In 1962, as the Civil Rights Movement was ramping up, 26-year old Stanley R. Puryear—now Dr. Muata Weusi-Puryear—enrolled in the PhD program in Symbolic Logic at UC Berkeley. When looking for housing, “he discovered the local apartment owners did not rent to Black people.” He then joined the Palo Alto-Stanford branch of the NAACP and “became the litigant in a case against a Sunnyvale apartment complex.” His “precedent setting case proved that the housing industry was not exempt from a California law that outlawed racial discrimination in businesses.” His focus on civil rights led him to drop out of his PhD program and become the president of the Palo Alto-Stanford branch of the NAACP in 1964 [1].

From 1966 to 1968, he was the general manager of the Stanford-Brentwood CAI Program led by Patrick Suppes (who was also a professor of symbolic logic), the largest computer-assisted instruction effort in the country. In 1972, he returned to completing a PhD, this time at Stanford—although his advisor, Edward Begle, was at Berkeley where most of the work was completed. His dissertation assessed the educational benefits of an educational game that he developed. In particular it sought to answer a very specific, but important question, that to this day seems to be neglected in the frenzy of creating educational games:

Can the game elements of a computerized tutorial/game motivate student involvement in the tutorial elements to a degree high enough to produce significantly greater achievement than the tutorial elements alone can produce? (Weusi-Puryear, 1975, p. vi)

As Dr. Weusi-Puryear recalls, some of his dissertation committee members—including Patrick Suppes, who in 1973 had become the President of the American Educational Research Association and the National Academy of Education—were skeptical that a game would have added benefits over standard tutorial software. (For many educational games, recent research would suggest that such skepticism is warranted!) Nonetheless, in an experiment with 258 students—mostly Black students from majority-Black schools—Dr. Weusi-Puryear found that

Students who had an opportunity to continue their partici­pation in a game if they correctly responded to randomized exercises, achieved significantly more than students who did not have the oppor­tunity to play, even though the game-playing students did fewer exer­cises. (Weusi-Puryear, 1975, p. viii, emphasis added)

This seems to suggest that the motivational benefits of an educational game can be more than added time-on-task. But surely it matters what the game is. The particular game that Dr. Weusi-Puryear made was called Gambo, later cleverly titled “Arithmetic-Tac-Toe.” The game works as follows:

The teletype output from the GAMBO system looks like a script of a play with three characters—Gambo, the student, and Jody. Each player in turn is given an arithmetical problem to solve; after he has committed himself to an answer, his opponent is given an opportunity to evaluate the player’s answer; the umpire presents the correct answer; if the player’s answer was correct, he gets an opportunity to make a move on a three-by-three Tic-Tac- Toe board; and if the player’s answer was incorrect he would forfeit his turn on the Tic-Tac-Toe board….In the GAMBO system a player can receive ten points for correctly answering the problem, five points for correctly evaluating his opponent’s answer, and fifteen points for getting three marks in a row on any Tic-Tac-Toe board. A new Tic-Tac-Toe board is presented after either player receives three marks in a row. (Weusi-Puryear, 1975, p. 11)

What makes the game particularly interesting is how Jody—the “AI” opponent—was programmed.

In general Jody attempts to imitate the student. In particular he matches the student's reaction time and tries to match the student's problem­ solving abilities (or disabilities)....The probability that Jody will give any particular type of response (correct or one of the various incorrect responses) is equal to the portion of the student's past responses in that given type to his total number of responses. (Weusi-Puryear, 1975, pp. 11-12)

To make Jody appear more human-like, their typing rate was adjusted to be one character per second. This made students perceive Jody as a human opponent, not a computer opponent:

On several occasions the students remarked that “Jody is stupid,” or “Jody is trying to cheat,” or attributing other human characteristics to Jody. (Weusi-Puryear, 1975, p. 19)

This dissertation was the beginning of Dr. Weusi-Puryear’s career in educational game design. In 1978, he founded Edutek Corporation, where he designed over 35 educational programs, many of which were games or had gamified elements. For ten years, the company had a successful run in selling educational games. The images below show pages from the 1983 catalog, including the philosophy behind the “Fun Drill” approach and a sample of the educational software they created.