...
I'm just wondering whether there is any merit in comparing 2-year periods.
...
I'm also a bit surprised to see how consistently close the curves are from one another at certain non-peak intervals.
...
There are many people on the Forum and on the ASIB who are quite sold on the idea of cyclicity, and the hypothesis that there is a recurring 5 year pattern has several vociferous advocates. Personally, I'm not convinced - either way.
Humans are great at seeing patterns, and that is why citizen-scientists (and the general public) are so useful in projects which require pattern recognition. See these two sites for example...
https://www.zooniverse.org/abouthttps://www.galaxyzoo.org/Unfortunately, this propensity towards pattern recognition is so strong (an evolutionary imperative perhaps???) that we can sometimes see patterns that are mere chimeras.
As you know, 2007 and 2012 saw jaw-dropping losses near the annual mins, and, should the 5-year pattern manifest again next year, then 2017 will be something to behold.
However, as I said, I'm unconvinced that there is any "
casus belli" to support the hypothesis. Basically, I think the noise level is still too great to expect any accuracy in short-term prediction.
A few years ago, both Rob Dekker and myself wrote articles on Neven's blog on the subject...
http://neven1.typepad.com/blog/2013/06/problematic-predictions.html#morehttp://neven1.typepad.com/blog/2013/07/problematic-predictions-2.htmlShould 2017 follow a similar pattern to 2007 and 2012, then it might start to show up quite quickly on a 2-year chart of the type provided above.
...
Regarding your comment about how those curves looked close except near the min/max turning points, I suspect that is more to do with the ways our eyes interpret such curves.
I've tried to demonstrate this effect in the attached chart. This shows a full two cycles of three distinct sine waves. The sine waves are identical in frequency, amplitude and phase, but there is a constant offset of +0.05 applied to the red, and a similar offset of +0.1 applied to the green.
I suspect that most people "see" the separation as being biggest near the turning points - even although it is identical throughout.