I built a Shiny app recently as part of an analysis of U.S. crime data from the UC Irvine Machine Learning Repository. I was doing a multiple regression analysis to find factors that correlate with upticks in community crime rates. I ended up making a game out of the analysis to make a point about how policy levers can drive changes in crime rates.
The goal of the game is to get the user to minimize the U.S. crime rate by moving levers related to family planning, employment, immigration, and housing. A user can learn about what really drives down crime by actually making policy decisions, rather than reading an interpretation of regression coefficients.
I also posted the data tables, some plotting functions, and other information in the link.