Anthony Lawrence Southwestern University Case Study B Problem: Southwestern University is experiencing a quickly expanding football program. As a result, attendance for home games is increasing and approaching capacity. It is in the best interest of SWU to forecast attendance to aid them in deciding when the best time to expand the present stadium, which now holds 54,000. Data: The following data is from the past six seasons, 2002-2007. Game Year – Game – Opponent 2002-1 Baylor 2002-2 Texas 2002-3 LSU 2002-4 Arkansas 2002-5 USC 2003-1 Oklahoma 2003-2 Nebraska 2003-3 UCLA 2003-4 Nevada 2003-5 Ohio State 2004-1 TCU 2004-2 Texas Tech 2004-3 Alaska 2004-4 Arizona 2004-5 Rice 2005-1 Arkansas 2005-2 Missouri 2005-3 Florida 2005-4 Miami 2005-5 Duke 2006-1 Indiana 2006-2 North Texas 2006-3 Texas A&M 2006-4 Southern 2006-5 Oklahoma 2007-1 LSU 2007-2 Texas 2007-3 Prairie View A&M 2007-4 Montana 2007-5 Arizona State Attendance 34200 39800 38200 26900 35100 36100 40200 39100 25300 36200 35900 46500 43100 27900 39200 41900 46100 43900 30100 40500 42500 48200 44200 33900 47800 46900 50100 45900 36300 49900 Anthony Lawrence An important thing to note is that the homecoming game of every season is the second home game (bold), and is always well attended. Also the forth home game always corresponds with a local festival that always draws from attendance (italics). Summary of Forecasting Methods: Below is a table of the forecasting methods. The correlation coefficient, bias, mean absolute deviation (MAD), mean squared error (MSE), and mean absolute percent error (MAPE) are shown. Naïve Moving Average (3 periods) Weighted Moving Average (3 period; .6, .3, .1) Exponential smoothing (alpha = 0.5) Trend Analysis Seasonal Additive Decomposition Correlation -- Bias 541.38 MAD 6865.52 MSE 69,856,200 MAPE -- 491.36 6,138.27 59,540,560 .17 -- 424.81 6,501.58 61,107,180 .18 -- 794.28 5,880.56 50,755,960 .16 .54 0.00 4,355.70 31,285,700 .12 .97 0.00 1,251.26 2,386,650 .03 .19 It is obvious that the superior method is seasonal additive decomposition. This makes sense because of the cyclical nature each season follows, largely due to the homecoming game and local festival. The other methods cannot take those variations into account, and this adversely affects their error. Because of this they are not suitable for predicting at which point the demand of attendance will surpass the stadium’s capacity. They will show when the average or smoothed demand will reach capacity, but this will occur after attendance for one or two games will consistently surpass capacity. Seasonal additive decomposition is definitely the best method, and all indications support that. The bias for trend analysis is just the same as seasonal, however the correlation coefficient is not acceptable for this method, well below .70, but the seasonal correlation is outstanding at .97. So, now the management has a decision to make whether or not it is worth putting off expansion of the stadium until the average attendance of a season is above capacity, and losing potential ticket sales for the games that are above capacity, or to go ahead and expand before capacity is ever reached in a single game. Anthony Lawrence Forecasting: The following table shows the forecasted attendance of the next two seasons of home games. Game 2008 – 1 2008 – 2 2008 – 3 2008 – 4 2008 – 5 2009 – 1 2009 – 2 2009 – 3 2009 – 4 2009 – 5 Attendance 46540.23 52555.73 50254.56 38370.06 50202.22 48784.39 54799.89 52498.72 40614.22 52446.38 According to this analysis, the second game of 2009, the homecoming game, will be the first game that the demand for attendance will exceed capacity. The management can wait until after the 2008 season to start the completion, so long as it is completed before the 2009 season starts. One thing they should consider is that a refurbished stadium would cause a spike in attendance because fans would want to see and experience the new facilities. With this in mind, these figures could be much lower than what would actually happen if an expansion were built. Revenues: Ticket sales will average $20 in 2008 and $21 in 2009 due to a 5% price increase. Total sales can be found for each season by summing the attendance from each game and multiplying by the ticket price. The results are displayed in the table below. Season 2008 2009 Total Attendance 237,922.8 249,143.6 Revenue $4,758,456 $5,232,016 Again, it should be emphasized that depending on when the expansion is built, these values could be lower than what actually comes in. It would be smart for them to build the expansion sooner than later, so that the spike in ticket sales will come sooner. This could offset some extra costs that could incur from building it sooner.