Students Step Up Their Game Using Statistics

What do space warriors use to get an edge in battle? Statistics.

Zach Hackett and Ben Fischer, students at Ocean City High School in Ocean City, New Jersey, needed an end-of-year project for their statistics class. Knowing that their teacher, Mrs. Cox, loved letting her students be creative, they decided to analyze one of their favorite video games, Star Wars: The Old Republic.

“It was something fun for the final project of the year,” Fischer says. “We thought it would round it off nicely.”

Students Zach Hackett and Ben Fischer used Minitab to examine the impact of "adrenals" on their video game performance.

Hackett and Fischer sought to determine whether “adrenals,” in-game items intended to temporarily improve damage and healing abilities, significantly increased mean healing and damage output. As in many games played collaboratively online, healing your allies and damaging your enemies is important in Star Wars: The Old Republic.

To assess the impact of adrenals, Hackett and Fischer used Minitab Statistical Software to perform two-sample t-tests. A two-sample t-test compares the means of two samples to determine whether a difference between the samples indicates a difference between the populations they were taken from.

Hackett and Fischer performed two two-sample t-tests, the first to see whether a significant difference existed between mean healing with and without adrenal use, and the second to see whether a significant difference existed between mean damage with and without adrenal use.

They began by making sure that their experimental data would meet the requirements of a two-sample t-test: the data had to be random, independent, and normally distributed.

When a character in Star Wars: The Old Republic performs a healing or an attack, the code of the game randomly generates a value influenced by factors such as the character’s weapon, so the students’ output data would indeed be random. Because each random output value is not influenced by previous values, the data would be independent. Third, because they planned to collect more than 30 observations for each sample, a statistical property called the central limit theorem states that the data would be approximately normal.

Having established that their data would meet the requirements for the two-sample t-tests, the students performed the same healing and the same attack 50 times each without adrenal use and 50 times each with adrenal use. They recorded the healing or damage outputs for each of these 200 trials.

They then used Minitab Statistical Software to calculate descriptive statistics, such as the sample standard deviation and sample mean, which are necessary for the t-test calculations.

From these statistics, they calculated t-values, which they used to determine whether the differences between the sample means indicated significant differences between the population means.


The boxplot of heals with and without adrenals reveals their impact on game performance.

These t-values showed that it would be highly unlikely for the differences Hackett and Fischer found in their sample data to arise out of mere random chance, so they concluded that adrenals significantly bolster performance.

Acting on this newfound knowledge, Hackett and Fischer have made adrenals a bigger part of their in-game battle strategy.

Hackett says that using Minitab Statistical Software was efficient. He and Fischer particularly appreciated Minitab’s graphs and the ease with which they could refer back to previous output and avoid having to recalculate statistics.

Their project not only improved their tactics in The Old Republic, but also gave them a better appreciation of statistics. They said it was a testament to how great their teacher is.

Hackett said that it can be hard, sometimes, to see the practical applications of math, but that their project helped with that.

“This experiment solidified in my mind that statistics is one of those areas of mathematics that actually does have real-world applications,” he said.

By using this site you agree to the use of cookies for analytics and personalized content in accordance with our Policy.

OK