I’m really busy at work so I have not had very much more time to look at the CEAC data today. However, I do have some thoughts to share.
First of all – as some are beginning to realize – it is VERY hard to predict final cutoffs this far away. Many things can happen between now and the end of the year. So – let’s take a look at the next VB and once we see that progress we can see how many interviews are being scheduled at the various embassies.
I also want to remind everyone that I cannot be 100% accurate, because I am not in charge of the DV program. In DV2015 for instance there were case numbers up to 89,000 – that year was massively over selected. I spent months trying to explain why I was sure we would see a lot of people miss out – and many just would not see the logic. Leading up to the final VBs I was convinced that AF region could hit 55,000, but I was concerned about numbers higher than that. In the final VB however, KCC inexplicably stopped AF region at 50,000. Months later I came to understand that my calculations were mathematically correct. There were enough visas for AF to hit the number I felt was safe, but the decision was taken to underfill AF that year, and globally also – and they missed the quota in AF. So – it just goes to show – we need to be careful about predictions.
Now – in talking about DV2015, that brings me to my point. Both DV2017 and DV2016 were underselected years. In DV2016 they only issued 46.7k visas because they underselected and then in DV2017 they underselected and then announced more winners in a second draw. Even with the second draw they still had far fewer selectees in than we have in DV2018, so we seem to be overselected. For that reason I prefer to compare this year to DV2015.
There are several ways to try and estimate final cutoffs – but there are many factors we have to take into account. Here are those factors listed.
- Quota for the region (set by a formula defined in last)
- Selectees numbers
- Starting derivative rate
- Response rate (some people don’t process their win)
- No show rate (some people don’t show up for the interview)
- Derivative growth rate (newborns and marriages increase the selectee numbers)
- Success rate (how many people get refused, or go on AP and cannot be approved in time)
- Processing limits such as the 7% rule
- Rule changes or political and economic influences – real and not real examples for DV2018 being:
- Travel Ban
- Trump is a moron so it puts people off living in the USA
- The economy crashes
I’m sure some people are more confused at what all that means – but let me give a MADE UP illustration. Let me say again – this is JUST AN ILLUSTRATION.
Assume I have case number AF45000. Because of the CEAC data and holes explanation in my last post you now know that there are NOT 44999 cases in front of you. Holes mean there are less cases than that – in fact in DV2018 there are 29336 cases in front of AF45000. So – already you are 15000 places in front of where you thought you were.
However, every case signifies the selectee AND any derivatives they have. They might have a spouse and 7 children, or they might be single. Until they are scheduled, we don’t know. So – we take the number of selectees in the published selectee numbers (49392) and divide by the number of actual cases in the CEAC data (32541). That means we have an average of 1.52 people for every case in Africa as a starting derivative number. So – by taking our example of 29336 cases in front of me at AF45000 and multiplying by the derivative rate (1.52) I can assume there are 44590 people in front of me. That is the length of the queue.
Now – because the formula for the quota is published, we know that the DV2018 quota for AF is about 21600 (based on 50,000 global quota).
OK – those were real numbers – now we move to guesswork. Again – ILLUSTRATION PURPOSES ONLY!
However, every time one of the selectees gets married, or has a baby, The queue in front of me get’s longer. That is the derivative growth rate. It varies by country/region – but for AF I have seen growth rates of over 20% – meaning the starting derivative rate becomes a rate of about 1.85 instead of 1.52. So now the queue is over 54,000 people!
Now – we need to know how many of those cases will be “non responses” or “no shows”. As desperate as I may be for my interview at case AF45000, there is some percentage of cases where people either never file their DS260 to say they want to process their case OR they do submit the DS260 but just never show up for the interview. In previous years that non response and now show rate has meant as many as half the queue don’t want their chance. So – maybe there are only 27000 people actually in the queue.
Not all of those 27,000 will be able to pass their interviews. Some will be unable to clear AP procedures before the end of the year, some will not meet the requirement, some will have lied on their eDV of “forgotten” they were married when they entered. The success rate is the rate of how many interviews actually get issued. It varies GREATLY by country – for instance in Ghana refusals are more likely than approvals. In Egypt you have to work very hard to get refused. So – then it becomes a matter of what countries those 27000 people are from. If they were all from Ghana I would be certain to get a chance, but if they are all from Egypt I don’t have a chance at all of getting one of the 21600 places.
See how complex that is? It is VERY HARD to predict the final numbers accurately. VERY HARD.
Now – to take all those into account takes time and extensive data from previous years. To an extent, some of it is guesswork. So – even if we really understand all the factors we STILL can’t be exactly accurate in predictions.
So – let me explain one quick way of doing the math.
Instead of trying to calculate all these numbers together I could look at how many cases were made current in a previous year to get within the quota. So – let’s do that, and give some other useful data.
For this method I will use DV2015. That was the first year the DS260 was used, it was over selected (like this year), and it is relatively recent. DV2016 and DV2017 are not good models because they were under selected. In underselected years some “odd” stuff happens. For instance, in DV2017 SA region received 1830 visas. That is over 400 visa MORE than the region’s quota for that year. That means SA took unused visas from other regions which will ONLY happen in an underselected year (like DV2016 and DV2017). DV2018 is NOT an underselected year.
OK – having said that – here are some numbers based on CEAC data from DV2015 and DV2018.
To explain, I list the quota from DV2018, published selectees per region and the number of cases for the region from the CEAC data. Using that, I have calculated the DV2018 starting derivative rate.
For DV2015 I give the TOTAL cases, and the final cutoff number. Then the cases that were in each region UNDER the cutoff and the number of visas issued. The final column (raw cases per visa issued) is a quick (approximate) method to apply the non response rate no show rate, derivative rate, response rate success rate and so on. By using ratio multiplied by the quota, we get to a ROUGH prediction of how many cases will go current in each region – ****ASSUMING**** all other things are equal. That is a BIG assumption – and an incorrect one for various reasons, but applying that method gives us the following numbers.
AF – 46600
EU – 33000
AS – 9600
OC – 1300
Now – I consider these LOW NUMBERS. These are what I would call safe – to a lesser or greater extent per region. In reality there are a number of factors that this method does not take into account.
AF region seems like a sensible safe number. It is broadly similar to my earlier thoughts. So – I consider that a good “lower number” range for safe numbers. That does NOT mean the number won’t be higher, but I would be surprised if it ended lower than 46600.
EU number is much lower than I would expect to be the final number. This year three countries (Russia, Albania and Turkey) experience a HUGE increase of selectees compared to previous years. That suggests to me that agents were somehow involved in registering entries. When that happens, the number of non responses increase, refusals increase and so on, because agents typically make things worse, not better – especially when their motivation is greed. Added to that the case density in EU is very low. So – from 33000 to 37000 for instance, there are only 1444 cases. So – it is not hard to imagine the number being MUCH higher than 33000. I would be SHOCKED if the number was anything like 33000 – I still expect the final number to be in the 36/37 range (a little reduced from my earlier assessment since max case numbers are actually lower than previously reported).
AS region is very affected by the travel ban. Iran, Syria and Yemen are all under the ban – accounting for nearly 40% of the total selectees in the region. Those selectees cannot be disregarded from the DV lottery. Some of them will not be subject to the ban because of dual citizenship and other reasons. Because of that, KCC can’t simply ignore those 6000 selectees. BUT I don’t know how KCC will treat the shortfall in visas being issued because of the travel ban impact. So – I would pretty much suggest we totally IGNORE the prediction for AS region. There is just no way to know what will happen. So – WAIT AND SEE is the only sensible answer.
For OC region I still believe the final cutoff will be between 1400 and 1500. The method above produced a lower number, but visas issued per 100 cases in OC region is low anyway – so the method could easily be wrong.
SA region again strikes me as low for the same reasons as OC. In addition, this year has a new country, so that has altered the game somewhat. However, the quota will be enforced this year, unlike DV2017, because there will not be spare visas coming from other regions.
I will be spending some time refining these numbers, they are best guesses at the moment, using one particular (approximate) method. NO ONE should bet the farm on these numbers. I could be wrong – just as I have been before.
Now finally let me just say, I provide this analysis not as numbers written in stone, but rather to allow analyitical people to understand what is going on. Many will be lost and confused by this whole post. I understand that, but I get frustrated when someone doesn’t even bother to TRY to understand and just wants ,e to tell them whether their number is safe. I expect people to use their own common sense – and of course feel free to calculate your own numbers using your own methods.