My way to execute this algorithm is to try to get as distant as possible to the problem in step 2, taking as much experience into account as possible, and then

My way to execute this algorithm is to try to get as distant as possible to the problem in step 2, taking as much experience into account as possible, and then hap-hazardly zooming in. Or something like that. And it seems to work.