Dr. Peter Andrew Jamieson — Academic in an information evolution

← Back to Meanderings, Opinions, and Articles

Games…What Are They Good For? It’s All Systems Thinking, All the Way Down

October 2026 · written with the help of Claude (Anthropic)

A paper that sent me back to games

I recently read a paper called Can Games Save Education? (IEEE Computer, October 2026) by Magy Seif El-Nasr at the University of California, Santa Cruz. I thought it was good, though it is probably preaching to the choir of people who already care about games. It is still well worth putting in front of everybody else.

I am in that choir. I wanted to make video games, and I had been building my own since high school, based on the tabletop games I was playing. My dad would help me financially through college only if it was in engineering, science, or mathematics, so I picked computer engineering, figuring those were the people who built games. Outside of a hobby, I did not get to build one until much, much later in my career.

My thesis is that games are systems, and that playing them is a good way to start building systems thinking, especially in the emerging space of artificial intelligence.

Systems thinking, more than critical thinking

When people talk about what students need, the phrase you hear most is critical thinking. My complaint is that we always say it, but I do not know that we really understand what it is, or how to teach it. We are talking about how we think, and we rarely break that apart systematically.

We do the same thing in coaching hockey. We all agree kids should have fun, but fun is hard to guarantee for everybody, because really it is a perception. USA Hockey has tried to pin it down. Its American Development Model piece “What makes youth sports fun?” leans on Amanda Visek’s research: when kids ranked what makes sports fun, trying your best, a coach who treats them with respect, and getting playing time came out on top. Winning came 48th.

I ran into this myself. I was having a hard time writing academically, basically because I was not a good writer, and I am not necessarily one yet. My advisor, Jonathan Rose, gave me a book on meta-thinking. It was interesting, but it was not tangible enough to be useful.

What worked was practice. You cannot just write to yourself. You have to write to the outside world. So I wrote about something I cared about, which at the time was Ultimate Frisbee. Because I cared, I looked at what I wrote, evaluated it, and asked questions of it. That is how the meta, critical side developed, and no book ever explained that to me.

I think we do the same with critical thinking. It is one of our main answers when we ask how to teach the next generation to use artificial intelligence. I do not think we know how to do it ourselves, let alone how to help someone else.

I think the better phrase is systems thinking: a way of breaking ideas down, combining them, and updating our mental models against reality so that we understand things better. That is the territory of Derek and Laura Cabrera’s book Systems Thinking Made Simple: New Hope for Solving Wicked Problems, and it is the feel of what I think students have to work on.

It is also why we do first-principles exercises to develop our semantic knowledge (as I argued in my last post). Semantic knowledge, to me, goes beyond understanding ideas one at a time. It is thinking about how those ideas, the systems of our world, actually do things.

Games are models

Games are models of systems. Modelling is much of what we do in the applied sciences. We use mathematics to describe a system, knowing the description is not an accurate representation of the thing itself, and then we extend it to see what behaviours emerge. With a computer to run the calculations, we can simulate a system we might want to build in the real world and estimate how it works much more cheaply.

A game is the same thing in simplified form: a representation of a real-world system. Monopoly captures the idea of getting a monopoly on properties and charging people for staying in them. Even a trick-taking card game models something: managing your cards to win the most hands.

Jane McGonigal, in Reality Is Broken, has a definition of a game that I think is really good. (She visited Miami University in 2013, when Reality Is Broken was our Summer Reading Program book, and gave the convocation talk “The Hard Part is the Fun Part.”) A game has four traits:

  1. A goal.
  2. Rules.
  3. Feedback, which tells you how well you are doing.
  4. Voluntary participation.

She relates it to golf, and all four line up:

  1. The goal is to get a ball into a hole. On its own, that does not sound that exciting.
  2. The rules make it interesting: the ball starts super far away, and you have to hit it with a stick.
  3. The feedback is how many hits it takes you.
  4. Nobody makes you do it. You choose to play, so it is a game.

Whether you chase the goal together in a co-op game or against each other in a competitive one, the feedback tells you how you are doing and you adjust your behaviour. Most of the time you are solving an optimization problem: what do I maximize or minimize to unpuzzle this?

Playing is the way in

That is why I believe simply playing games is a wonderful place to start on the path to systems thinking. Systems thinking is the more formalized way of working through these ideas, but playing games gives us an entrance into that space.

Take a game I am pretty good at: Five Tribes. It is an abstract game, meaning you are moving pieces around with some purpose, and it does not really relate to anything. It is still a system, and the goal is to finish with the most points.

There is really one action. You pick up a group of meeples, which is just a name for coloured wooden blocks, and trace a path across the board. Where the last meeple lands, you match it with the others of its colour, and that colour, one of five, sets off a special ability that usually gets you more points. Before that, players bid points in an auction to decide who acts first.

Looked at simply, it is a game of actions and arithmetic. But the complexity and randomness of the board mean you have to work out what to do every time.

Five Tribes is much more tactical than strategic. A tactic is a good move on a single turn; strategy is long-term building toward an idea. I look at the board for a situation that gets me a large number of points, while thinking about what my move then leaves for someone else. Taking points for myself removes that possibility for another player, but my movement also lets other possibilities emerge.

It is very difficult to know what people will do. That is the emergent behaviour of the system: it comes out of two, three, or four people each making different moves. The state space of what could eventually happen is huge, and you cannot look at the starting board and know how the game will end.

The first time you play, you are just learning what is allowed within the system: what a legal move is. Only in your follow-up plays can you start looking for strategies that are advantageous and tactical moves that are solid.

Where the systems thinking comes in

So where is the connection between games and systems thinking? It is likely in looking at how you are playing and trying to understand the game as a system: a group of elements that combine into a greater whole.

The Cabreras break systems thinking into four elementary pieces, DSRP: distinctions, systems, relationships, and perspectives. Five Tribes has all four.

Even that one element, looking from different perspectives, changes how you see the system significantly.

Games in my research

I have tried this in my own research. With Lindsay Grace, Boyu Zhang, Naoki Mizuno, and our artist John-Rhys Garcia, I built verilogTown, a game whose main goal was to get people playing more with how to use Verilog to design hardware.

With Eric Rapos I run Let’s Play, which uses games for faculty development: helping teachers become better educators. A game is a system you have to teach someone, so teachers can test out interventions relatively quickly, in a safe space where there are not many consequences other than winning or losing. We have written articles, we are writing a book chapter, and we run workshops.

The cost of seeing everything as a game

The weird part for someone who studies, uses, and builds games is that I start to think of every other system I am part of as a game. That is both good and bad.

Suppose someone designs a system to motivate me, or to distribute resources. Because it is a game, it has rules. If you are into optimization, you look at those rules and ask two questions. Have they been designed well? And is there a loophole that lets me win, maybe at the expense of the institution or the people I am playing with?

Many of the systems I am involved in carry a presiding assumption: don’t game the system. The institution designed its game with a goal, say fairness. But the rules as written can still hand an advantage to someone who chooses not to be fair.

Higher education as a game, and what AI does to it

This is where I think the main argument sits. A system designed for a real-world purpose, to make money or to change people’s behaviour, is a little different from just playing a game. Education is one of those systems, and I think its transactional ideas turn it into a game where winning means getting good grades. Artificial intelligence can help you win that game, and that is a bad thing.

Given that AI exists, our major approach to assessment has been to go back to the blue book or the oral examination, to check the individual’s thinking. I think playing a game for the sake of assessing someone’s thinking is an interesting space. For one, the AI probably cannot handle the complexity of the system.

The downside is that these are fuzzy things, and it is hard to count fuzzy things. Real thinking, and then assessing real thinking, is very tricky.

Systems thinking, all the way down

The big point is that games are first-principles exercises. They let us experience designed systems, play them, and try to understand them.

Once you see a game that way, you start to see the systems around you the same way: the incentives, the classroom, and what AI does to both. It is systems thinking, all the way down.

References

This post was written by me, with the help of Claude (Anthropic) for editing and for tracking down the citations.

Peter Jamieson is an Associate Professor of Electrical and Computer Engineering at Miami University. More of his work, publications, and projects are at drpeterjamieson.com.