After the participants briefly introduced themselves, I summarized the June 1998 Game Developer Soapbox column written by MIT Professor, Seymour Papert.  Papert has two main points:

1. Most educational software is bad. Such software strives to be "easy", which isn’t what kids want or need. Furthermore, most educational software breaks knowledge up into isolated small pieces that are boring and lack personal meaning.
2. Games are more educational than most software that tries to be educational. Kids improve their learning skills by playing hard games that truly engage them. And game developers know how to make software that can be learned. Games that kids like are "hard fun". And kids want to be in charge. Furthermore, the learning experience can be enhanced by encouraging children to talk and to think about how they learn to play these games. Finally, kids should be encouraged to be game designers and builders according to Papert.

In each roundtable discussion there was some debate as to whether, indeed, all games that kids like really are "educational" in the sense that something worthwhile is learned. There was a general consensus that many games not considered educational are so. Some claimed that even the worst games teach hand-eye coordination. Some argued that while important learning was taking place, it couldn’t be marketed as such.

An issue that came up frequently was that game players learn many useless fictional details. Some responded to this by saying that game designers should replace the details with ones that are educational. An example we discussed in one session was Wolf – an adventure game where the player is a wolf. The more you know about wolves the better you can do in the game – not by answering quiz questions but by acting out the knowledge. Other examples of this sort discussed were Civilization, Age of Empires, a chemistry game that was being exhibited as part of the Indie Game Festival, and Oregon Trail. It was claimed that a slice of reality can be very attractive and educational. But the slice needs to have broad appeal (e.g. wolves do, while marmots don’t). In one of the sessions, we discussed the tradeoffs that ensue. For example, Age of Empires is largely historically accurate but accuracy was sacrificed when it conflicted with game play (e.g. making the different civilizations better balanced). An analogy with these kinds of games and historical novels was made.

Another issue was whether the basic thinking/learning skills acquired while mastering games transfers to other domains. For example, in many games you need to be good at systematically searching a space. Does this transfer to school or real life? Many opinions were offered, including that the transfer happens but is very hard to measure. Others agreed with Papert that transfer could happen if the player is encouraged to think about his or her learning, but that that parents and teachers don’t (and don’t know how) to discuss learning and make learning strategies explicit in the mind of the learner. Some questioned whether it was possible to measure or prove that a player has improved his or her ability to learn by playing games. Another participant argued that what should be measured is whether the player can solve the same problem in multiple interesting ways.

Another thread was open exploration versus constraints. Many argued that true freedom wasn’t needed and sometimes isn’t even desirable, but that the illusion of freedom that most games provide (but many edutainment titles don’t) is important.

Some claimed that it is hard to mesh education and entertainment. To be entertaining, only a small amount of time is spent on curriculum material. And making a game is too hard for this market. Entertainment wins when there is a conflict with education. Some argued there is a role for both curriculum-based learning and critical thinking, but that curriculum-based systems are very constrained by standards.

In every session, the phrase "chocolate covered broccoli" came up. Amy Bruckman introduced this in her talk "Can Educational be Fun?". Many agreed with Amy that much educational software is like broccoli in that in order to make it appealing, it has been "sweetened" with chocolate. This is like Papert’s "Shavian reversal" where the child inherits the worst of both parents. One participant claimed that the game industry produces "chocolate" but needs to learn to eat broccoli.

During every session there was a discussion of the conflict between the needs of the curriculum (and measurements of progress) and more abstract cognitive learning (e.g. critical thinking skills) that Papert claims happens when game playing. Some participants thought it was very important to match the approved curriculum since that is what both schools and parents demand. Others claimed that parents were more forgiving and are happy if their child has more physical engagement with the software or if their children create something they can be proud of. There was much talk of how the industry is very risk-adverse, how marketing studies keep showing that there is demand for software that helps directly with schoolwork, and how hard it is to market software that emphasizes thinking and learning skills. Many discussed how very good software that is open-ended, encourages creativity and independent thinking, and the like have had a very hard time in the marketplace. And things are worse now than a few years ago. One reason given was that discount stores like Wal-Mart are now big sellers of consumer software and they are very conservative in what they carry. Market studies show that titles positioned as "logical thinking" were addressing a small market. Some countered that these companies were looking at the market through the "rear view mirror". Someone said that the educational software companies have given up on the 12 years and up market because parents are no longer in control.

One participant claimed that edutainment titles strive for a kind of "intentional learning" which is very "orchestrated". Games provide a kind of learning that is more subliminal. Combining them is bad – they are 2 different products and should not be mixed up. Others claimed that titles like Incredible Machine and many puzzle games were in between these extremes.

Some claimed there was a contradiction between learning and what parents want and buy. Others thought that software that enables children to create things was the salvation to the curriculum problem and that parents see the value of making things.

Another theme was how different the home and school markets are. It used to be common to make software that worked in both markets, but this is no longer true. One participant claimed they were a "force fit". Schools and teachers want to design their own teaching plans. They want a large source of components from which they can pick and choose. Others countered that few teachers know how to incorporate software in their curriculum plans. Another participant said teacher creativity is needed to integrate software into the classroom. One participant pointed out that school supply stores were open to open-ended thinking titles, unlike school systems and many parents. Another said that teachers expect titles to help students memorize facts because they want their students to do well on standardized tests.

Many agreed that it was easier to make game designers into educators than to make educators into game designers. Good educators have good person-to-person skills but are not likely to truly understand game design. Some argued that good collaborations between game designers and educators was the answer and others told of cases where it hadn’t worked.

Some game designers criticized educational titles. One complaint was that most educational titles do not totally immerse their players into a game world. Another is that don’t give the player enough control. Players want to "push" and not be "pulled". One participant argued that educational games don’t have enough violence to appeal to him and his younger brother.

One participant argued that good delivery is the most important element. She urged the others to aim for software that shares features with the great teachers that teach at differently levels simultaneously. What is needed is software that is entertaining, charismatic, and filled with the right content. Blue’s Clues was claimed to be a good example.

One participant argued that good educational software should be like a Montessori school where the students have freedom and learn by playing games.

There was a discussion about math teaching. Some argued that behaviorist teaching is a good match with 3rd grade math. Others countered that math needs to be relevant. It isn’t fair for a teacher to argue that a student should learn math because otherwise they may build a bridge that would fall down. The student would really care about getting the math right if they really were building a bridge. Students ask themselves, "Why am I learning this?" Some argued that the answer is to make the learning relevant to the life of the learner. Reading and math, for example, are needed in life, so they should be integrated into games. Building bridges came up in another session about the hard problems that come up when constructing things. The problems were based on math, physics and mechanics. Another said software should support a continuum from using to modifying to building. Level editors were presented as a common example of support for modification and limited game building. Someone else claimed that building things is ultimately a communication act – that things are built to show to others.

This led to a discussion of physics and how it should be incorporated into games so that you need to understand physics to succeed in a game. Another idea is how much kids could learn about physics by building some kinds of computer games.

In the third session there was a long discussion of why hint files, walk throughs, and cheat codes were popular with many players. Doesn’t this contradict Papert’s claim that children want "hard fun", since walk thrus and cheat codes make games too easy? Some claimed that hints were only needed by players to work around bad designs. But others claimed that for some players the only goal is to win and "cheating" is OK. In another session, someone said it was important to pay attention to the "granularity of failure" and how it could lead to a good path.

In the third session, a long discussion followed when someone claimed that Papert was a "quixotic figure" that was "losing his battle". He asked how we can help turn the tide. Some said that Papert was inspirational to many in the industry. It was claimed that Papert has lost the battle in the schools and is now focusing on parents and the home.

Another thread was the difference between being educational and having an educational element. Someone claimed that it was far better to produce a title that was 5% educational if 1,000,000 players played it for 50 hours than to produce a title that was 100% educational that sold only 5,000 copies and was played only 2 hours.

Many of the discussions could be roughly put into two categories – those issues which are really about how our society, industry, or educational system needs to change and those issues which the game developer community can directly affect. Personally, I thought the discussion was more satisfying and coherent when focused on games and not on the problems everyone is having dealing with parents, teachers, standards, retailers, etc. But to be fair, it was clear that to many these issues were intertwined – you can’t just make a worthwhile game (Droidworks from Lucas Learning was often given as a prime example) and expect it to sell on anything other than its entertainment merits.