Cool...
I'll run the Numbers
thru the Curve-O-Matic 2000 and see what it Spits Out.
OK...
I got out the Curve-O-Matic 2000 and it Wouldn't turn on. Grrrrrrn It. The
batteries were Dead.
So I did some Quick and Dirty Scatter Plots of the Post Counts in Excel. Here is what it looked like ...
http://farmerengineering.com/ECF/All_Posts.jpg
The first thing I Noticed, which was a Surprise, was how Linear the Plot Looked. I was Expecting kinda a 1/2 Smile type Plot. I slammed a Regression Line on the Data Points a Ga-Zing, Not a Bad Fit. "Check out the Big R^2 on that Line".
Some would say that a Regression Line with that High of an R would be Good Enough to do a Projection. But Check out the 2013 Portion of the Plot. I see a Little Wiggle on the Squiggle. And the Ratio of the Range to the Domain is Huge. I mean Ginormous.
So I
threw out the 2012 Data and did a New Scatter Plot.
http://farmerengineering.com/ECF/2013_Posts.jpg
Much Better. But another Surprise. The Plot was Increasing (No Surprise) but at a Decreasing Rate (Big Surprise). Clearly this Plot needs a Higher Order Function to do a Projection, so I tacked a 2nd Order Poly on it and got another Outstanding R^2.
But the Poly looked Wacked because it was in Engineering Notation. And the Numbers Didn't make and Sense. It was those Damn Date Formates I used for My Domain. I couldn't figure out how to Edit or Reformat the Display. So I applied a 1/1000 Scalar to the Post Data and then Converted the Date Data to Julian Data Format. (Maybe it would have been Quicker to get Batteries for the Curve-O-Matic 2000?)
So solving the Pretty Poly for 10,000 (Ten Million) I got Two Solutions. 296 (Rounded off) and Six Hundred and something plus change. Obviously the 6xx result was on the Other Side of the Horizontal Tangency of the Curve. So let's Toss Him in the Bin.
So for the Data I got, the 296th Day of 2013 should be about when we Hit 10,000.000.
Give or Take 2 Years.