On the Tigers project, NEAS adds a comment explaining the negative autocorrelation at lag 1 of the first differenced time series as resulting from the draft.  However, even the best draft picks typically take longer than one year after they are drafted to impact the team at the major league level.  Further, the draft order is not based on a team's improvement (first difference), rather it is based on the winning percentage (original time series).

Chicago Bears football team TS sproj

This student project examines the won-loss record for the Chicago Bears.

The adjusted R² decreases for AR(3) and longer ARIMA models, even for football, which has many players and a slow shift in team quality. The AR(2) and ARMA(1,1) models don't improve predictive ability, even though they do better for the in-sample goodness of fit tests. Your student project need not give a complex model. Explain the statistical techniques you use – adjusted R², p-values, F test, Durbin-Watson statistic, Bartlett’s test, Box-Pierce Q statistic, mean squared errors, types of forecast errors, and so forth – and what each one shows.

This analysis of the Detroit Tigers won-loss record fits simple ARIMA processes in Excel. The analysis is very good, and the candidates selects the best model.