Registration 2017

[Jon Stratis]

Ok, I've rewritten this like 5 times, I keep going back and forth.

The first one could be a useful comparison next year with the new preference system Frank has talked about in the blog - does the introduction of a new system for initial event registration significantly change how registration goes? But at the same time, I feel like it probably overstates what would have happened if registration proceeded normally this year, and could lead to false conclusions.

The second one I like for the simple fact that it represents reality. It shows that, despite being a week late, the initial surge "made up" for that lost week of registration immediately.

Looking at these graphs today, I was struck with a burning desire to see a third graph (or a third and fourth), one showing the difference, day to day, in registrations between years. So one line that shows (2017-2016), another for (2016-2015), etc. each day as registration proceeds. Does that difference generally stay the same? Does it increase at a predictable rate? Are there sudden surges some years that we can't explain? And, perhaps most appropriately, how does it show the difference between this year and last year?

[Mark McLeod]

The second plot is not actually reality either, so you may want to see a fully uncorrected plot.

Opening day in the 2000's up until about 2014 was typically halfway between this year's delayed opening and last year's opening. That's why I've been using Oct 3 as a normalized starting point for past years.