A STAR question that I have not been able to get an answer to

classic Classic list List threaded Threaded
19 messages Options
Reply | Threaded
Open this post in threaded view
|

A STAR question that I have not been able to get an answer to

Joeinpb
CONTENTS DELETED
The author has deleted this message.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
Since the FPR is supposed to based on vehicles from the last cycle, I would guess it could take 2 years or longer.  BAR did address this situation in the public meetings in saying that a technician with volume below a certain threshold would receive a NO SCORE.  
EE
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

EE
In reply to this post by Joeinpb
I have a tech   His name is still on the machine and hasn't worked since dec 2010 not here not anywhere else and still has a score of .23. So it's been 2 years now   What a program   Sry joe but I think accused is right   Up to 2 years for no score to show up. Maybe longer.   It's in God's hands
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Karupt
In reply to this post by Joeinpb
Joe in Pb , i hate gray areas and questions that have no answeress, therefore i have not read your post ,just the headline and the feild i work in has lots of nonanswered questions , so since these are supposed to be PROFFESIONNALS THAT WE ASK THESE QUESTIONS TO  ,my only answer to someone who HAS NO ANSWER IS TO ASK THEM TO PLAY BLACK OPS TWO ON THE ZOMBIE MODE WHERE I COULD BLOW THIER BRAINS OUT . just my opinion and we are talking about Black Ops two !
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
In reply to this post by EE
You know that BAR will take him out of your machine if you call and tell them right??  They are required to when an employer calls.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Joeinpb
In reply to this post by Karupt
CONTENTS DELETED
The author has deleted this message.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Banned
In reply to this post by Joeinpb
One thing we constantly, and sometimes conveniently forget, is that we are taking the BAR's word for
it, when they say that the FPR is a calculation. They have provided NO data, and NO explanation of
how the data is used. When any government agency gives you a "reason why," it's an automatic red
flag if all we have is their word, and nothing else. I refer to it as the "number," not the score.

We went to war with North Vietnam, because of the FABRICATED Gulf of Tonkin incident. The official
story of 9/11 is so full of problems, that there is NO WAY it happened as concluded. The Warren report
on the JFK assassination has serious problems.

It's become a "rule of thumb" that if an agency tries to feed you an explanation for something, you'll
need some Tums, Alka Seltzer, Beano and Nexium, just to keep it down.

++++++++++++++++++++++++++++++++
What came first - the chicken, or the egg?
++++++++++++++++++++++++++++++++

Just when you thought that things were hopelessly lost, I want you to know that a good friend of mine
did come up with the answer to this age old mystery. You have to word this correctly, if you want to
dazzle your friends. You say, "I have figured out the age old question, was it the chicken that came
first, or was the egg laid first. It was neither, because it was the rooster who came first, while he was laying the chicken."
 

Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
What??? No Bilderburgers, moon landings or Free Masons??

The question isn't if the FPR is a calculation but how the calculation is made.  I have a theory on the basic process but the variables involved can produce about any number they want when applied to the techs statewide so the question of how the calculation is made is still really open.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Banned
In reply to this post by Joeinpb
And on what verifiable evidence to you base the assumption that it is a calculation?
Or are you taking their word for it?
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
You're right, they won't verify it.  I look at what is most likely though.  It is far more likely that they created an algorithm with variables that they can manipulate to make everyone's life hell than they have a few people behind the scenes making up scores for the entire state.  It just makes more sense.  It doesn't make it fair for most involved - for sure.   But consider this...the idea of basing a scoring metric on factors completely out of the control of anyone connected to the inspection process is so fucking stupid that it almost has to be true.  It is exactly the kind of thing some cubicle monkey with no connection to the real world would dream up.  And...in typical government fashion, they get to blame everything on supposedly indisputable statistical data so they can step back and say "figures don't lie".  As I see it, it's just the most likely scenario.  

We see the same approach all around us every day.  Statistical data is used to prove and disprove any and every theory desired.  The "science" of statistics is extracting a desired proof or result from seemingly insignificant data.  From global climate change ( you notice I didn't say "warming") to second hand smoke to the health benefits of shark cartilage it's all been "proved" through statistical analysis.  All of which is easily accomplished by adding or removing data points until the desired result is achieved.  This is most likely the same thing contained in the sierra report and the reason that the "final report" has not surfaced.  

It is my belief that BAR started with a desired outcome and reverse engineered an algorithm using failure rates on the follow up tests that would create a positive and negative outcome for a few stations that they thought they knew what the outcome should be and then applied that same template to the rest of the state.  But more than that, I think this all grew out of existing data mining.

It starts something like this...the year is 2002 and some guy in a back room deep within the bowels of BAR engineering is having a conversation with a PR2.  He says, "I could tell you who is cheating out there just by looking at their VID data".  A PR1 listening in says "you should put together a demonstration of that so we don't have to go out and actually work anymore".  So this self assured smart guy that rarely interacts with the real world seeks out the help of some other guy that is equally self assured of his own brilliance and sets about the task of "proving guilt without physical evidence".  They make some general assumptions about the smog check industry and start looking at failure rates, failure types and inspection times. Having access to data from the entire state, they create a set of data points that the PR1's can access prior to making their quarterly QA inspections. The PR3's are so impressed with their results that they require all PR1's to use this data every day.  
Over time, new data points are added to pick out vehicles with RPM readings that don't fit a statistical profile, rates of visual failures with no emissions failures, vehicles being tested with no OBDII data, etc.
All of us got to see this first hand when every 90 days a PR1 would show up with a list of VIR's and invoices to be pulled.  Or even worse, 2 BAR reps show up and accuse you of clean piping because they used VID data to "prove" it.  They tell your tech he might as well admit it now or they would show up with the police and confiscate all of the equipment and close down the shop.  The tech would also be arrested for a felony and go to prison (no lie - this actually happened to one of my employees).   All based on VID data.

Now bring in someone as fucking stupid as Sherry Mehl.  She says, the smart guys in engineering can tell who is cheating just by looking at the VID data. Why are we wasting time having PR1's driving around all day to ask why they are cheating when we can just start sending them UC cars and camera crews so we can fine them all out of business. As a matter of fact, why are we even testing the emissions on most of these cars.  The REALLY smart guys over at ARB tell me every car has a computer in it now that can tell us if a car is polluting and states all over the country are just using that to test cars.  Why do we need all these cheating "mechanics" to tell us what the computer already knows.  Now keep in mind that at one time, there were people that worked for BAR that actually knew something about cars.  When these people tried to tell Sherry why her opinions might be wrong, she forced them all out so she would never hear any disagreement. Around the same time, one of those stupid smart guys in the basement of Sherry's bowels whispered that he had finally come up with the complete solution to "proving guilt with no evidence".  He calls this new data set a Follow-up Pass Rate.  It's really quite simple he explains..."We already know that at least half of the smogs performed are cheated through. So take 100 87 Ford Mustangs that should be failing a smog check.  50 are correctly failed, repaired and certified.  The other 50 are incorrectly tested, improperly preconditioned or clean piped.  Which group do you think is more likely to pass a smog test 2 years later?"  "This is very smart" Sherry says.  "I am really glad you hung around and kissed my ass all this time".

And STAR is born.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Banned
In reply to this post by Joeinpb
SHIT, this is elegant!! And here I thought that I had a "heads I win, tails
you lose" type of question. I guess I'm wearing the egg on the face, but it
doesn't turn out that way too often. So I'll survive this.

Yes, I agree, when it comes to probabilities, this is a highly likely
scenario. It wouldn't surprise me if you've hit the nail, nearly square on
the head. Since we are in the realm of theories here, I could add a couple
of other items to your very well put together thesis.

Somewhere in the story, someone also decided, "Hey, we can look like heroes,
if we can incorporate this whole scheme into LOWER PRICES." Why not create a
new class of techs, (SFB). They will work for next to nothing, and the tests
they have to pass will be so damn easy, it will be impossible for them to
fail. Who gives a flying fart if they don't notice things like a bald tire about to
explode, or a serp belt on the verge of breaking, or dangerously low oil pressure,
during the test. The PRICE is what the motorist will remember. And in addition,
we can rake in some serious "chump change," by making it look like we crafted
the whole joke of licensing process, just so the greedy Jiffy Lube bastards can
replace the legitimate test sites. Let's see, free reign to fuck with any tech or
shop, elimination of 50% of legitimate shops, new markets for Jiffy Lubes and the
SFB techs, major kickbacks in my pockets, look like heroes to the motorists,
I think we have a winner."

So how's that as an addendum? There is just one thing they seem to have
forgotten in this whole story. This country is founded on the presumption of
innocence, which is totally missing from the debacle. They have no problem
using the FPR on us. But as I mentioned in another thread, what if the exact
same logic was used on PR1s.

Every time a PR1 certifies a shop to perform smogs, in 2 years if another
PR1 finds anything missing, or out of place, it goes back to the original
PR1. The original PR1 can lose his job, just because in 2 years the shop was
not in the state it was 2 years before. If they really believed in their BS
logic, they should have no problem using it here as well. If you believe
they will ever admit this, or use it, then you probably also wait for Santa
to come down the chimney.

By the way, for those who haven't figured out what SFB stands for, it's
"shit for brains."
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

barrym95838
Administrator
In reply to this post by Accused
Accused,

Very well thought out.  When you put it that way, it seems to make sense, until ...

I look at my three 0.00s and wonder out loud how someone who does "proper" inspections over 99% of the time can find himself sucking the snail slime at the bottom of the FuPR barrel.  As I stated before, poor maintenance, improper diagnoses, and improper/incomplete repairs could go a long way toward explaining why my FuPR is far from good, BUT A PERFECT 0.00?  It defies logic.

I have not completed my compilation of all of my inspections from two years ago (another big item on my gitterdun list), but I can see how my "raw" FuPR could be significantly higher than my "calculated" FuPR due to the collection and averaging of the all the zeros and ones from each vehicle profile category, as you described in a previous thread.  But ABSOLUTE ZERO?!!?  It seems like a one-in-a-million situation, unless there is an undisclosed variable, like a "Zero this Scumbag out for Good" button that someone clicked on over there.

Just my take on the situation,

barrym95838
EA144107

Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

FRUSTRATED
In reply to this post by Accused
Accused, this is freaking awesome, well put!
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Banned
In reply to this post by Joeinpb
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
But ABSOLUTE ZERO?!!?  It seems like a one-in-a-million situation, unless there is an undisclosed
variable, like a "Zero this Scumbag out for Good" button that someone clicked on over there.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Barry, I used to work as a programmer using Quickbasic 4, and I can tell you that it's not hard to
generate a subroutine that could end up in this Barry-zero state. It would just take into account
various variables, and in the if/then part, it would automatically give a zero under certain conditions. Sometimes it's necessary to do this, when division by zero is possible. This is the monkey wrench
that all programmers try to avoid, because the whole process explodes at that point.

Another possibility is that truncation is employed, rather than rounding up. Banks have used this for
years, because it saves them tiny fractions of a cent. If you saw the movie "Office Space," the
employees reprogrammed some bank(s) software to deposit this tiny amount into an account, instead
of truncating it. With the process happening millions of times a day, they had a BAR like kickback in
this hidden account, after a few days.

So, in my estimation, if it is a true calculation, it may only calculate to a point. After that, it's the
goose egg for the poor bastard.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
Banned,

You nailed it on the zero calculation.  I have spent a lot of time doing large spread sheets that involve similar built in calculations.  I can see easily that while a ZERO score is unlikely if calculated correctly it is still possible to see it occur hundreds of time across the state.  Again the variables are the question.  I suspect it also has to do with the number and size of each vehicle group and the volume threshold for considering a group scorable.

As I stated or at least eluded to, I suspect that they beta tested this using a small number of "known quantity" shops and techs to produce the scores they wanted to see - and this probably included a zero score for some in the beta group.  Being very satisfied with themselves over these desired results, they then just applied it to the rest of the state.  Why would they care if someone received a zero score...they must have earned it because the beta testing worked as planned.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
In reply to this post by barrym95838
MBarry,

If you start with the supposition that BAR intends this 0.00 "score" to indicate an OPINION of your performance rather than your actual performance it starts to make more sense.  And I don't mean that someone in BAR has purposely singled you out but rather they have chosen a given statistical outcome to produce a 0.00 "score" that is supposed to mean "we are sure that inspections are being done incorrectly".  Somewhere in the materials explaining this program, they used language very similar to this.

What I mean by this is that the FPR does not in any way correlate directly to a pass/fail rate.  

My theory is this:

For purposes of explanation lets assume that there are 1000 vehicle groups.
Your raw pass/fail data for each group is compared to the state average for that group. If your pass rate is within a given percentage of the state average you receive a 1.  If you fall outside of that average, you receive a zero.
Groups that you have not performed any test receive a "no score".
Simply averaging the sum of these zeros and ones will create a raw "score between 0.000 and 1.000.
In a simple average showing only 2 decimal places, a score of 0.004 or less would show up as 0.00. But as Banned stated, they could easily have decided that any score below some other arbitrary number is considered a "zero" as well.
Introducing other variables further complicate this metric and they could easily introduce some secondary algorithm in where they identify some other suspected behavior through VID data mining and have it deduct from your raw scores.

It is these automatic "opinion" based corrections to the raw score that I suspect creates the published FPR.

So unfortunately for you and many others, zero scores are a reality of this scoring metric.  

I have spent way too much time thinking about this.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Joeinpb
CONTENTS DELETED
The author has deleted this message.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Accused
I think this validates a lot of what I said in earlier posts on this thread.

They don't care if it actually works

They don't care if people get hurt

They don't care if it's unfair

They don't care if the air gets any cleaner

Basically, they just don't care.
Reply | Threaded
Open this post in threaded view
|

Re: A STAR question that I have not been able to get an answer to

Banned
In reply to this post by Joeinpb
There is ONE thing they do care about. They may even care about THIS, more than doughnuts. It's
the illusion that they are doing something constructive, and responding. That's why they won't
release the Sierra report. As long as it's hidden, they can just say they did it because of the Sierra.
A huge illusion, and a BS explanation. But this, they actually do care about. It's keeping politicians
and consumers THINKING that they are operating in a "constructive" manner.

As we all know, if they had nothing to hide, the report would be freely available. A judge will want
to see the ORIGINAL report, when the "why did you create this disaster" question comes up. When
he sees that the debacle has nothing to do with the report, then he will have to look at the hidden
agendas we've been talking about for quite a while.

In my opinion, that's why the Wallauch's March 9th total BS post disappeared. Because it can be
proven to be a fabrication, it had to go. Once there is one obvious lie, the old cliche of "one seldom
finds a lonely lie," comes in to play. In other words, one lie leads to another.