I have mentioned before that there has been developed a really good, statistically robust model that describes Oregon Coastal Coho survival. It won't work as a predictor because the 4 ocean variables are so independent that the fourth (about the time the fish hit the bays) can take a bad year and make it good or a good year and make it bad.

If that is the case in WA then the best we can hope for for accurate numbers is in-bay/river. For the ocean, some very conservative fishery may be possible.

It should be obvious that major ocean fisheries are based on shaky assumptions that won't get any stronger no matter how much computer power is thrown at it. The manager's hope is that there won't be gross failures of return. Blow the fishing and you piss off fisherman this year. Blow the escapement and you screw the future.