StatsVille is a game where the player assumes the role of a city health administrator. On their first day on the job, they have been notified that the city is faced with a potentially severe epidemic that they need to stop.

In the initial stages/levels of the game, players are asked to determine an appropriate treatment strategy (i.e. whether treatment A or treatment B will more effectively stop the spread of the disease). Players can simply make choices based upon the visualizations, but they will develop better strategies (i.e. improve their chances of winning) if they thoroughly understand the data and techniques for analyzing it. Players get daily feedback based upon their strategy. If they have a good strategy they will see that they are stopping the spread of the disease.

Multiple levels of the game are developed by varying model parameters. In the initial stages, the game is fairly easy to win, but at each level the game gradually advances in difficulty and requires students to build upon strategies used in previous levels.

The following link allows you to play the StatsVille Game.

You may be asked to install Unity Web Player, this may take a few minutes.

Many browsers will require you to allow popups before they will run these stat2labs games.

These games may not run on the most recent versions of Chrome or Firefox. We are currently working on fixing the issue.


Learning Goals and Class Discussion:

  • Students should see how larger samples sizes provide more accurate estimates of than small sample sizes.
  • The last activity in this lab asks students to read about the ASA’s statement on p-values at ( Note you can download the pdf for free. Students pick one of the six principles and discuss how it relates to this StatsVille activity. More than one of the principles can apply to this lab.
  • Addressing the challenge of multiple hypothesis tests conducted on the same population.
    • The game is designed so that each day is independent of other days. The effectiveness of Treatment A and Treatment B do not depend on the day. However, what cautions need to be considered when multiple tests are conducted from the same population?
    • While day-to-day samples are independent, explain why using cumulative samples each day is not independent.
  • Discuss the use of graphs and statistics. Are counts (number cured) or percentages (percentage cured) a better way to evaluate a treatment’s effectiveness? Can multiple tests be used to evaluate this data? Should a two-proportion test be used with this data? Could we use a Chi-Square test or t-test instead?
  • The spread of the disease follows a basic SIR (susceptible, infected, and recovered) differential equation model. We then use an underlying binomial model to simulate natural variability in the cure rate. Knowledge or discussion of the SIR model is not needed in any part of this lab.


LAB 1: Multiple Two-Proportion Tests: Student Handout

Additional Game Instructions: StatsVille Instructions.


This activity uses the StatsVille game to demonstrate the challenges of using multple hypothesis tests to draw conclusions. Students recognize that each random sample from a population varies, and this causes the corresponding p-value to also change with each sample. The emphasis is on using data to draw conclusions, even when the standard assumptions for hypothesis tests may not hold.


Data visualizations for all data from Level 1 is available at StatsVille App.

Data visualizations for all data is available at StatsVille App.


All data from the game is available at StatsVille Data.



Thanks to Professor Anya Vostinar and Grinnell students Mariam Nadiradze, Tianhao (Mike) Zou, Ritika Agarwal, Jimin Tan, Jemuel Santos, Hoang Cao, Kevin Connors, Houfu Yan, and Anaan Ramaay for creating, editing and maintaining the on-line game. Thanks to Yuanqi Zhao, Yuyin Sun, and Matthew Palmeri for assistance with the data visualizations.