Mike Frye

Jim is right, getting the correlation between the VIN and the owner (or even former owner) would probably be a deal breaker no matter how you want to obtain it and definitely for reproducing it.

However, if you just call the nice people at Porsche and give them a VIN, they'll tell you all of the options it originally delivered with.

Obviously that info is in a single DB somewhere, and they don't consider it proprietary since they don't require any kind of verification or charge you to give it out. Maybe they could be persuaded to publish a list of all VINs of all years, with the codes and what they mean (which I know is already available elsewhere).

Then you could cross reference that with the carfax data (proprietary alert) in order to get what you're looking for (I think). As far as publishing it, I think that's going to be a problem.

Cool idea though . Maybe it would be worth money, but I doubt enough to cover the cost of licensing the data from Carfax or wherever else you obtained it.

kary4th

I'm working on all VINs for all years. I have that data in the book I've mentioned.

danglerb

To publish or even make public we would have to go to original sources for the data, not just carfax, but something with a doorway for access so distribution is limited to rennlist members or something like that might be ok, or if not ok something we could get away with.

928 Intl might want to consider "renting" their contact info for selected mailings. Its profitable for many companies that accumulate mailing lists, and the data is protected by using a bonded mailing house so the person renting the list never has access to it, address labels just get attached to the item they want to send and those go directly to the postal system.

Non specific information like how many this or that, or state by state, that kind of thing, even the figure for how many are still on the road is very interesting.

kary4th

I'd like to have an idea of how many of each year remain - how many are no longer registered or are known to be destroyed.

danglerb

Year World USA VIN
77 175 13 928 820 xxxx
78 1070 1584 928 820 xxxx
79 5438 1626 928 920 xxxx
80 4278 1824 92 A0 81 xxxx
81 3821 2036 WP0 JA0 92xBS 82xxxx
82 4570 2140 WP0 JA0 92xCS 82xxxx
83 3999 2746 WP0 JB0 92xDS 86xxxx
84 5095 2175 WP0 JB0 92xES 86xxxx
85 4460 2582 WP0 JB0 92xFS 86xxxx
86 4892 2780 WP0 JB0 92xGS 86xxxx
87 4791 1967 WP0 JB0 92xHS 86xxxx
88 4358 1427 WP0 JB0 92xJS 86xxxx
89 3627 835 WP0 JB0 92xKS 86xxxx
90 3196 620 WP0 JB0 92xLS 86xxxx
91 1792 263 WP0 AA2 92xMS 81xxxx
92 1132 183 No US Production
93 730 120 WP0 AA2 92xPS 82xxxx
94 470 84 WP0 AA2 92xRS 82xxxx
95 104 95 WP0 AA2 92xSS 82xxxx
96 5
Totals 57,998 25,105 Note "x" is generic and varies
Copied from chart on 928 spec, world includes USA
I would have just linked to them, but didn't see how with frames etc.

I think jim said about 16,000 were still registered in the US, which is an incredible 75% or there abouts.

danglerb

I need to play with some valid numbers to figure out how to calculate the check digit.

In the bad ebay deal in FL thread the fellow says FL DMV has searchable DMV by VIN, so maybe other states do as well.

In Calif the smog tests are searchable, which gives license number and indicates at least its registered and pass/fail smog history. http://159.145.15.175/vehtests/pubtstqry.aspx Once I know how to do check digit I will try a few there.

Vilhuer

Someone in here has made page where he shows how to count check digit using S4 VIN as example. I can't never find link to it when its needed. Here's generic example how to do it.

If you actually plan on doing complete list of US VINs I suggest you at least start from correct production numbers. Those you list are in reality how many cars were registered in USA or USA+Canada each year. VIN of all those could be basically anything as long as car actually existed at time of registration. So you can't make accurate VIN list using those yearly totals. You end up doing thousands of unnecessary checks unless list is based on right kind of data. Only place I know which has it is Jörg Austen's book. Even it has problems in at least '89 MY data IIRR.

Also if you base VDS code part of the VIN on that 928GT.com site list you might end up with wrong VINs. I think 4th to 6th digit are at least wrong on '90 when airbags were introduced after only few '90 MY US cars were made without them. Other problem area is early (really '92) and late (real) '93 GTS. At least actual production numbers are known for them so list can be done fairly easily.

Correct VIN list is certainly doable to some degree. Checking Carfax agaist it could provide answers like how many of prototype and special 0001...0060 VINs were made each year. Which in turn would tell what was actual number of cars made each MY. Knowing largest VIN for each year will not in itself tell anything other than maximum number of US car for said year. Actual total can only be known when its know how many protos were made.

Would be very interesting to find out how much data Carfax has for all of them. I'm just afraid that their site would not like that large data mining operation no matter what 30 day free use purchase agreement says.

Normy

All this is interesting [and you people know a hell of a lot more about computers than I do!], but the very minor problem that I can see with setting it up is that you're probably not going to be able to search out ROW cars. That's not many cars, but it still misses a few.

Many state's DMV computers reject the ROW VIN's, and I think CarFax does too- the "ZZZ" of the ROW VIN throws them off: WP0ZZZ92ZFS841526. When I titled my car in 1999, the lady at the tag office had to call the state DMV headquarters in Tallahassee because the computer wouldn't let her register the thing-despite the fact that it had spent the previous 14 years registered in Florida.

Interesting project...

N!

Vilhuer

Had few hours spare time other day and fooled around with Visual Basic code. It was fairly easy to make code which creates US '81-95 VIN numbers on fly, enters it to Carfax site, reads result web page and stores relevant info into tab delimited text file. It takes about 8 seconds to get info for each VIN. Started code last night and in the morning it was going in middle of '82 MY and had checked about 3500 VIN's. I expect it to get last '95 MY info and everthing in between in about 36-48 hours.

As I don't have payed account, code is just collecting basic info of how many records are available for each VIN. I expect it would be similarly possible to do code which gets whatever data is available by paying. It would take about weeks unlimited access to get all that info into Excel sheet or database table. This should work fine unless some session variable in the site or something similar does prove to be too difficult to go around in code at payed pages.

So far even this little amount of data I have has proven interesting. For example it seems all early '81 MY VIN's were used meaning there is records for almost all of the 0001...0060 sequence numbers. For '82 MY records start at 0050ish. On other hand '81 MY records stop approximately 100 cars short of what is listed by Jörg Austen as largest '81 MY US VIN. Data will most likely reveal other similar facts also. Code is checking all MY's from 0001 to largest VIN listed by Austen + 100 to be sure what is highest VIN with records for each year. In case there are records for cars close to upper limit in current seach its easy to do other one with smaller set to find out true upper limit for each MY.

Search is limited to only to US models in USA I think but data will give climpse into Canadian cars also if they show up as 0 results VIN's in Carfax site like I suspect. In '81 and early part of '82 MY there are surprisingly few VIN's which have 0 results. Maybe 1 in 20 on first 3500 cars do not have anything recorded. I think this could be fairly close to percentage of cars which were sold new into Canada.

Mike Frye

Erkka,

That's cool. Nice work. It's nice to see that I'm not the only geek that sits around writing code when he's not writing code for a living.

Now if you could come up with a way to make money on this...

Rick Carter

I'm posting just so I'll know when there are info updates.

928cs

Just curious, could someone ask CarFax this VIN:
WP0ZZZ927FS841526 ?
Does it provide information on Normy's car?

(If i'm right, 7 is the check digit for its VIN).

Samuel

928cs

Interesting, it seems that the two US '88 CS I look for are not WP0AB092ZCS865061 & WP0AB092ZCS865062 (from Jörg Austen's book) but WP0JB0922JS865061 & WP0JB0924JS865062.

Can someone have more information about those cars?
Thanks in advance.

Samuel

tdelarm

Nooooooooooooooooooooooooooooooooo...was the first thing popping into my mind as my soul fell to the floor in Zen fashion, suddenly, as the crowd dispersed, and defibrillators pulled, sounds of Whoville once again echoed through my mind and I realized...this is an old Danglerb thread...not a new one...yes my son...Rennlist lives yet again

Vilhuer

7 is correct check digit for it. As with all ROW ZZZ VIN's, Carfax can not understand them no matter if actual cars are in US or not.

JB0 is correct VDS code for '88 and your VIN's have correct check digit. Carfax do not show any records for them. If my suspicion is right this means they were registered into Canada or two US CS have in reality another VIN numbers we don't know yet. I'm running few hundred VIN's at end of large list trying to hunt down these two cars. Most likely do not find anything but it doesn't hurt to try. I expect them to have normal '88 US VIN somewhere in middle of close to 2000 S4's made in that MY. This makes them impossible to find with Carfax.

Program code I did can easily calculate check digit for just about any VIN one wants it to. Carfax site recognizes for example WP0JB0926JS869999 as US model '88 S4. Since car with so high VIN sequence number obviously do not exist this proofs Carfax site do not have any idea how many 928 were done in each year. It only knows what could be '88 S4 VIN based on WP0JB0926JS8 part of the VIN. Site does not even know that US model should have WP0JB0925JS86 instead of any other number as 13th digit. As long as check digit checks out it is happy to report there are no records for such VIN. Value in search I'm running is that it will tell which VIN's have records in Carfax database. This will tell pretty accurately how many cars were made in each MY for US market.

For example '82 MY records also end 100 VIN's short of what Austen says was largest VIN for that year. So either factory exported last 100 cars to Canada or somewhere else or more likely '82 is already second year out of only two where Austen numbers do not match reality. To find out such info is whole purpose of this exercise. During past day code has gotten slower and it might take few more days to complete list its going through.