This is going to be a long post and possibly difficult to understand – especially for people not comfortable with English. I have included links to other blog posts so it is helpful to read those other posts as you read through this one.
OK, I have been “studying” DV for a couple of years now – and my knowledge about the subject is mainly gained through my interactions with others, reading of the laws etc. and having access to the CEAC data (thank you Rafik!). Without looking at the full breadth of information we can only make wild guesses – rather like being put in a large pitch black room and being told there is an animal in there and being asked to guess what animal it is without turning the lights on. You can get clues and bit by bit you get the whole picture.
Well yesterday USCIS gave us a flashlight. It is still a dark room, but we can see a lot more. What was the flashlight? They published the CEAC data to their website and using a computer program I have extracted all the data as a snapshot. As the year progresses I will publish updates to the data at least every month, but faster if possible as my time allows.
CEAC data for DV2015
Last year the CEAC data was useful but we only got information as the cases were scheduled. This year they have loaded a lot more data so we can “know” a lot more – or at least confirm things that I have previously described in theoretical term and explanations. The flashlight let’s us see a bit more….
First let me show the density charts for the five main regions (I ignore North America generally because it is too small to be significant to the process).
So what can we see from the charts
If you have read my post on holes theory then you should understand the DV draw process. To get to the 125k selectees that were informed of their win, USCIS would have started with more cases identified as winners. There would then have been a verification process looking for duplicate or bad entries – and those cases create “holes” in the numbering. That is why the numbers from CEAC have gaps in the numbering. If you look at the charts you can see the disqualification rates for each region. The holes rate should be (more or less) constant throughout the region because the disqualified cases would be randomly distributed throughout the region. So – the holes rate for a given region is the difference between the cases per thousand (or hundred) and the one thousand mark (or one hundred). Showing this graphically for Africa as an example the disqualification holes are marked in yellow on the graph below:-
So – the disqualification holes rate for AF is just over 20% – meaning 1 in 5 cases in Africa were disqualified and never got notified that they were winners.
The draw process disqualification rates for each region are:-
AF – 22%
EU – 30%
AS – 5%
SA – 7%
OC – 8%
In reality this rate would vary by country. Some countries would have almost no disqualification rate whilst others would have very high rates. The reasons for that difference are cultural (some countries have more “creativity”) and also there is an influence of unscrupulous/clumsy agents. Agents in a place like Nepal for instance are charging low fees and don’t seem to try to cheat/blackmail people, so the entries are genuine and rarely disqualified). In Ghana however, there agents are clearly using unscrupulous tactics that will cause their high refusal rates (i.e. many failures during interview), but they seem to at least be passing the initial disqualification phase better than their counterparts in Ukraine. As I have described before, the Ukrainian and Ghanaian derivative ratios are indicative of agents being involved in the process (because they will often register an applicant “proactively” without family members.
In my draw process post I describe the draw process and point out that some countries are limited during the draw process. I have explained this in further detail for some of the limited countries such as Ethiopia, and Egypt.
The limited countries are Ethiopia, Ghana, Egypt, Ukraine, Uzbekistan, Nepal and Iran. If we look at the density charts for the three large regions we can almost pinpoint the cutoff points for these countries. I have increased the detail of the grouping in a couple of cases to make things clearer.
Remember these are where the countries hit their limits during the draw process – the VB limits for these countries could be lower.
It is hard to know which country is which in EU – one cuts off just below 14500 and the other at about 15000.
For Africa there are 3 certain cutoffs as shown above (Ghana is more like 16XXX) and a possible cutoff for another country which I suspect might be Cameroon (it could also be Liberia or DR Congo). This suspected cutoff happens around 53k and the reason I did not know about it before is because I don’t have the entrant numbers for DV2014 and DV2015 (only 2013 and before). That would indicate a sudden increase in the entries for one of the countries (Cameroon, Liberia or DRC) – possibly because of agents operating in that area.
Just a quick note about the cutoffs. IN each year there seem to be a handful of cases that are “outliers” – a country is limited but you find one or two cases way outside the normal cutoff. Those cases can sometime be explained (as someone who is charging to a non limited country) BUT there have been a couple of cases that were unexplained. That doesn’t ruin the theory – it simple shows the limiting process is not perfect.
SA and OC regions
There are no limited countries in these two regions.
The Nigeria impact for Africa
As I have explained in this post before the exclusion of Nigeria has affected the case numbers. Before seeing the CEAC data I could not be sure how big was that impact, but the impact is more clear now. The highest AF case in the file is 89799. There is a chance they have cut that off slightly early – but given the number of cases for AF region is 39.2k, I actually think we have pretty much every case (let me know if you feel differently).
DV2014 has case numbers as high as 116k. So – there has been a 26k plus reduction in case number – or increase in density. As I mentioned in the earlier post this means that a medium high case number for DV2015 is not as low as it was in DV2014 – so the AF max case number in DV2015 will not reach as high as the 81100 we saw in DV2014. I am not certain where the axe will fall and the DS260 change could have an impact, but it seems sensible to consider that cases above 70k are in a risky range.
Final year ending case number predictions
OK – this is hard to be accurate this early in the year, but the completeness of the data makes things a little easier. The DS260 fiasco would certainly have had an impact, but as yet is isn’t clear what impact that could be. The quotas will have been increased as explained in this post. Local political turmoil, Ebola, and the ability to resolve AP cases will all have an impact, but that being said – let me make some rough guesses in the areas where my understanding has changed based on seeing the data.
I expect a final cutoff around 68XXX to 74XXX. This far out that is as close as I can get – so don’t bet the farm on these numbers – I would not give up at this point, but at least higher numbers can get their backup plans in place. We know a heck of a lot now but most crucially we do not know the “non response” rate – the number of people that won’t even return their forms. That rate is significant – especially ion this year with the DS260 introduction.
Of the countries that experience a slowdown during VB progression, I expect all Ethiopian selectees to go current and almost all Egypt selectees (there might be a small cutoff in place at the end of the year for Egypt at around 27500).
I think Asia will be largely a repeat of last year. However, that assumes Iranian AP cases are held back again as we saw in DV2014. If those cases are resolved faster (not very likely in my opinion) than last year the AS final number could be smaller than last year.
Nepal and Iran could be limited – and I see Nepal hitting the limit again this year perhaps being limited in the 8600 to 8800 range (compared to the 9500 limit in place at the end of last year).
So – Asia should reach around 13XXX if Iran behaves the same way or as low as 11XXX if Iran AP cases resolve faster.
I was very surprised to see the 45k high case number. It almost feels like a “mistake” – but if true it represents an increase in density (probably due to agents in Ukraine and Uzbekistan). Up to now I have been assuming that the quota increase for EU would mean a slight increase in the final case number above last year (40150). However, given the density and lower max case number I now feel that we might fall a bit short of the 40k mark. So – I would guess a final number in the 38XXX to 41XXX range (allowing for a margin of error).
OC and SA
There were no big surprises in the CEAC data for these two regions so my estimates are unchanged (OC 16XX to 17XX and SA around the same as later year).
Final comments about case number predictions.
I am very aware of how important the DV lottery is to people. I am not looking to cause anyone to lose hope – I am simply trying to offer informed opinion. I could certainly be wrong and I would be delighted to see numbers go higher than the amounts I have said. I remain utterly supportive of all selectees, whatever your case number so if you have a high number and want to ignore my predictions I will totally understand that and will happily help in any way that I can.