October 2017 I-129 K-1 filers

[Chris J 111]

3 minutes ago, Tpienkowski said:

If things stayed as slow and tedious as they did for May & June, I would agree. However July & September have fewer cases & USCIS seems to be picking up the pace as well as the NVC waiting time decreasing. If things stay like this I would like to think that POE for October’s by the end of June/ first week of July is feasible 

So even factoring in the NVC wait time, as well as the Embassy interview wait time? 

[DandY 134]

On 1/21/2018 at 12:53 PM, Daniel P said:

In that time, with a total of 1877, 818 and 243 approved cases (after removing some outliers) respectively:

 

• May had a global mean of 175.25 days from Received to Approval date, with a standard deviation of 14.06 days.
• June had a global mean of 186.05 days from Received to Approval date, with a standard deviation of 9.16 days.
• July had a global mean of 193.91 days from Received to Approval date, with a standard deviation of 3.28 days.

Good stuff.

 

I have not pulled the numbers since early December so I'm not sure how it is going. But a couple of points:

- To get mean numbers like this, you have to remove a large quantity of "outliers"--about 15-20% of cases back when I was calculating. What quantity did you use as "outliers" to reach these numbers?

 

- Back in October when I was using a predictive algorithm to determine June's average wait time (which I calculated at 184-186 days), I was only calculating for 80% of filers. For the last half a year 15-20% have been waiting in excess of 230 days (though I have not pulled numbers in over a month, so this may be decreasing but anecdotally it does not seem to be).

 

- If the outliers are still 15-20% of the total, it means that 1 in 5 to 1 in 6 people will be waiting much longer than the above mean + standard deviation and that is a high enough percentage (and painful enough) that it should be considered. As of now, there are still May petitions not approved.

 

Good luck to all!

[Daniel P 2,022]

25 minutes ago, DandY said:

Good stuff.

 

I have not pulled the numbers since early December so I'm not sure how it is going. But a couple of points:

- To get mean numbers like this, you have to remove a large quantity of "outliers"--about 15-20% of cases back when I was calculating. What quantity did you use as "outliers" to reach these numbers?

 

- Back in October when I was using a predictive algorithm to determine June's average wait time (which I calculated at 184-186 days), I was only calculating for 80% of filers. For the last half a year 15-20% have been waiting in excess of 230 days (though I have not pulled numbers in over a month, so this may be decreasing but anecdotally it does not seem to be).

 

- If the outliers are still 15-20% of the total, it means that 1 in 5 to 1 in 6 people will be waiting much longer than the above mean + standard deviation and that is a high enough percentage (and painful enough) that it should be considered. As of now, there are still May petitions not approved.

 

Good luck to all!

Only already approved and updated cases were considered. From that group there were some outliers, but not too many, definitely less than 15-20%. The criteria I used was ± 2*sd. Please, note that that chart is not intended to predict anything, just to compare under the same conditions every day that received cases in May and June as a way to see what the trend line is.

Edited January 23, 2018 by Daniel P

[DandY 134]

38 minutes ago, Daniel P said:

Please, note that that chart is not intended to predict anything, just to compare under the same conditions every day that received cases in May and June as a way to see what the trend line is.

I realize that it was not predictive though some might take it that way. You have all the data you need to create a predictive algorithm if you want since it seems things are back to a "predictable" environment. You can calculate the "velocity" of approvals (how many per week) and the backlog to determine the probable timeline. That was essentially the third version of my algorythm that I used when things "stabilized" (earlier versions did not have enough data so I had to make assumptions and use historical trends). 

 

40 minutes ago, Daniel P said:

Only already approved and updated cases were considered. 

Ah, your dataset was not all petitions but only "touched" petitions (i.e. those in "recieved" status were not considered), thus your "outliers" only includes touched ones. I was saying that 15-20% of May and later filer petitions have been staying in a "received" state for longer than 200 days (and if you hit 200 days it seems like you have a very high probability to be over 230 days though I have not done all the math on this). You have left them out of your data, so the true mean will be worse than your numbers--you essentially have what I was doing--the mean and SD for 80% of the cases, not 100% of the cases. There are MANY "untouched" cases in those time ranges. For instance there are May cases at over 260 days. June cases at over 230 days. Cases that went over the 200 day mark were about 15-20% (I can't remember off top of my head the actual percentage and it is on my home computer) of May cases (I have not looked at June cases in a while so I don't have final figures for the percentage of June that went over 200 days). 

[L and E 305]

I'm driving myself crazy lurking on the june and july threads looking for good news of approvals... I did so well up until January with this wait... Now I feel like I'm clawing my way through each day. Can it just be April already pleeeeeeease 

[Daniel P 2,022]

4 hours ago, DandY said:

I realize that it was not predictive though some might take it that way. You have all the data you need to create a predictive algorithm if you want since it seems things are back to a "predictable" environment. You can calculate the "velocity" of approvals (how many per week) and the backlog to determine the probable timeline. That was essentially the third version of my algorythm that I used when things "stabilized" (earlier versions did not have enough data so I had to make assumptions and use historical trends). 

 

Ah, your dataset was not all petitions but only "touched" petitions (i.e. those in "recieved" status were not considered), thus your "outliers" only includes touched ones. I was saying that 15-20% of May and later filer petitions have been staying in a "received" state for longer than 200 days (and if you hit 200 days it seems like you have a very high probability to be over 230 days though I have not done all the math on this). You have left them out of your data, so the true mean will be worse than your numbers--you essentially have what I was doing--the mean and SD for 80% of the cases, not 100% of the cases. There are MANY "untouched" cases in those time ranges. For instance there are May cases at over 260 days. June cases at over 230 days. Cases that went over the 200 day mark were about 15-20% (I can't remember off top of my head the actual percentage and it is on my home computer) of May cases (I have not looked at June cases in a while so I don't have final figures for the percentage of June that went over 200 days). 

I thought about it but I don't think the data USCIS provides is ever stable or reliable enough to make good predictions. Besides, I'm afraid that any kind of prediction can only bring more stress and frustration than anything else so I'm not really up to it. That's only my opinion. If someone else (you?) is willing to do it, I'll be happy to take a look to the results, either if they end up being correct or not.

 

In the text before the charts I tried to explain in a short way what I was exactly doing with the data. I'm aware that the results provided aren't "complete" because they don't include all the cases. The problem with received cases is that some stay that way indefinitely and only one part change to approved. Since there's no way to know about what cases were approved but not updated (and WHEN that happened), and what cases are still pending, I didn't think they should be included. When and if they change to approved, then they will be included and the average increased, but until then they can be anything and so I just prefer to "temporarily" leave them out. Furthermore, and this is also my personal opinion, I don't believe USCIS don't update only "long" cases, and so, if they don't update "short", "average" and "long" cases all in a proportional way, then the average should remain very similar even if those "still" received cases could be included.

 

Anyway, my main purpose is to just have an general idea of the trend through the comparison of a set of days under the same conditions, that's the reason that I decided this approach.

[DandY 134]

2 hours ago, Daniel P said:

I thought about it but I don't think the data USCIS provides is ever stable or reliable enough to make good predictions.

I think it is close enough. More than 40 days before June started seeing approvals I was predicting that 80% of early June filers would be mid-180s and you analysis shows I was pretty much right on.

 

2 hours ago, Daniel P said:

If someone else (you?) is willing to do it, I'll be happy to take a look to the results, either if they end up being correct or not.

Honestly, I'm too lazy. We are approved and in transit to the embassy so I don't really want to. It was a lot of work and I lost my taste for this site (today was first posting in a looooong time). When I posted the results of the analysis and an overview of my methodology back in the June thread I was ruthlessly attacked (shoot the messenger). A group of people joined the June thread in October and with no data or facts thought June approvals would by 160 days--despite the fact there was absolutely no way to realistically interpret the data to back this up. When I pointed this out I was made fun of for being an idiot that didn't know what he was talking about. So, lets say that VJ isn't really the place I hang out anymore.

 

If you change your mind, send me a message (I'm not on much) and I'll explain my methodology and the code to support it if you want to get into the prediction biz.

 

 

[Naes 12,205]

Nobody attacked you, we actually asked for the methodology and you took offense on everything, and was extremely rude. In October with only April and May data I did not say 160 days on average, I said to start around 160 days, and it started on 168 days. Which was never the problem on who was right other than your rudeness.

[DandY 134]

9 minutes ago, Naes said:

Nobody attacked you, we actually asked for the methodology and you took offense on everything, and was extremely rude. In October with only April and May data I did not say 160 days on average, I said to start around 160 days, and it started on 168 days. Which was never the problem on who was right other than your rudeness.

I had no desire to bring this ####### to this thread, I'm not posting anymore so feel free to say what you want. You are welcome to search the thread and see the attacks from Squirrel and others (not you initially and I didn't imply that you did) including posted gifs to make fun of my predictions. Anyone that cares (no one) can go search the threads chronologically from August on and see my posts. But you are welcome to see it how you want and say what you want. I'm done here, I was trying to be helpful since I saw someone looking like they were starting to do some of the math for the October filers. 

 

As for the rudeness, I only got rude after the attacks and a final "I told you so" on the average wait time since my prediction was so accurate. Yes, the "I told you so" was a bit juvenile, but I thought maybe some people would learn a lesson and try to understand the message before shooting the messenger. 

 

Feel free to go back to attacking me and bring the others here to help if you wish--I will not respond. But I implore you not to sully the October thread with this nonsense.

[Tpienkowski 210]

11 hours ago, Chris J said:

So even factoring in the NVC wait time, as well as the Embassy interview wait time? 

That’s what I believe, but like I said, ONLY if the improvement we saw last week stays consistent and we don’t go back to the 100 cases processed a week. Today in the July mid day scan, we saw 60 cases from today approved. So I’m starting to think a bit more optimistically 

[Chris J 111]

32 minutes ago, DandY said:

I had no desire to bring this ####### to this thread, I'm not posting anymore so feel free to say what you want. You are welcome to search the thread and see the attacks from Squirrel and others (not you initially and I didn't imply that you did) including posted gifs to make fun of my predictions. Anyone that cares (no one) can go search the threads chronologically from August on and see my posts. But you are welcome to see it how you want and say what you want. I'm done here, I was trying to be helpful since I saw someone looking like they were starting to do some of the math for the October filers. 

 

As for the rudeness, I only got rude after the attacks and a final "I told you so" on the average wait time since my prediction was so accurate. Yes, the "I told you so" was a bit juvenile, but I thought maybe some people would learn a lesson and try to understand the message before shooting the messenger. 

 

Feel free to go back to attacking me and bring the others here to help if you wish--I will not respond. But I implore you not to sully the October thread with this nonsense.

Good Lord, sounds like someone s*** the bed... 

[Chris J 111]

7 minutes ago, Tpienkowski said:

That’s what I believe, but like I said, ONLY if the improvement we saw last week stays consistent and we don’t go back to the 100 cases processed a week. Today in the July mid day scan, we saw 60 cases from today approved. So I’m starting to think a bit more optimistically 

Yeah, I hope you're right.. I know the Russian Embassy has a 58 day wait for an interview.. Can't wait to see some people from August start getting approved, then we know they're really rolling lol.

[Naes 12,205]

Whatever 🤷‍♀️

[Tpienkowski 210]

10 minutes ago, Naes said:

Whatever 🤷‍♀️

Happy thoughts, think happy thoughts  

[Tpienkowski 210]

15 minutes ago, Chris J said:

Yeah, I hope you're right.. I know the Russian Embassy has a 58 day wait for an interview.. Can't wait to see some people from August start getting approved, then we know they're really rolling lol.

Yeah I can’t vouch for your embassy, my fiancé is from England and I’ve heard that they’re one of the quicker ones but I hope you get a quick embassy appointment when the time comes