my take

O

orca

Registered Users (C)
even at the risk of sounding ludicrously naive and shamelessly optimist (especially considering we've got stalwarts like paulclarke and funkyjunky to educate us on the raw data side), a rather inane and simplistic inference has dawned on me.

it seems if we trust rupnet (which is according to most statistics, about 0.5-0.9% of total 485 filed with the CIS), then NSC is beating VSC almost 2 to 1 as far as speed (of approval) goes. i haven't tried to get into the RFE vs. no-RFE scenarios, but if we just look at approvals from these two centers over the last two months (including an incomplete December), here are the snapshots (as of December 30th):

Service Center
Average of Jan-Apr 2002 cases on Rupnet
Approvals in November 2003 on Rupnet (% Approvals)
Approvals in December 2003 on Rupnet (% Approvals)
---------------------------------------------------------------------
VSC
291
70 (24)
74 (25)
---------------------------------------------------------------------
NSC
138
47 (34)
73 (53)
----------------------------------------------------------------------

So percentagewise (which is all that really matters), NSC is over twice VSC's success rates this month. Could it be that we're chasing down and stigmatizing the wrong guys and NSC is "statistically" the best Service Center of all? or are we skewed by some data that's so little (even when compared to VSC) that it doesn't make sense to talk about this?
 
Last edited by a moderator:
Orca,

I've had the same thoughts for the past 6 months or so! A simple calucation of %s on Rupnet always indicted that NSC was better than VCS. However, common notion is NSC is the slowest. I even heard it indirectly from someone working at NSC acknowledge that they are the slowest in the country at present - this was sometime in Aug - Sep, though.
 
ORca
the no of applicants in Nebraska is the highest.
10,000 per month
for vermont it is 7000 per month
 
frodo: somehow i don't find it logically comprehensible that the movement at NSC is even slower than the infinitesimal movement sometimes seen at TSC.

dadagiri: would that explain what we see at rupnet, though? to me this would mean the opposite. clearly, if what you say is right, then by normalizing the rupnet data of NSC to that of VSC, we have an average of 97 cases (138 X 0.70) posted from NSC as opposed to 291 from VSC, or, exactly, three times less from NSC users. this would then simply mean that NSC has a rupnet user database which is 33% that of VSC. it still doesn't solve the problem why out of these smaller posted cases from NSC, the "percentage" of approvals is over twice that from VSC.

obviously, the only other way to counter this fact (and i think one which is a particularly non-trivial one) is to say that due to the low posts from NSC at rupnet, we are merely seeing turbulence ("activity") below noise level. in other words, we're dealing with a very insignificant proportion of users to confidently say anything useful. but, on the other hand, as frodo says, if this has been the consistent pattern over the last several months, then maybe there's more to this pattern than meets the eye.
 
Orca

maybe more people from NSC list their case on RUPNET compared to those from VSC. I don't think you can use RUPNET data to see the full big picture.
 
The problem with NSC is that
the out of order execution randomness is highest in NSC.
they have cases approved to april 2002 it has diluted the no of approved cases in cases in 2001 . this is misleading as no of cases are being approved the backlog is increasing.
 
Dadagiri

I think you are confusing the meaning of the word "backlog." To quote a dictionary backlog is "An accumulation, especially of unfinished work or unfilled orders." Hence, even if NSC processed only 2002 cases they would still be reducing the backlog as long as they processed more cases than they received.


minn_labor

On the contrary, there are much more VSC cases than NSC cases on RUPNET. As a frequent RUPNET visitor, I also got a feeling the NSC has sped up while VSC has slowed down since November.
 
Pysar
I think so backlog is if inflow of applications is larger than outflow.
I Nebraska the applications are highest. I dont see nebraska being any better than than VSC. .
everybody is waiting on "luck based picking"for 2002 cases.
 
Says he, hopefully!

Originally posted by orca
.... obviously, the only other way to counter this fact (and i think one which is a particularly non-trivial one) is to say that due to the low posts from NSC at rupnet, we are merely seeing turbulence ("activity") below noise level. in other words, we're dealing with a very insignificant proportion of users to confidently say anything useful.
Click to expand...

Based on anecdotal evidence, I think VSC is considerably faster than NSC. There is no hard evidence that I can cite in support of this though. There are a couple of circumstantial observations that point in this direction. The fact that VSC has approved some June cases (much the same way as NSC approving some random April cases) should be one. The published processing dates of VSC are well ahead of those of NSC. Also, for what it is worth, VSC does not routinely RFE cases before approvals thus resulting in increased throughput, atleast in theory. In recent memory, I do not think there is even a single approval from NSC without an RFE. The last one that comes to mind is of a bloke named Await_GC a few months ago.

On whether one can draw statistical inferences from rupnet data of the sort that you have attempted, my feeling is that the sample size in itself is more than adequate, but it is hard for the data to be any less random. So, at best rupnet can guide us to what RD range is in play currently, within the limited set of countries represented therein, and at worst, the data is worthless.
 
Last edited by a moderator:
i argue

sankrit: i have friends and colleagues at VSC who i know have gotten 485 approvals lightening fast (12 months or less) and others who are so-so (centred around 18 months); however, none i know have taken as long as it's currently taking at NSC. this doesn't mean that VSC is faster per se. it could just mean that:

a) VSC has a wider spread (maybe more random?) than NSC.

b) also, over 33% of rupnet Dec 2001 cases from VSC, according to rupnet, have yet to be even initally adjudicated. even buffering for the fact that not all approvals are posted, this is a large enough number in my opinion to counter the official processing times of VSC. hence, the Aug 2001 official crap from NSC could be (and by all probability, IS) to deter (or defer) the floodgates of harassment from opening wide. then if the official times don't hold any meaning, is there a real difference between Aug 2001 (a time-dilated NSC) and Jan/Feb 2002 (a riding-the-photon VSC) that these guys advertise?

c) if, like you say, the rupnet data "might" just point towards a more accurate RD spread even within the limited set of countries, then, by corollary, people posting in this forum (i take it there's a close to 100% match in the countries represented, or that folks contributing here are a subset of those who've posted on rupnet) should still be elated to see a good representation of adjudication trends by closely watching rupnet statistics. in other words, if the rupnet data is a good match for the processing rates of VSC, NSC etc. from say, the subcontinent, then AT THE VERY LEAST, these should give us an indication of the vicissitudes of the denizens of the subcontinent (OURS?) by the verisimitudes of the cases represented.

i apologise in advance: it's not my intention to praise the beast that's been haunting us for almost two years. i'm just trying to play the devil's advocate on the one hand, and trying to fight the remotely possible and probably erroneous "grass-is-greener-on-the other-side" concept we might be nurturing on the other. albeit, i sincerely hope i'm right and you guys are all wrong and that NSC is truely faster than the rest of the SC's, even if there's a slow, perhaps irreversible, trickle from the clogged drains where the backlogs have been piling up.
 
Last edited by a moderator:
Re: i argue

Originally posted by orca
sankrit: i have friends and colleagues at VSC who i know have gotten 485 approvals lightening fast (12 months or less) and others who are so-so (centred around 18 months); however, none i know have taken as long as it's currently taking at NSC. this doesn't mean that VSC is faster per se. it could just mean that:

a) VSC has a wider spread (maybe more random?) than NSC.

b) also, over 33% of rupnet Dec 2001 cases from VSC, according to rupnet, have yet to be even initally adjudicated. even buffering for the fact that not all approvals are posted, this is a large enough number in my opinion to counter the official processing times of VSC. hence, the Aug 2001 official crap from NSC could be (and by all probability, IS) to deter (or defer) the floodgates of harassment from opening wide. then if the official times don't hold any meaning, is there a real difference between Aug 2001 (a time-dilated NSC) and Jan/Feb 2002 (a riding-the-photon VSC) that these guys advertise?

c) if, like you say, the rupnet data "might" just point towards a more accurate RD spread even within the limited set of countries, then, by corollary, people posting in this forum (i take it there's a close to 100% match in the countries represented, or that folks contributing here are a subset of those who've posted on rupnet) should still be elated to see a good representation of adjudication trends by closely watching rupnet statistics. in other words, if the rupnet data is a good match for the processing rates of VSC, NSC etc. from say, the subcontinent, then AT THE VERY LEAST, these should give us an indication of the vicissitudes of the denizens of the subcontinent (OURS?) by the verisimitudes of the cases represented.

i apologise in advance: it's not my intention to praise the beast that's been haunting us for almost two years. i'm just trying to play the devil's advocate on the one hand, and trying to fight the remotely possible and probably erroneous "grass-is-greener-on-the other-side" concept we might be nurturing on the other. albeit, i sincerely hope i'm right and you guys are all wrong and that NSC is truely faster than the rest of the SC's, even if there's a slow, perhaps irreversible, trickle from the clogged drains where the backlogs have been piling up.
Click to expand...

It neither helps nor hurts to be optimistic at this juncture and I am happy to jump on to your boat. However, I do not think we suffer from the grass-is-greener-on-the-other-side syndrome as is apparent from the fact that all of us appear to be of the same opinion with respect to TSC!
 
You must log in or register to reply here.
Top