Nice long term effort to help try and solve the problem insofar as prediction goes.

I've a few questions:

1) Did you formalize your method (& put it in writing too)?

2)Did you publish anything on it ?

3) Did you use any statistical procedures to see if the method you were using was legit ? Otherwise you might have been chasing an illusion of sorts (thinking that you were seeing precursers instead of an early warning or cause and effect sequence...).

I agree that lots of signals are being dismissed as noise...but this requires documentation to make one's case. Are you going to do this? A publication on this could trigger others (with the resources) to study things further....

Yes...the thinking 'outside the box' theme has been harped upon a bit before...glad we agree.

Just my .02 worth