Video games model good teaching

Video games model the best teaching strategies, writes neurologist Judy Wills on Edutopia.

Games insert players at their achievable challenge level and reward player effort and practice with acknowledgement of incremental goal progress, not just final product.

When players meet a challenge, dopamine is released, producing deep satisfaction, Wills writes.  A task that’s too easy or too difficult won’t activate the dopamine reward circuit.

MRI and cognitive studies reveal that the brain “evaluates” the probability of effort resulting in success before expending the cognitive effort in solving mental problems. If the challenge seems too high, or students have a fixed mindset related past failures that they will not succeed in a subject or topic, the brain is not likely to expend the effort needed to achieve the challenge.

Video games let players move quickly to their level. They’re not stuck at level one just because other players haven’t mastered the skills yet. They want to work harder — to move to a higher level — because it feels so good to win. That model works for classroom learning too, Wills argues.

The best on-line learning programs for building students’ missing foundational knowledge use student responses to structure learning at individualized achievable challenge levels. These programs also provide timely corrective and progress-acknowledging feedback that allows the students to correct mistakes, build understanding progressively, and recognize their incremental progress.

Game designers provide “just in time” information to players, said researcher James Gee at a MacArthur Foundation seminar reported on HechingerEd.

. . . (In) the strategy game Civilization . . .  players can access a “civilopedia” that contains details on the various game concepts, civilizations and units featured in the game.

“In school, information is given to you whether you want it or not and never just in time,” Gee said. “You’re not going to use the 500 pages until you finish them, and by that time you can’t remember what was on the second page.”

Many games track players’ performance throughout their playing time, showing where they faced problems. That kind of detailed feedback would be useful to teachers, Gee said.

Update: “Just-in-time learning” doesn’t work for math, argues Barry Garelick.

About Joanne


  1. Roger Sweeny says:

    Maybe this is too obvious to be worth stating, but the strategy outlined in the post is impossible in any remotely heterogeneous classroom.

  2. SuperSub says:

    More importantly, to achieve the feedback necessary any current conceptual lessons would need to be broken down into individual concrete skills, which seems to be a no-no currently.

  3. tim-10-ber says:

    Just how does what one “masters” on a video game get captured on a real test?

  4. Mark Roulo says:

    Just how does what one “masters” on a video game get captured on a real test?

    I think the point is that video games tend to ramp the players up so that after ‘N’ hours the player is measurably better. One of the ways that they do this is by providing incrementally harder challenges. The games also provide a tight feedback loop and the players can move at their own speed … I don’t have to go to level 18 just because everyone else is. If I’m having problems with level 10, I can (and probably will) stay there until I’ve mastered/beaten it.

  5. “Just-in-time” learning? Wow, we’re not even pretending that school isn’t an assembly line anymore. You know, I can teach you how to identify a metaphor just in time for test prep. This method is EXACTLY how one teaches to the test.

  6. Stacy in NJ says:

    Most games have three levels: beginner, intermediate, and advanced. You can “win” the game (travel through all individuals elements or challenges) at each level. It incrementally ramps up, and then up, and then up. You can repeat any particular challenge as many times as you need to to master it. It may only take player A 10 hours to beat the game at the most advanced level; Player B may need double or triple the play time to reach the same level. But once a player has beaten any game, it is likely it wil take them less time to beat the next game and less time for the next. And that’s the really interesting aspect of game theory.

  7. Cranberry says:

    Civilization’s a great game. It is, however, only a game, and the players do learn how to beat the game. They don’t learn all the definitions, only those which are most effective or most lucrative.

    I find this statement hard to accept: “In school, information is given to you whether you want it or not and never just in time,” Gee said. “You’re not going to use the 500 pages until you finish them, and by that time you can’t remember what was on the second page.”

    Some of us remember. If I have to choose between Gee and E.D. Hirsch, I’ll choose Hirsch. If everyone had to look up (or be given) facts “just in time,” no one would ever be able to study law, history or medicine.

    Gamers can create challenges for one another that create learning opportunities. Gee cited an example focused on the popular life-simulation game The Sims. Players were challenged, in one case, to play the entire game as an impoverished single parent and then create a graphic novel about their experiences. “That sounds like a pretty good social-sciences assignment, right?” Gee said.

    That sounds like a very slow-moving social sciences class, with a very undemanding, but time-consuming project. I’d rather my child were required to read “How the Other Half Lives,” by Jacob Riis, then to write a book report. ( I think he’d learn more from reading that book than from playing a game set in an artificial environment.

  8. I find this very interesting.

    I’ve been trying to write educational software for several years and I’ve found there are numerous factors involved.

    It’s a shame that good educational software doesn’t really exist. It’s a shame because “programmed learning” works.

    Fewer than half of all adults know when to use lay or lie but most students who run through 100 short questions in a programmed learning setting can master it.

    The technology exists for creating great math and grammar software but, for some reason, it hasn’t been done.

    In fact, we’ve moved backwards. The software in the 1980’s was better than what we have now.

    One of these days, we’ll have great educational software, but I have only one more year before I retire so I’m not going to see it.

  9. George Larson says:

    Robert Wright

    This is a very interesting observation. Can you speculate why?

    Are the best programmers and software designers involved in other types of programming?
    Is programming no longer interesting or too difficult for teachers?
    Are instructional programs selected by incompetent committees?
    Are programs designed to appeal to adults and not the target audience?
    Is this type of programming done on the cheap or as an after thought to support textbooks?
    Do programs emphasize multimedia too much?
    Programs are not tested for effectiveness?

  10. tim-10-ber says:

    Mark — Surely you jest. I understand how video games are played…but mastering a level in a game and truly learning something that can be put on paper…well, I have to see it to believe it can be done. Other than the basic technology classes in school that teach kids how to use a computer, word, power point, excel are not working…not in many many government schools. The private schools will tell you even though they require the computer it will not enhance academic performance. Do government schools tell this to taxpayers and parents we are requiring/providing a computer but we are in essence wasting more taxpayer money on something that does not enhance academic performance.

    I grant you the way parents have chosen to raise kids today with the bombardment of stimulus and quick “reward” probably has done something to kids brains. Heaven help us when these kids get to the work force…

    Surely (and maybe there isn’t) schools can help rewire the brain rather than succomb to another silly fad that will not produce real learning…please give me a break…I am not a fool…

  11. Roger Sweeny says:

    The technology exists for creating great math and grammar software but, for some reason, it hasn’t been done.

    The greatest software won’t do anything if kids don’t use it. Kids just aren’t interested in most of what they are expected to learn in middle and high school. They will not sit alone at a computer screen and work their way through an educational program (at least not for long).

    Right now we have them with 20 or so of their peers and a flesh and blood teacher. They still don’t learn much but they are forced to have some interaction.

  12. tim-10-ber says:

    @ Mark — Okay..I have to take back some of my comments…I still believe the video games impact the brain and technology does not produce higher academic achievement…

    but…back to mark…now after my second cup of coffee…I understand better the concetpt…I am reading Teach Like a Champion and you are right…the constant checking back before moving to a higher level is exactly how champion teachers teach….

    I have also read “The Flickering Mind” and am very frustrated about technology in school, the improper use and the waste of taxpayer dollars and teachers’ time…

    My apologies…

  13. SuperSub says:

    Robert –
    I completely agree with your criticism of current teaching software. I remember using a series of math and reading/writing programs in elementary and middle school… just due to those programs alone I skipped 2 years of math between 6th and 7th grade. I think what has hampered current programs is the curricular denial of the importance of basic skills… the current software does not provide the same focus and repetition of individual skills that is required to work.

    Learning is learning, and Wills hits the nail on the head with her comparison. Video games do exactly what school is supposed to do – build intrinsic motivation for performing seemingly pointless tasks. For those who played and remember the original Super Mario brothers… think about the first and last levels and the skill required for each. If the entire game had the same difficulty as the last levels, players would immediately get frustrated and give up. If the entire game remained easy like the first levels, players would get bored and get distracted or give up.

    As for the just-in-time argument, I think Wills is discussing the need to break large tasks into smaller manageable ones that actually allow the student to learn. If you assign a whole book to be read and then give a test on the events of the book, most students will fail miserably. Instead, if you give chapter quizzes and then a test at the end, students will retain information better.

  14. GoogleMaster says:

    No one has mentioned the phrase “game cheats” yet. What is learned or accomplished when players level up using cheats they’ve downloaded from the internet?

  15. GoogleMaster says:

    George Larson —

    Having a member of my household who was employed by a university department that produced science games for middle school students, I have a few insights.

    In my experience (sample size of one):

    – Some (many? don’t know) of the games are produced by grant-funded institutions.
    – University pay for a programmer/designer/artist is about half what those people could get in the open market.
    – Grant-funded projects have limits on resources and time for a particular game.
    – The PI may be more interested in lining up the next grant than in overseeing the work on the current one.
    – Some PIs are crazy. (Again, sample size of one. YMMV.) The one I know of drove away so many people that the department completely turned over two times in three years.
    – The programs are targeted at the age range, and they are tested and evaluated on that age range for effectiveness. (Kids take test, play game, retake test to evaluate learning.)

    I think the big problem is that the people with the skills to create great games are at actual gaming companies producing entertainment games, not at grant-funded institutions producing edutainment. The second half of it is that edutainment doesn’t bring in nearly as much money as a first-person shooter or the latest NFL whatever, so the gaming company isn’t going to bother.

  16. Cranberry says:

    SuperSub, James Gee spoke of “Just in time,” not Judy Willis.

    If you assign a whole book to be read and then give a test on the events of the book, most students will fail miserably. Instead, if you give chapter quizzes and then a test at the end, students will retain information better.

    Is it essential that students remember the events of the book? Doesn’t that depend on what sort of book it is? Some auto-didacts memorize the dictionary. Whether it’s done by sitting down with a dictionary, or in a structured, artificial environment, does memorizing the dictionary count as a good education?

    What you’re really speaking of at present with games-facilitated instruction is a variation on efficient drilling of basic information. Increasing reaction time by practicing facts (such as multiplication tables), is very useful.

    I would decline to enroll my child in a school which relied upon this sort of artificial computer environment to teach. Yes, it would do a great job for test prep for state tests with a limited, predicted curriculum. It wouldn’t teach what would not be on the test, though. Why do that, when there’s no reward to the school, or its software contractors? There’s also no provision for more complicated learning, such as writing or debating. Most video games are very simple versions of shoot-em-ups. I’m sure they’re very effective in educating future soldiers. Future diplomats, however? Not so much.

    I would greatly prefer the Montessori method, or a school which had the courage to track children. Allowing children to progress at their own pace is more effective. One need not turn to computers to find it.

  17. George, you speculations are good ones, GoogleMaster, your info is interesting, and SuperSub, i know what you mean.

    I wish I knew why the educational software is worthless, why many video games make much better use of the science of learning, and why the educational software of 30 years ago is superior to anything you can get today. I wish I knew. I don’t.

    But I have just a couple ideas to add.

    30 years ago, the best educational software was teacher made. BASIC was used and didn’t have to be cross-platform or compatible with HTML. These days, it pretty much has to be tied to the internet, if not completely on the internet, and JavaScript and the like is much, much more difficult to master than BASIC. Back in the day, most schools had four or five teachers who could program in BASIC. These days, there are entire districts without one person who knows beans about JavaScript. And even the best JavaScript programs, from what I’ve seen, haven’t incorporated some simple learning steps that were easy to include with BASIC

    When software companies first entered the education market, they combined Space Invaders with multiplication problems. You had to solve the math problem as it was falling to Earth inside an astroid. That was a bad marriage of ideas. It was torture.

    When they put out the glossy stuff with killer graphics like JumpStart, they sold very well and didn’t teach anybody anything except how to insert a CD and use a mouse.

    Oh, then there were a lot of programs that gave such wonderful graphics when you got a problem wrong, students provided wrong answers on purpose. Not good.

    And another hugh setback was the idea that since computers were such amazing devices, they should be used as tools for creativity and not “drill and kill” learning machines. That idea still is dominant, which is too bad. As a consequence, the “drill and kill” programs haven’t improved or kept up with technology. Computers need to return to the less glamorous role of teaching machines. Why? Because they really work. “Drill and Kill” got a bad rap. We need to move beyond how that ugly phrase sounds. Do you want a student to learn how to distinguish between a direct object and an indirect object? There is nothing at all more effective than a so-called “drill and kill” program to do that. The same with learning the difference between lay and lie, unless it’s a teacher with a club standing over a student who’s going through a programmed learning book to make sure he’s not skipping pages.

    One summer, a long time ago, I taught five 8th graders who didn’t graduate because they could’t do long division. I used a simple teacher made program written in BASIC and after a few weeks, long division was a piece of cake. The constant trail and feedback these students had to go through just could not have been achieved through student-teacher interaction even with a ratio of 5 to 1.

    Some learning is best done by teacher led direct instruction, but other learning is best achieved by computers set up as teaching machines with a programmed learning design that would include some techniques developed by the well-funded video game industry.

    One more thing to add.

    The very best educational materials I’ve ever seen have been the H&R Block materials on how to become a tax preparer, the Radio Shack book on how to pass your ham radio license and the attitude-leads-to-behavior booklets you get a Traffic School.

    I really hope that the day comes when computers can be teaching machines again.

    If somebody could take the 440 lessons and 88 quizzes from
    and put that in JavaScript that would allow for time-delay suspense, branching, auto remediation, with awarded levels of achievement and some other lessons learned from the gaming industry, it would be wonderful.

    Right now, the mandated book we use at school is a revamped Warriners book that had to be renamed a “handbook” because it was clearly too awful to be used as a workbook. Why we still have Warriners grammar has to do with money, not education.

    What we need is something like in a way where it could be interactive.

    Too bad I’m not smarter. And too bad I’m going to retire before I see this done.

  18. Roger Sweeny says:

    The very best educational materials I’ve ever seen have been the H&R Block materials on how to become a tax preparer, the Radio Shack book on how to pass your ham radio license and the attitude-leads-to-behavior booklets you get a Traffic School.

    Interesting. The first two are things that the student has voluntarily chosen to learn. He wants to be able to prepare taxes. She wants to be able to use a ham radio. The third has an imposed motivation: no pass, no license.

  19. Mark Roulo says:


    No worries 🙂

    -Mark Roulo

  20. Roger, I see your point that motivation to learn is a key factor in these educational materials.

    The reluctant learner has zero entry-level motivation to engage in educational software and a lot of motivation to play the stupidest video games imaginable.

    Yet once the reluctant leaner feels the rush of dopamine in well-desgined educational software, that might create all the motivation that’s needed to see it through.

    If you break down the rules regarding depreciation and make it easy enough for a dimwit to achieve success, there’s a drug like rush. A well designed program can build on that with the right pacing. Content isn’t important for the computer user. It’s the drug rush.

    How to get a video-crazed zombie to sit down and give a good educational program an initial two minutes of engagement is all you have to do. That’s the good news. But it might require a shotgun. That’s the bad news.

  21. Cranberry –
    Thanks for the “just in time” correction.

    As for my book example, it was just that. You could replace it with just about anything, including mathematical operations, writing essays, or memorization (gasp) of scientific facts. The point was that retention of anything occurs best when it is done in manageable chunks with immediate practice and feedback. This is necessary no matter what type of school you belong to… that’s the benefit of real neuroscience to learning… no matter who you are, what race you belong to, how much money your parents make or where you live, 99.9999999% of human brains all function the same. None of this edupsychobabble that has infected our society.

    The real sad part is that 20-30 years ago, teachers understood how learning worked. Reaching for abstract teaching styles as opposed to time-tested (and scientifically sound) pedagogy has given us the unrealistic expectation that we can force students to learn more than they actually can.

  22. Roger Sweeny says:

    I’m not sure a few successes (or a lot of successes) will be enough to motivate a kid to really try to make it to the end of a piece of educational software.

    I have tried to play video games a few times. I make some progress and then get stuck. I try a few things and they don’t work and I think, “I don’t get it, this is stupid, this is wasting my time, I’d rather do something else.” I just don’t care if Mario gets to where he’s trying to go, or whatever the object of the game is.

    Many of my students say the same things about school learning.

  23. Mark Roulo says:

    “I’m not sure a few successes (or a lot of successes) will be enough to motivate a kid to really try to make it to the end of a piece of educational software.”

    I don’t think the claim is that more-like-video-games solves the problem. I think the claim is more of the “more like video games in the following ways moves us closer to the right way to teach.”

    I expect that motivation is a huge component to learning. And I don’t see this addressed at all by the video game analogy. But other bits matter, too, and for some of those video games seem to have some lessons.

    Still, no silver bullet. Maybe remove some of the unnecessary obstacles, but the path to learning will still be steep for many subjects.