Let’s start with one of the oldest and most popular games: Super Mario Bros for the NES. In it, the player controls Mario past a number of hazards, through a number of levels to rescue the princess. Young children in the 80s and 90s would spend hours upon hours playing, beating and (important for today’s topic) *learning* this game. See, beating a video game is a learning process. As kids play, they learn to play better and better until they play well enough to beat it. Then they keep playing and learn to play better and better to beat it faster, more consistently, more stylishly or what have you. Some of this learning is physical, you can train your reflexes to work faster, but a lot of it is actually learning how the game works and what you need to do to complete it.
On the surface, learning how the game works seems kind of basic, but is it? The game has a large number of enemies with their own patterns, and those enemy patterns can combine in a large number of ways to challenge the player. The player has to learn how to approach each situation, and how to adapt to a situation that isn’t going how they expected it to. They may even plan ahead and devise multiple strategies before testing each one out in turn and going with whichever is best. The player might also be memorizing the map layouts of the levels, the locations of secrets, and all sorts of other things. It’s safe to say that a lot of real learning is taking place, even if it isn’t “school learning” like what we’re used to.
Kids do a lot of work learning to play video games, and thus since the very dawn of video games parents and teachers have wondered if that energy could be more productively transferred towards academic learning. This eventually morphed into a “gamification” push, where many modern schools will put at least some effort into having gamification elements in their teaching in order to motivate students to work as hard at academic learning as they do at gaming learning. Now, gamification is an INCREDIBLY broad topic and it doesn’t just cover video games that try to teach you things. I can personally remember playing cheesy point-and-click video games that tried to teach me the parts of the body or the planets of the solar system, but video games themselves are only a small facet. Gamification can also be as simple as having class leaderboards to encourage students to do well and get good grades, badges or points for completing certain tasks, there are all sorts of ways to gamify a learning task.
But this brings me to today’s question: can you gamify science education? Now first off it’s very clear that you can gamify *early* science education (thinks like elementary or middle school) since we’ve had those sorts of things for years. Teaching a student about the human body, or the planets, or the teacher creating a whole Jeopardy! setup to help them learn the parts of a cell, these are all gamification aspects that were used to teach me and many others about science over the past few decades. But post-secondary education is a different beast and often entails learning things on the cutting edge that aren’t always fully accepted by the entire community. Science does have its internal struggles, and if a student learns by reading papers (which is necessary to study topics on the cutting edge) they will by necessity be learning about at least some ideas which will later be proven false. That’s ok, science isn’t a set of facts, it’s a process for discovering the truth, but that does make it harder to “gamify” since you can’t just program a game with right and wrong answers, because on the cutting edge *we don’t have all the right and wrong answers* and we’re learning new things every day.
I thought long and hard about this question: can you make a video game (or something like it) that would allow students to study a cutting edge topic like proteomics? I pick this topic because it’s one I know a lot about, and I came away thinking the answer is “no.” A proteomics game would either be highly simplistic and thus not very useful for cutting edge studies (high school studies perhaps), or would be so complex that you were really studying someone’s protein simulation and not proteomics itself. Let me explain.
A video game for proteomics would have to have certain limitations. The first limitation is the pre-defined actions that the programmers allow. Mario can’t climb walls in Super Mario Bros because the programmers didn’t program that, they only programmed certain actions. As far as I know, all proteins are biologically synthesized in an N-to-C direction. So presumably the program would only allow synthesis in this direction, but what if we discover some organism that can synthesize C-to-N? What if we discover organisms that synthesize or modify their proteins in ways we did not expect, and what if those proteins become scientifically or economically relevant? A programmer can’t exactly predict every possible action that biological proteins could take, and so can hardly program every possibility.
OK, so they can’t program every possibility, but what about creating an open-ended system that would allow the “players” to create their own actions? That brings limitation number two: the approximations used. An open ended proteomics game would by necessity need to employ certain approximations in the code to allow for proteins to be synthesized and moved around at will, it isn’t feasible to create a perfect simulation that can calculate the effects of every atom and bond in a protein. So a game would have to use a number of approximations to allow for this open endedness, but then you end up with the problem where students may not be studying anything real but studying only an approximate model that doesn’t work in the real world. My most notable reminder of this is the game Kerbal Space Program which is a fun little astronaut simulator that, due to computer limitations, has to use a set of heavy approximations for gravity that make it very inaccurate with the real world. This leads to some fun but physically impossible creations such as perpetual motion machines and giant mecha.
It’s not just the scope but the scale. You can do so many things with proteins, there are 20^10 combinations of 10 amino acids. All those possibilities can’t be programmed in. The best molecular dynamics currently has is the ability for super computers to roughly approximate the actions of proteins by simulating all the atoms and bonds, but even those simulations require heavy approximations. So if you try to make more and more approximations, you end up with a program where students aren’t studying proteomics but rather studying the approximations that are built into it.
The final, most important piece of this is: how would you make such a thing fun? Science, as in actual science, is fun to me because I get to learn and discover new things. As said before, a video game would necessitate such approximations that nothing “new” could really be discovered. Games like Kerbal Space Program are fun because they give you the means to perform some of humanity’s greatest feats for yourself like going to the moon or launching a robot to Mars, but what are the equivalent actions that could be done in a proteomics video game? I honestly can’t think of anything proteomic that makes me think “man I’d like to do that for myself!”
So yeah sorry to be a debbie downer but I think the idea is unworkable for now. Stick with fun little games for early childhood education and then read papers when you go to college.