nsxtasy

The question: How many ITRs were sold in the U.S. market?

EDIT: The answer to this question has been copied into the end of this initial post. Look for the <FONT COLOR="red">red heading</FONT>.

The good news: It's possible to come up with a definitive answer to that question.

The bad news: It's going to take a bit of computer skills and a lot of time.

The background:

Between a previously posted letter and personal contacts, several of us have heard an answer to this question from folks at American Honda, that the numbers sold for model years '97, '98, '00, and '01 were 318, 1023, 1347, and 1157, respectively, for a total of 3,845 ITRs. But some folks here have been looking for documented proof of those numbers, beyond that already available. It is now possible to get such proof. Here's how.

Every ITR sold in the United States has a VIN (vehicle identification number) in the format JH4DC231cySnnnnnn where:

c is a check digit, calculated as shown on this web page, where it says "What is a Check Digit and how is it calculated?" (The examples and information shown are for NSXs, but the calculation itself applies to ITRs as well)

y designates the model year of the car (V=1997, W=1998, Y=2000, and 1=2001)

nnnnnn is a sequential numbering of the cars produced (that does not have ANY relationship to the ITR numbers on the armrests)

Therefore, it is possible to use a computer to calculate the check digit, and therefore the VIN, for all possible VINs for ITRs sold in the United States.

It is also possible to take each VIN and verify whether it was or was not ever a valid VIN for an ITR, by entering it into the "My Acura" section of Acura's website. The website will indicate whether or not it is a valid VIN, and if so, will also indicate the color of the car. If someone does this, he can compile a list of all valid VINs of every ITR sold in the United States, and identify the color of each one. This will give us a total number of ITRs and a breakdown by color.

The problem in doing so is that this process involves entering thousands of VINs into the website to verify which ones were actually ITRs. This will require a fair amount of work. In fact, it might be so much work that it's impractical; I don't know. More specifically, I don't know how high the VIN sequence number (that nnnnnn part of the VIN) goes each year. If there was one VIN sequence for all Integra versions, including two and four doors, GS-R and SE and LS and GS as well as ITRs, there may be so many possible VINs to check that it's impractical to check all of them. (Integras were selling at a rate of approximately 40,000 cars a year in the States when the ITR first came out in 1997, although that figure dropped to around 14,000 a year by 2001, as shown in the figures here; note that those figures show sales by calendar year rather than model year.) But if the ITRs were only produced in a small number of "runs" during the manufacturing process, it might be feasible to check only the ranges of VIN numbers in and around those runs, so that only a few thousand VINs for each year would need to be checked instead of tens of thousands. Unfortunately, I don't have any idea how to go about isolating which portions of the sequence number series would contain all the ITRs. EDIT - since originally writing this post, I found out that there are four sequences of USM Integra VIN numbers each year, and exactly how high they go for each year's sequence for manual hatches. See this later post in this topic for details.

If this sounds too complicated and impossible, it's not. Someone on NSXprime went through exactly this process to determine the VINs of NSXs sold in the U.S. in recent years, identifying each one by color, transmission, and body style. Granted, it was easier with NSXs because it was only a few hundred each year, all using one sequence of VINs with no non-NSXs mixed in, so there were only a few hundred to check without having thousands of GS-R or LS Integras mixed in (and again, I don't know whether or not that's the case with the VIN sequence numbers used for the ITR). But still, if it's possible for someone to do that with the NSXs, it's possible for someone to do that with the ITRs.

Maybe there is someone here (or several someones here) who might be interested in taking this on. Someone with the computer skills to program the algorithm to calculate the check digit for each possible VIN, and someone to spend the time entering the list of possible VINs into Acura's website to verify which ones were sold as ITRs (and find out the color of each).

EDIT - Added bolded sentences above.
EDIT - Added the answer below.

<FONT SIZE="large"><FONT COLOR="red">andyt posted the following answer to this question on page 5 of this topic. Here is a copy of that post:</FONT></FONT>

Modified by nsxtasy at 8:53 PM 1/20/2006


Modified by nsxtasy at 8:54 PM 1/20/2006


Modified by nsxtasy at 6:43 PM 6/29/2008

1GreyTeg

Sounds like a job that would keep Bbasso busy and out of trouble for some time

RTW DC2R

yeah, good luck with that

Bradstard

It would be nice if we could just get a copy of the VIN database from Acura, but I doubt that would happen...

George Knighton

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by RTW DC2R &raquo;</TD></TR><TR><TD CLASS="quote">yeah, good luck with that </TD></TR></TABLE>
LOL.... That's about all I have to say about that, too.

I'm curious, but not enough to go through that. Maybe Mike DiCarlo can find a way to do it with a measure of automation.

canuckr

Great idear!

This might make it a little easier to firgure out an itr vin from another integra.. I think you left this part out in your post, if I read it right or wrong.. Still sleeping

According to my 97 and 98 helms:

JH4DC231cySnnnnnn

1=Vechile Grade
1=Type R
4=RS
5=LS
6=LS with moonroof
7=GS
8=GS-R

So all you really need to do is start with 00001 and work your way up.

I can do all the 97 ITR's

ZygSpeed

Nice detective work Guys.

tommymoose

I'm not sure about doing all that work, but at the very least its interesting. The timing of this post is funny because I just started a class that is all about applications programming with algorithems. Hmm, maybe by the end of the semester....

Rene M might be able to help with it. I remember seeing that he's a M.E. graduate so he's already done with all the courses like that.

nsxtasy

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by canuckr &raquo;</TD></TR><TR><TD CLASS="quote">Great idear!

This might make it a little easier to firgure out an itr vin from another integra.. I think you left this part out in your post, if I read it right or wrong.. Still sleeping

According to my 97 and 98 helms:

JH4DC231cySnnnnnn

1=Vechile Grade
1=Type R
4=RS
5=LS
6=LS with moonroof
7=GS
8=GS-R

So all you really need to do is start with 00001 and work your way up.</TD></TR></TABLE>

No, you misunderstood.

We know the format of the ITR VINs. However, if the last six digits are mixed in the same sequence with all the other Integras, you would have to try all of the numbers in the sequence (as many as 40,000) to find out which ones were ITRs and which ones didn't exist (because the last six digits were used on a GS-R or an LS).

IOW, I don't know whether or not the ITRs were numbered consecutively as 000001, 000002, etc. I don't think so, based on my own VIN.

tommymoose

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by nsxtasy &raquo;</TD></TR><TR><TD CLASS="quote">
if the last six digits are mixed in the same sequence with all the other Integras, you would have to try all of the numbers in the sequence (as many as 40,000) to find out which ones were ITRs and which ones didn't exist (because the last six digits were used on a GS-R or an LS).</TD></TR></TABLE>

Just wondering how you came up with 40,000? I would think that if its a 6 digit number with 10 possibilities for each place (0-9) that there would be 10^6 possible combinations (aka ALOT)

len

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by 1GreyTeg &raquo;</TD></TR><TR><TD CLASS="quote"> Sounds like a job that would keep Bbasso busy and out of trouble for some time </TD></TR></TABLE>

Hahahaha...

1GreyTeg

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by canuckr &raquo;</TD></TR><TR><TD CLASS="quote">Great idear!</TD></TR></TABLE>

Well since some of us have lotsa free time...


On a serious note hows the job prospects coming along Ry?


<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by canuckr &raquo;</TD></TR><TR><TD CLASS="quote">Still sleeping
</TD></TR></TABLE>

As always.

jOOr like the only guy I know who is pretty much still sleeping going into his first session. Quick but still sleeping

nsxtasy

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by tommymoose &raquo;</TD></TR><TR><TD CLASS="quote">Just wondering how you came up with 40,000? I would think that if its a 6 digit number with 10 possibilities for each place (0-9) that there would be 10^6 possible combinations (aka ALOT) </TD></TR></TABLE>

They start with 000001 and they increment by 1 from there, and they start over each year. I'm pretty sure they have a different sequence for each model, so that you have NSXs numbered 000001, 000002, etc, and Integras numbered 000001, 000002, etc. (I'm just not sure that there is only one sequence for ALL Integras, or if they have one sequence for ITRs and a separate sequence for other models, or one sequence for two-door Integras and a separate sequence for four-doors, etc.) At worst, this means that if they have one sequence for ALL Integras, and they sold 40,000 Integras in a year (like they did in 1997), then the highest sequence number would be 040000, and nothing higher. Just like with NSXs, where they sell 200 in a year and the highest sequence number is 000200.

canuckr

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by nsxtasy &raquo;</TD></TR><TR><TD CLASS="quote">No, you misunderstood.
</TD></TR></TABLE>

Yea I did misunderstand..My bad..

So far I have found 97 ITR's listed from 002891 to 003026.

Also created an excel page to determine check numbers (just enter the last few digits), but for some reason I can not get the right check number..

andyt

I cannot find anywhere to enter a VIN on the website?
I have a "My Folder" option, but not "My Acura".
Could you provide a URL to where the VIN entry is?

NM found it.



Modified by andyt at 11:44 AM 1/14/2006

Dave-ROR

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by canuckr &raquo;</TD></TR><TR><TD CLASS="quote">

Yea I did misunderstand..My bad..

So far I have found 97 ITR's listed from 002891 to 003026.

Also created an excel page to determine check numbers (just enter the last few digits), but for some reason I can not get the right check number..</TD></TR></TABLE>

I just created an Excel spreadsheet to provide the first 40 thousand VINS of each year. Anyone wanna spend some time testing? lol

I was going to post the spreadsheet but I figured some paranoid fool would cry about ITR thieves being able to steal more ITRs now.

Then I was going to say expletive it, but it's a 66mb excel file lol

12mb zipped. I am uploading it but I gotta go so if people really can't figure it out i'll post it later. I'll IM Ken the link in case he wants to post it in a bit.

hushypushy

by the way, you don't have to check every LS, RS, SE. you only have to check the DC2s, that's the GSR and the ITR. the others are DC4.

Dave-ROR

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by canuckr &raquo;</TD></TR><TR><TD CLASS="quote">

Yea I did misunderstand..My bad..

So far I have found 97 ITR's listed from 002891 to 003026.

Also created an excel page to determine check numbers (just enter the last few digits), but for some reason I can not get the right check number..</TD></TR></TABLE>

That's not how you generate the check digit (I read that to mean that you are using just the last digits to generate the check, I may have misread but read on anyways), it's just like any other check digit (credit card, bank scanline, etc), different methods/weights, but the same thing in the end.

In this case you convert letters for numbers, multiple by the weight, then add it to the rest.

For example, for the VIN: JH4DC23141S000001 (2001 possible ITR, seq number 1)

your formula is:
J-H-4-D-C-2-3-1
8+56+24+20+12+6+6+10

that part will never change (for an ITR), so no need to go into detail.

next add the year, which is the alpha converted to number, or number multipled by it's weight, the weight for this position is 9, so 2001 is 1*9, 1997 is a V, which converts to 5, so 5*9=45 (just to show a conversion).

Anyways back to the VIN at hand, a 2001.
so 1*9=9 so +9

now the S is constant so +16

so now we are at:
8+56+24+20+12+6+6+10+9+16

Now for the rest of the series, the weights go 7, 6, 5, 4, 3, 2.

So in the case of 000001 we have ((0*7)+(0*6)+(0*5)+(0*4)+(0*3)+(1*2) or basically +0+0+0+0+0+2

sooo...

8+56+24+20+12+6+6+10+9+16+0+0+0+0+0+2

is the base number for the check digit, now we sum them to get: 169

Now we do a mod 11 on that value, a mod is basically the remainer of a division of a number by another number (11 in this case. So, 169/11 = 15 with a remainder of 4, so the check digit is 4.

Now it is entirely possible that I f$cked up on one of those numbers, but it worked for my GSR VIN. (with DC239 of course) After checking it once I didn't recheck so if there's a mistake, oh well live with it.

Dave-ROR

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by hushypushy &raquo;</TD></TR><TR><TD CLASS="quote">by the way, you don't have to check every LS, RS, SE. you only have to check the DC2s, that's the GSR and the ITR. the others are DC4.</TD></TR></TABLE>

doesn't matter, the Integra line is numbers as a series, so you still have to check 000001 through 040000. There won't be a 000001 LS and a 000001 GSR.

Haleiwa-Brando

Not speaking on the behalf of Brad, but he's a pretty smart whip when it comes to computers and what not. Maybe we can get him to help out and write a program.

Dave-ROR

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Haleiwa-Brando &raquo;</TD></TR><TR><TD CLASS="quote">Not speaking on the behalf of Brad, but he's a pretty smart whip when it comes to computers and what not. Maybe we can get him to help out and write a program. </TD></TR></TABLE>

Getting the numbers isn't hard (in fact, it's already done), do a POST to acura's server to check and pull the returns isn't the most complicated (depending on how that vin lookup works anyways).. convincing acura that a flood of 200,000 hits being generated from your server isn't an attack against them might be...

You'd have to break it into MUCH smaller groups. Seems like a lot of work for not too much value.


I'm reuploaded that ZIP Ken, I forgot the S in my concat so I added it...

Haleiwa-Brando

I think this is a great idea. I wouldn't mind knowing so I can finally get a good nights rest for once. I've always been curious how many 97's there reallyare. If I had the computer skills I would do it.

George Knighton

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Dave-ROR &raquo;</TD></TR><TR><TD CLASS="quote">

That's not how you generate the check digit (I read that to mean that you are using just the last digits to generate the check, I may have misread but read on anyways), it's just like any other check digit (credit card, bank scanline, etc), different methods/weights, but the same thing in the end.

In this case you convert letters for numbers, multiple by the weight, then add it to the rest.

For example, for the VIN: JH4DC23141S000001 (2001 possible ITR, seq number 1)

your formula is:
J-H-4-D-C-2-3-1
8+56+24+20+12+6+6+10

that part will never change (for an ITR), so no need to go into detail.

next add the year, which is the alpha converted to number, or number multipled by it's weight, the weight for this position is 9, so 2001 is 1*9, 1997 is a V, which converts to 5, so 5*9=45 (just to show a conversion).

Anyways back to the VIN at hand, a 2001.
so 1*9=9 so +9

now the S is constant so +16

so now we are at:
8+56+24+20+12+6+6+10+9+16

Now for the rest of the series, the weights go 7, 6, 5, 4, 3, 2.

So in the case of 000001 we have ((0*7)+(0*6)+(0*5)+(0*4)+(0*3)+(1*2) or basically +0+0+0+0+0+2

sooo...

8+56+24+20+12+6+6+10+9+16+0+0+0+0+0+2

is the base number for the check digit, now we sum them to get: 169

Now we do a mod 11 on that value, a mod is basically the remainer of a division of a number by another number (11 in this case. So, 169/11 = 15 with a remainder of 4, so the check digit is 4.

Now it is entirely possible that I f$cked up on one of those numbers, but it worked for my GSR VIN. (with DC239 of course) After checking it once I didn't recheck so if there's a mistake, oh well live with it.</TD></TR></TABLE>

<FONT SIZE="4">Oh, my. All these years and we have only now just discovered God's Only Total Nerd amongst us.</FONT>

nsxtasy

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Dave-ROR &raquo;</TD></TR><TR><TD CLASS="quote">I'll IM Ken the link in case he wants to post it in a bit.</TD></TR></TABLE>

It's here, a zipped Excel spreadsheet with separate worksheets for each year. (Don't ask me why Dave included one for 1999. )

12 Mb zipped, 67 Mb unzipped.

andyt

I wrote a little program to do the calculation and then request the info.
As of right now I am in 1997 up to JH4DC231VS004000.
I count 280 Champ White cars so far.
Its really not taking that long, nor taxing my network connection, so I'll keep running it for each year, if you guys want?
I'm running this single-threaded and sequentially by vin, so it can't be hurting the honda/acura website.