Authors
Dr. DeWayne Derryberry
Idaho State University
Objective
Determine if a backgammon program has been upgraded by comparing the performance of a player against the computer across different time periods.
Background
A college professor played an online backgammon program eight games per day, during lunch periods, for several days in 1998. Each day he kept score, positive scores when he won, and negative scores when the program won. Each game can be a simple win for the professor (+1) or the program (-1), but it can also be a gammon (± 2) for an extremely one-sided win or loss, and even occasionally a blowout, or backgammon (± 3). Further, there is a doubling cube, which allows the stakes to be doubled and redoubled any number of times. At the end of the 50 days the professor tried to determine if he was better than the program.
The professor then returned in 2011 and played the software 60 more times (eight games/day) and 56 more times in 2012. The purpose of the 2011 games was to determine if the program had been upgraded, resulting in different daily scores. The professor continued to play the game in 2012, creating a third set of data.
The Task
Determine if the program is about the same in 2011 and 2012 as it was in 1998, or if there appear to have been upgrades at some point. Also, determine whether the professor is demonstrably superior in play to the program.