feedback that lands two rounds later teaches judgment, because by then you have to live with it.
5. The debrief. Written before the game, not after. If you cannot draft the questions the game will make answerable, the game does not yet have a point.

A serious game teaches what its rules reward, not what its learning objectives claim. When those two disagree, the rules win. Every time.

Mechanics that produce judgment, and mechanics that only produce noise

Designers reach for points, badges, leaderboards and timers because they are quick to build and read as a game on a slide. They are also the four mechanics least likely to produce the behaviour you asked for. Points reward completion. Badges reward collection. Leaderboards reward the already-confident and quietly tell the bottom half to stop trying. Timers reward speed, which is the opposite of judgment in nearly every scenario worth simulating.

The mechanics that do the work are less decorative:

• Asymmetric information. Different players hold different pieces and nobody holds all of it. Nothing else surfaces who shares, who hoards and who never asks.
• Irreversible choices. A decision that cannot be undone forces a conversation before the click. Undo is the enemy of deliberation.
• Scarce resources. Time, budget, headcount, goodwill. Scarcity is what turns a stated preference into an actual priority.
• Costly verification. Let players check a fact, but charge them for it. This is the cleanest way to separate the people who verify from the people who assume.
• Roles with conflicting incentives. Give two players goals that cannot both be maximised and you have rebuilt the org chart in miniature, along with its conflict.

Fidelity is a trap

The most expensive design mistake is buying realism in the wrong dimension. Nobody learns more because the interface matched the real system or the video was shot on decent kit. What has to be faithful is the shape of the decision: the same pressure, the same missing information, the same colleague who will be irritated whichever way you go. Everything else can be a cartoon. Plenty of serious games that change behaviour are played with printed cards and a facilitator holding a stopwatch they mostly ignore.

Design one this week: the one-page brief

Before anyone quotes you for software, run a 45-minute session with three to five people who know the problem first-hand, and fill in seven lines on one sheet.

1. The sentence. After this, people will ___ instead of ___ when ___.
2. The situation. One real incident from the last year, anonymised enough to say out loud in a room.
3. The decision. The moment in that incident where it genuinely could have gone either way. That moment is your round.
4. The missing information. What the person did not know at the time. Do not hand it to players either.
5. The trade-off. What the cautious option costs. If caution is free, you have no game.
6. The reveal. What players learn afterwards that they could not have known during, and the moment you release it.
7. The debrief question. One sentence, pointed at Monday. For example: where in your actual work do you make this call with this little to go on?

If the sheet fills up easily, you have a game worth building. If lines four and five stay blank, you have a briefing, and you should write it as a briefing. The test costs 45 minutes, and stopping a procurement is a perfectly good outcome for it.

When a serious game makes the training worse

This is the section the vendor deck leaves out. Six cases where we would tell you not to:

1. The objective is knowledge. If people need to know a policy, a well-written page beats a game and costs a fraction. Wrap a rule in a story and people remember the story.
2. Failure is not affordable. If players believe their score is being read by their manager, they play safe and you learn nothing. A serious game needs psychological safety more than it needs a scoreboard.
3. The players are unequal and the ranking is public. Put new joiners on one leaderboard with fifteen-year veterans and you have taught the new joiners to stop volunteering.
4. Nobody will facilitate the debrief. Without that, you have bought entertainment and an invoice.
5. The group is small and already trusts each other. Below about a dozen people, a skilled facilitator with a whiteboard usually gets further, faster, for nothing.
6. The subject is genuinely painful. Bereavement, redundancy, harassment investigations. Game framing trivialises them, and the trivialising is what people will remember.

The retention percentages you will be shown

Somewhere in the pitch a pyramid appears: people remember 10% of what they read and 90% of what they do. Those numbers have no research behind them. They were bolted, decades later, onto Edgar Dale's Cone of Experience, which contained no percentages at all and was never a claim about retention. Treat the slide as a signal that the supplier has not checked their own sources. The case for doing rather than being told stands perfectly well on experiential learning grounds without inventing arithmetic, and you should be equally suspicious of any engagement uplift quoted to one decimal place.

Judging whether it worked, without inventing numbers

You will not get a clean before-and-after on judgment, and anyone promising one is selling. Use the Kirkpatrick model honestly: reaction scores are cheap and nearly meaningless, and behaviour is the level that costs something to measure. In practice four signals tell you most of it.

• What they argued about. Facilitators can record the disagreements. Recurring ones are your real training needs analysis.
• Whether anyone changed their mind mid-round. A team that never reverses a position under new information has told you something important, and it is not good.
• What language survives. Six weeks on, is a phrase from the game being used in a real meeting? That is the cheapest evidence of transfer available.
• What a manager can name. Ask team leads for one decision made differently since. One concrete example beats a page of satisfaction scores.

What we have learned building them

We design and code these for one client at a time, which is where design ideas either survive contact with a real cohort or quietly do not. HEARTWARE is five rounds and ninety minutes on judgment in the AI era, written for new joiners: teams decide what to trust, what to verify and what to own, and can question the characters in free text rather than choosing from a menu, which is where the interesting behaviour turns up. Operation Keystone is an induction for two hundred joiners in sixteen teams, half of it mental ops inside the app and half physical ops run by facilitators away from the screen, where solving the puzzle earns the passcode that unlocks the next level.

Two decisions carry more weight than any mechanic. Everyone plays in their own mobile browser with no installs and no accounts, because the first ten minutes of a session should not be spent on logins. And reveals are released by the facilitator to every phone at the same second, so no team can race ahead of the conversation. The longer version of how we build gamified learning apps covers the rest, including the parts that are not worth paying for.

Whatever you build, the last twenty minutes decide whether any of it was training. The game produces evidence; the debrief turns it into something that survives contact with Monday. A serious game without a debrief is just a game, and an expensive way to buy one.

Tour De Force runs this as live, experiential training for teams worldwide — online and in person. Talk to us, or play Gamified learning appsThe Weekly Challenge to see the method in ten minutes.