I posted a discussion prior to the epic collapse of how I thought the weeklies were deriving at those epic looks. Then I reposted again after they collapsed. They were based on algorithms programmed into guidance based on analogs from seasons past which matched things as pdo, qdo, enso as well as current data when models/weeklies were ran. With that said, if we had similar conditions to say late Jan 2010 when those weeklies were ran, a heavily weighted analog was used thus showing us something which looked like Feb 2010. Honestly it wasn't surprising things flipped but I played along for a bit with the positive vibes gang because I actually enjoy posting here. Just have to walk on eggshells at times.