Ask and ye shall receive 3yards .3yardsandacloud said:Working from my daily usage numbers, buckiprof did some calculations and just ran some "interesting" numbers (read as "scary") by me in a PM. Hopefully he'll post them here.
I see prof just posted ... ooops. Please put the numbers up here buckiprof. And thanks for extending the anaysis.
Excellent stats 3yards! I couldn't resist playing around with them a little bit, so I started by determining averages for each month. They are:
May 2004 144.52
June 2004 161.9
July 2004 227.77
August 2004 266.77
Septmeber 2004 280.77
October 2004 308.1
November 2004 354.4
December 2004 422.71
January 2005 515.55
February 2005 965.93
March 2005 783.29
April 2005 752.73
(Of course April 2005 is the average to date.)
Making a plot of this averaging data, time vs the averages, and excluding the February 2005, since it can be considered an outlier, and you have a plot that is damn near a quadratic. In fact, excluding the March 2005 data as well, yields a plot that can be modeled by a quadratic regression model that has a coefficient of determination of 0.9899. In other words, the quadratic model fits the data with an accuracy of about 99%.
So what in the hell does that mean?!? Well, if (and as always in making a prediction it is a big if) the growth continues following this model, we can expect the average in October 2005 to be about 1291.5. Whereas if we employ a little calculus and assume the rate of growth for April (as given by the model) continues for the foreseeable future, the predicted value for October 2005 is about 1205.3.
Again, there are some big if's involved and assumptions made and I am by no means saying that this is what is going to happen for October 2005. There are simply too many unknowns/variables involved and the calculus utilized is good for making a short term prediction, say one month into the future, but I decided to jump to the heart of the football season
(Oct. 2005) anyway for illustrative purposes.
Upvote
0