It sounds like a conspiracy theory because it is not based on any facts and despite that, you keep believing in it. Sure, it is possible that everything is a lie and that you cannot trust anything the government tells you, but that is not a theory that I embrace.
What you say is that a biased drawing will bankrupt these "pay after win" companies.
It looks like you do not understand what I was saying. It is not biased drawing. It is just another way to do an unbiased one. However, this way gives a much higher variance than the previous one.
We already established that their costs per entry are very small,
No, we have not. Even though Fox25 said that, he might not have known all detais. They hire people on the streets that try to get personal data for new people. Fox25 is just paid for writing scripts (and running them) for entering all entries into DOS system and then checking the results. He does not know all additional costs. I am positive, they are not small.
fraction of a cent for electricity and that a uniform distribution of their entries over the entire period will hedge against any big loses. Moreover, sometime they can also win more. They could even charge more for a win if they found to be at a loss, but that would not be necessary.
After they are at a loss one particular year, they would think twice before investing more into it. Also, if all entries which use the same photos as in DV-2011 are disqualified, they will just through all those away.
Despite all this, you still believe that this is such a smart tactic that some genius at DOS has created.
I have some doubts, because of new facts - other regions. That is what I did not know before.
Second, there are better, more direct ways for KCC to invalidate entries if they are fraudulent and if they really wanted to. You will argue that the law requires all entries to have an equal chance,
correct
but the actual wording is "A qualified electronic entry submitted directly by an applicant has an equal chance of being randomly selected by the computer at the KCC, as does a qualified electronic entry received from an outside intermediary on behalf of the applicant."
An entry submitted by a company without the consent of the applicant cannot be said to be on behalf of the applicant.
KCC is not able to distinguish that one on the stage of drawing. That could only be done at the consulate. But the drawing is done before, and the equal chances have to happen on the stage of drawing.
Thus KCC could choose to invalidate them without breaking any rules.
No, they cannot.
Making a concerted decision to select both partners of a married couple would indeed affect this type of fraud. However, this has the effect that for the same number of selectees (let's say 100K), there is a smaller pool of possible applicants for visas. You know that the 50K visas include dependents as well. If you selected 100K principal applicants and each applicant has on average 0.5 dependents, there will be 150K total applicants for those 50K visas. By selecting both members of a couple, they reduce the average number of dependents and they would have to select more than 100K people.
Correct. And the official instructions do not mention 100,000. They just say "more than 50,000". Absolutely true, they would select about 150,000 this year.
Anyway, I agree that this could be a valid tactic to reduce one type of fraud.
Good.
But a better explanation for why both husband and wife were selected this year is because they entered their applications around the same time and the selection was not random.
No. This is not a better explanation. This is the new way of drawing - based on time of submission. Truly random, however. In order to combat this type of fraud they changed the drawing. That is it.
So all applications entered during a specific time frame were identified as selected by the online status check. Show me an example where the husband and wife submitted entries on different dates and they both got selected, and then you might have a leg to stand on.
It looks like there is some kind of miscommunication. What you are proposing is illegal. What they did - based on time, is not illegal. However, being based on time, it would automatically do with family applications what actually happened with them. They did a truly random drawing, just based on time of submittal. And that would automatically lead to combatting this type of fraud.
The selection algorithm does not check if two of the selected applications correspond to married couples or not.
Of course, because it is illegal. That would contradict equal chance requirement.
With a random drawing, the odds of that happening are small. With the bogus algorithm there were multiple instances reported in this forum alone.
No, it is not bogus. It is correct.
Or it may have been a bug, as the online status check has been bugged since last year. It does not surprises me at all that they did not catch the bug before because I have some insight into how government agencies operate and the lack of communication between the entry level people answering the phones and those making the decisions. You only need to look at how the phone operators give wrong answers about what documents people doing AOS need to submit to KCC.
Come on, I have never led a big software project, but I led a few small ones, working for private companies. To me it is clear that statistical tests should have been satisfied, and I know perfectly well which tests to satisfy. I am sure they would hire a person not worse as myself to do the job.
Now, if we consider your theory, is not DOS afraid that the cancellation of results also results in some lawsuits and they would not be able to prove the existence of any bug?
If I were in charge of the project, I would create proof of running those statistical tests beforehands, and logs for each random number selected. That would be a proof for the court.
Then they would be in even more trouble. I do not want to go deeper into this topic. When you do not believe any facts or any evidence that we have, even if it is mostly circumstantial, you can make up any theory that you want. But it becomes a conspiracy theory in that case.
The thing is that the word random has a number of meanings. The one defined in 22 cfr 42.33 is more or less useless.
(c) Processing of petitions. Entries received during the petition submission period established for the fiscal year in question and meeting all of the requirements of paragraph (b) of this section will be assigned a number in a separate numerical sequence established for each regional area specified in INA 203(c)(1)(F). Upon completion of the numbering of all petitions, all numbers assigned for each region will be separately rank-ordered at random by a computer using standard computer software for that purpose. The Department will then select in the rank orders determined by the computer program a quantity of petitions for each region estimated to be sufficient to ensure, to the extent possible, usage of all immigrant visas authorized under INA 203(c) for the fiscal year in question. The Department will consider petitions selected in this manner to have been approved for the purposes of this section.
Because it does not allow to figure out whether it is in fact random or not. So, the law does not actually define it well. What is random from the point of view of math (and would work in casino) might not be understood by general public.
Wait, are you saying that you have some doubts about your theory?
What I am saying is that logically the drawings for different regions have to be independent. That is why truly random drawing worldwide based on the submittal date (what is truly random) would most likely select different winning dates for different regions. However, 90% winners worldwide came on two days. That is really suspicious. Three largest regions (Africa, Europe, Asia) - all on the same date. The probability is 1/36. Not really very small, but not very large as well. Enough to suspect there could be a bug with selecting first interval instead of a random one. Random interval would not contradict the law. But it is unlikely all 3 regions would have the same random interval.
Come on, do not give up so easily. Think about it, DOS foresaw that if they chose different days for different regions, then the biased results could not be blamed on a bug in case of an uproar from the public (which they have foreseen as well).
Right. here you are correct. If those dates would be different for different regions, I would not have doubts. And they would not have a valid argument to cancel this selection.