OK, as requested here is the updated graph to 4 March 2015:
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FB_ROCy0XEAAwWDI.png%3Alarge&t=665&c=2MNXU3Wf9lLE0w)
Now clearly we are falling behind the original trend, so the model is falling apart (I can't help but notice how much this graph actually is starting to look like Wobber's prediction!). However, I think there is room to wiggle with. For example, if I align the initial high points of the second peak in each cycle, we get a 694-day cycle which is still OK:
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FB_RRvDQUwAAm9MN.png%3Alarge&t=665&c=WIJj1VNYhQ6iaQ)
In this model, the run-up to the second peaks align across both previous cycles, and the general run-up to the first peaks also is more or less in agreement. While we have been in a decline over the first part of the current (third) cycle, if we are indeed currently seeing a turnaround and a general upswing that would be in agreement with this revised model.
As I said before, I am mainly having some fun with this. The data is what it is, but I am not (yet) convinced that this is really indicative of a recurring pattern, although I still believe there are natural patterns that will occur in nature (that is the basis for all TA, now, isn't it?)