Before I went on my rant in early Feb then subsequently took some time away, I noted in the Philly forum to tread lightly with the weeklies and the epic looks. Of course I got booed off the forum even suggesting that. But in all seriousness I will reiterate what I did at that time. Many of these extended products are heavily weighted on past analogs when there are rolled out. So if a group of indices like the QBO, PDO, enso etc match a particular year (let's use 2010 as an example because this year had many similar key features), then the weeklies would tend to spit out something similar that transpired during those analog years. Not saying this is sole reason we got the head fake, but partly is due to this, of that I have no doubt. Many of us actually discussed this in the past and there is truth to the algorithms involved in creating an extended product. Now could I tell you which specific products use more of the analog data? No, I dont have that I fo, but maybe a red tagger would?