I believe the thread title makes the issue seems a lot easier than it actually is. I am trying to forecast how many applications our office with receive by the end of the application period by comparing how many we have today to the average of the last four years (generally speaking). Last year this worked fine because there was not much variance in the date (couple days) at which the application period started in the first three years. This year the application period opened nearly an entire month earlier. Also the priority application period has ended on November 1st for all prior years and will as well this year. After that date the numbers are fine, however because the forecasting needs to consider the entire application period to predict how many applications we will have received in entirety when the year is said and done creates an issue. I've tried normalizing the data based upon the percentage of days the application has been open relative to the entire period, but that did not work. I feel as though there may be some way in which solver can help me with this issue without having to analyze trendline equations for previous years? I'm sure many of you have encountered this issue and your direction would be much appreciated. I want to be sure that I forecast with utmost accuracy and do not want to rush and roughly draw conclusions about how previous years differed. Certainly it is also difficult because no one can actually know whether starting the application a month earlier actually results in individuals being more proactive about applying, but one must consider it when determining if the organization is on track to reaching the goal (which the forecast informs if will be achieved). The earlier applications are not simply a "bonus."

Thanks in advance for your help. Here are the application period dates for each year to help give you more insight (Fall 2015 being the current application period):
times.PNG