Thanks very much for that. It now shows about 58,000 visas would be needed to satisfy the likely demand (based on the response rate, interview rate and issued/abandoned/denial/admin rate being the same as 2013).
Sloner, before you say a thing - those three things (response rate, interview rate and issued/abandoned/denial/admin rate) have NOTHING to do with the software used to select the winners.
Given that the CP numbers need to fit within the 50k LESS the AOS cases it looks like there will be about 10k too few visas. So - if we wanted to predict the visas issued then we would expect about 20% less than your issued number for each region. Correct?
I would reduce that by a bit because of the CEAC data concern I have explained above - so lets knock everything down by 15%
AF - 20625
EU - 16867
AS - 9739
OC - 939
SA - 1255
Those numbers sound about right to me. They come in just under 50k BUT don't leave a lot of room for AOS, so I suspect these would be about the highest we could possibly see - likely a tad less in some regions.
However, when I do the same exercise on the max case number interviewed column (which does not influence the projections above), it doesn't look right - and I think that goes back to this odd thing about the CEAC data cutting off too soon.