DV Lottery 2012 July 15th statistical breakdown

[Concerned_]

It looks like the numbers are really intermixed between regions. One AS number is really low, below 4000.
It could be that numbers are going in the increased order according to country of birth. From 1 fo N1 - Europe, from N1+1 to N2 - Asia e t.c.
And later letters are assigned according to native country that could be different from country of birth. That could be why a small AS number is among EU numbers.

generally it looks like numbers 1 through 31000 are EU numbers. 32000 through 45000 AS numbers (about 15000) e t.c.
Very logical. Few exceptions could be because of differences between country of birth and country of chargeability. The case AS3753 could be explained as the primary applicant born in EU but took chargeability from AS (by spouse or parent). That is why his number 3753 got AS letters in the end. The same for AS15893. Or vice versa. It could be they are chargeable to Europe, but country of birth is AS (it would be logical to change chargeability from AS to EU, but not from EU to AS)
Dig for cross-region chargeable cases from forums for checking the idea.
That would be in line with the idea that each number corresponds to a person, not a case.
That would also mean that if case 1 corresponds to a family of 4 people and gets number 1 (country of birth EU, letters are assigned later, according to country of chargeability), numbers 2,3 and 4 will not be allocated, the next allocated number will be 5 (country of birth EU, letters are assigned later, according to country of chargeability)

If it works this way, then why do all DV numbers on visa bulletin begin with plus or minus 8000? If your theory is correct, then majority of applicants from AS region should have case numbers above 30K. So they will not get their interviews until April or so? Same thing with Africa.

 

[tmfl]

http://forums.immigration.com/showt...umber-Or-Low&p=2295365&highlight=#post2295365

http://forums.immigration.com/showt...umber-Or-Low&p=2295919&highlight=#post2295919

I was born in Japan and I've never had nationalities except Japan, I was single at the time of DV2010 entry

 

[raevsky]

2011EU25201
(snip)...
2011AS25201

then why do all DV numbers on visa bulletin begin with plus or minus 8000?

That would really mean numbering in different regions is completely independent one from another. That is also said in the description how numbering is done.

Allocation method of CN has been changed since DV2009, but at this point nobody knows if CN allocation returns to that of DV2008 or before, or DOS would take new algorithm

How do you know it was changed in 2009?

Obviously they have a lot of spaces between numbers (unused). That is probably a property of selection algorithm why they have them.

Still strange EU has only numbers below 31000 while other regions have much more than that

 

[deepfrrezed]

If it works this way, then why do all DV numbers on visa bulletin begin with plus or minus 8000? If your theory is correct, then majority of applicants from AS region should have case numbers above 30K. So they will not get their interviews until April or so? Same thing with Africa.

Yes, I think above assumption is not correct. Numbers are intermixed between regions. When they say 8000 for AS, that means there are roughly 2K AS candidates in there up to AS8000.

 

[radovic]

There is winner from EU (Serbia) 33xxx

 

[Arekee]

Yeah

 

[raevsky]

OK.
Then probably what is happening upon removing duplicates from each region they select new and new winners to substitute the ones removed as duplicate entries. There are a lot of them in Africa(about 10 countries), less in Asia (maninly because of Bangladesh), and even less in Europe. That is why numbers for Europe are close to the number of winners, and numbers for Asia and especially Africa are not close.

 

[dis66]

or if you are a winner and have a family they "reserve" some numbers for you and your family. So if your case number is 456 and you have a wife and 3 children they reserve 457,458,459, and 460 for you.

 

[raevsky]

That is the thing I said already

 

[Concerned_]

OK.
Then probably what is happening upon removing duplicates from each region they select new and new winners to substitute the ones removed as duplicate entries. There are a lot of them in Africa(about 10 countries), less in Asia (maninly because of Bangladesh), and even less in Europe. That is why numbers for Europe are close to the number of winners, and numbers for Asia and especially Africa are not close.

I concur. This explains it. Question, though. Have you looked into previous years DV data and did the numbers match with winners by region? Is this the first year when numbers are off?

 

[raevsky]

I have all possible statistics for the number of winners, starting 2007, but not for their rank numbers.

 

[tmfl]

How do you know it was changed in 2009?

Visa Bulletin and this forum. It looked like cut off numbers just before being current for all applicants were less than winners for DV2008 and before, and since DV2009, AS cut off consistently exceeds winners (30000 or higher before being current). However I've noticed AF case numbers slightly exceeds the total number of winners in DV2008 (visa bulletin Sep 2008, cut off of AF was 62300 > AF winners published). So, in terms of case number allocation, some additional procedure might have begun around DV2008.

 

[raevsky]

Here are my calculations for fraud levels and probabilities to win in DV-2007 through DV-2012

You might need to extract the .xls file from the archive and save it onto your hard drive first before you open it

 

[Tarukh]

It doesn't work that way. Those allocation are mixed randomly. For example first 10000 CN includes 5000 from AF, 1500 AS, 3100 EU and 400 from OC, NA and SA (randomly mixed).

I tend to agree with this, since my colleague has number 43xxxx, and I believe what has been shown on the visa bulletin are the combines cut off numbers, where all the numbers are dispersed between region, so basically the DoS will process the first 8000 in the month of October for people that would come from all over the part of this world, whether from AS, EU, AF and so on, because each number is unique and won't be duplicated to another region.

 

[Mele]

Here are my calculations for fraud levels and probabilities to win in DV-2007 through DV-2012

Thanks for that piece. But ur attachment (publish.zip) is not a supported file type or it has been damaged while u attach it. Its a .file extension & Adobe Reader, MS Word, WordPad, Internet Explorer all says it's damaged. Pls be kind enough to re-attached.

 

[raevsky]

You might need to extract the .xls file from the .zip archive (open by winzip) and save it onto your hard drive first before you open it with Microsoft Excel
It looks like I am not able to open it in Excel directly from winzip without saving first

 

[Samusoni]

Here are my calculations for fraud levels and probabilities to win in DV-2007 through DV-2012

You might need to extract the .xls file from the archive and save it onto your hard drive first before you open it

Does DOS define "Fraud Countries"? How do they define them? What entry is fraudulent...using same photo over and over, using multiple names, ....?

 

[raevsky]

DOS does not define anything. Government documents from GAO and congressional testimonies list major types of fraud.

What entry is fraudulent...using same photo over and over, using multiple names, ....?

According to the description of procedure, probabilities to win are the same within region. However, frequencies to win for different countries are usually not the same. I assume that is because different percentages of entries from different countries are filtered out. They could be filtered out automatically because they contain invalid photos (like a blank square), or photos are exact duplicates (as a file) or on some other reasons.
My definition of fraudulent entry (might not be a fraudulent one, just a dummy one or whatever) is the entry that is filtered out before it reaches the consulate (and that is why no appointment is scheduled) or USCIS.
Countries with high percentages of fraudulent entries are fraudulent countries (definition is mine).


http://www.gao.gov/new.items/d071174.pdf and http://judiciary.house.gov/hearings/pdf/Edson04052011.pdf are examples of government documents describing fraud that is being filtered out before the entry reaches consulate or after that.

What I do in my calculations is I figure out the maximum frequency of wins for some quite enough amount of entries in the region and assume that is about the real probability to win. If a country has less (and significant amount of wins still), I assume there is some fraud involve and calculate it.

Imagine a middlemen company bought 100,000 entries by bribing local authorities (from passport bureau) and submitted all of them without entrant's knowledge. Next year it split into two companies each of them submitted exactly the same entries with old photos. Those 100,000 will be filtered out as duplicates then. Those are fraudulent entries.

 

[pru]

...yeah but if the numbers just increase regardless of region - why are there no Oceania CN
s higher than 2000? surely thar should mean it is possible for anyone in any region to have a CN 88***

 

[raevsky]

A concrete example of a fraud country would be Bangladesh (97% of entries are filtered out before consulate in DV-2012) or Nigeria (83% of entries are filtered out before consulate in DV-2012)