Quote:
Originally Posted by NUKaydee
It is common on my campus that the strong recruiting chapter have slightly less than 2x quota at their pref parties because the RFM specialist told them they had to release X number of girls. One of the strongest chapters on my campus has had to snap bid or COB up to total a few times because of this.
|
Okay, so if the prediction is that more than half the women will rank this group as their #1, they absolutely SHOULD have less than 2x quota at their pref parties. The question is why it would happen that the prediction should be off several times in a short time frame.
The problem, of course, is that formulas are not humans, and while they can make good predictions, they sometimes miss the obvious. Let me give an example:
Let's say there are 10 groups on campus, and return rates for the first invitational round at XYZ are 85% in 2009, 90% in 2010, and 95% in 2011. A basic trend projection would obviously predict a 100% return rate in 2012. But you and I, who are human beings with real thoughts, know that no chapter is ever going to achieve a 100% return rate. In fact, it's very possible that the chapter will top out at 95%.
Further, the RFM models are all based on past years' data, i.e. they are strictly time series models, not associative models. They don't know that XYZ's main competition just built a gorgeous new house or there is a rumor going around that XYZ is on social probation, or any number of other things that can cause a chapter to over- or under-perform in a given year.
So, in short, forecasting is very, very difficult, and the outcomes are not perfect, but RFM is miles ahead of its predecessor and has done much to level the playing field.
Hybrid Mode
