1967 GTO exterior color distribution calculation

[squidtone]

Hello all,
A couple weeks ago someone posted about how many Plum Mist GTO's were made. For years I also wondered how many Linden Green GTO's were made. This kind of inquiry is made often here on PY; not only for exterior colors, but for all sorts of other options and combinations.
As most of us know this information has never been parsed out of the records. PHS probably has most of the raw data, but of course it would be a huge job to tabulate all the colors and options, especially since there were 81,000 GTO's, and then all the other Pontiacs too.

So, a few years ago I started to do my own data mining. As some of you might guess, most of the data comes from online sales and auctions. Basically I was interested in the trim tag info. I gathered the body and roof color (whether vinyl or paint) and interior trim color.

I have been able to categorize 317 GTO's according to color. That is a sample size of about 0.4% of total GTO production for 1967. Note that I only gathered GTO data, no Tempest or LeMans data. I am no statistician, but I think the information I got was sufficiently randomized considering location, surviving cars, etc.
I put this information into a spreadsheet and calculated the breakdown of colors. I am using Open Office, and I'm not used to it and I had a hard time putting it into a decent format. (I don't have excel, which I am more used to)
Just to refresh: The trim tags (or PHS billing card) had a two digit code for body/top color, and a three digit code for interior color. I'm just posting here on exterior colors.

Of note on the exterior colors: I recorded several special color cars; three tiger golds and one sunfire yellow.
There were also a few cars that had two tone paint. (not resolved in this post)
I'm sure with the small sample size, there is some statistical "noise", but still, it's intersting that Starlight Black, Fathom Blue and Plum Mist are just about equally "rare".
Interestingly, what I found significantly more "rare" is a convertible with a color top (not black or white) (again, not shown here in this post)

Here's the list showing most popular down to least popular. I can't format it nicely, but you can see what's what. I'll see if I can post the bar chart later....

1967 GTO Exterior Color distribution
Colors / Percent
Tyrol Blue 16.1%
Signet Gold 15.5%
Regimental Red 14.2%
Cameo Ivory 6.9%
Burgundy 6.6%
Mariner Turquoise 6.3%
Silverglaze 6.3%
Montego Cream 5.4%
Montreaux Blue 4.7%
Linden Green 4.1%
Gulf Turquoise 3.5%
Champagne 3.2%
Starlight Black 2.2%
Fathom Blue 1.9%
Plum Mist 1.9%
Tiger gold 0.9%
Sunfire Yellow 0.3%



Sampled quantity: 317 / 100.00%

Total 1967 GTO production 81722

Sample size of total production: 0.4%


I just thought this would be interesting. I fee pretty confident it's probably pretty close to actual percentages. As time goes on, if I keep gathering info, I'll update if it changes much.
Any statistics buffs care to chime in?

[rare4k]

great work !

[60sstuff]

great work !

Very impressive work and informative numbers!

Thanks for posting.

You may already have these but here are a few 67 GTO's I have on file.

Ironically two of the below are the popular "Tyrol Blue" with White tops

[67GTONUT]

You can add mine to the list of Special Colors

Buick Riviera Charcoal Gray

Data plate is blank where color is supposed to be.... and I have original Window Sticker if you need documentation....

But I have talked to the original owner also....

[johnta1]

Do you also keep track of the VIN/body # for each of the tags?

If using ebay ads, a lot of times some get sold over and over (over time).

Cool job on the data tracking.

:)

[squidtone]

To be on the list (of 317 cars so far), my rule was to have a unique identifier so that I do not count the same car multiple times. So, I use the VIN. The "data mining" I do is from publicly posted information. Some folks get squeamish about VIN's, so I want to be respectful and never re-post any VIN's I come across.
(sometimes I feel like some of those English dudes who do "plane spotting"; they sit at airports and just record the registration (wing) numbers of planes that take off and land at airports....)
Anyway, in the grand scheme of things, for most vehicle sales today, particularly for modern vehicles, VIN's are prominently displayed on the vehicle, as well as on the information page when they're up for sale on the internet.
On certain online auction sites, it's pretty common for a vehicle to be posted several times, and sometimes with long stretches of time in between. Using the identifier (the VIN) eliminates "recounts" on my part.

I've actually got info on about 750 cars, but about half of them didn't have trim tag info, so in that case I just recorded what the car colors were presented as.

It just occurred to me that I should calculate what the color percentages are on those cars to see if it that sample lot has similar percentages.

Overall my "goal" here is to just have fun with information about 1967 GTO's. I never want to parlay this information for my own benefit; I want to let anyone have this statistical information who wants it. The big deal is to make sure facts get spread around, not misinformation.
As time goes on, perhaps interest will wane and no one will care, on the other hand maybe this type of stuff will help people enjoy the hobby and help it grow. Some folks don't give a whiff about "numbers" and options, but I think it's human nature for some of us to get interested in this stuff.

[squidtone]

Oh, and thanks guys for the positive comments!

As an aside, I wonder also what the color distribution is for 1967 Super Sport Chevelles....where Chevy buyers similar in their tastes? Or, was it really dealers who ordered the majority of cars (and what colors they were)

[Keith Seymore]

Statistically that sample size is known as the "how many can I get?" sized sample.

[Gator67]

This is cool Dave. To get a feel for the bounds of error given your sample size, take the square root of the following quotient: (percentage of a certain color*1-percentage of a certain color)/n. This gives you a standard error of the proportion. Then multiply by the value from the normal distribution at the confidence level your interested in (z). Then subtract (and add) this z value to the percentage of the color you're interested in. This will give you the lower and upper bounds of the percentage at the confidence level you chose. So for Linden green (.041*.959)/317=.00012403. The square root is .01113709. If I want 95% confidence I multiply by 1.96, which gives me .0218287. I subtract and add that number to .041 and I get 1.92 to 6.28. This means I'm 95% confident that the percentage of Linden Green GTOs falls between 1.92 and 6.28, or between 1,567 and 5,134 cars. You'll see that if you want to be less confident, the interval will narrow. You'll also see that the standard error will be wider the higher the percentage--probably closer to 4% for the top three colors. If you want to make comparisons between colors, look to see if the confidence intervals overlap. I suspect you'll find three bands at the 95% confidence level.

[gtospieg]

Interesting info squidtone...My car is Regimental Red, not all that uncommon, but it has a red interior (and it's a post coupe). I wonder how common (uncommon) that is? I would guesstimate that if I've seen 100 Regimental Red 67's 90% of them have had a black or parchment interior. That would really make for some interesting statistics.

[Diego]

While I am not always thrilled by estimates, this may interest you:

http://www.camaros.org/exterior.shtml#ColorPercentages

They can't be analogous for several reasons (different target market, several different colors), but it bears mentioning nonetheless.

[b-man]

Standard 335 HP 400 engine with factory A/C, TH400 auto with his-n-hers Hurst shifter on the console, power steering & power drum brakes, rear window defogger, Tyrol Blue paint/parchment interior/white vinyl top.

http://i137.photobucket.com/albums/q230/b-man64/1967%20GTO%20hardtop/DSC00070.jpg

[carcrazy]

Thanks for putting in the effort. Useful information!

[squidtone]

Gator,
WOW, I find that formula fascinating. What is that process called? I notice it doesn't seem to reference the actual sample size compared to the total cars built. At what point does my sample size matter less and less? What if I eventually got info on 50 percent of all 67 GTO's built...wouldn't that formula give me the same confidence numbers? Even though (in my mind) I've really got actual numbers on half the cars, and presumably better confidence in the accuracy?
Well, it'll be interesting to see how the percentages track as I collect more data. I'll keep going as long as I can.

GTO Spieg, I have also collected interior info...I'll try to break down those combos soon. But I just looked and of the 45 Regimental GTO's, 7 of them, or 16% of all Regimental Red cars had Red Bucket seat interiors.

Diego, I just looked at those "estimates" for the Chevy Camaro of 1967, and the numbers tracked pretty closely with the GTO colors. Although they had a popular butternut yellow that cut into the Marina Blue (Tyrol blue for Pontiac), Red, and Grandada Gold (Signet Gold for Pontiac) I'd guess.

B-man, wow that is a beautiful GTO on the trailer. Something about patina'ed, straight, rust free GTO's. Sometime I think a semi-gloss or flat finish GTO option would have been popular new! (Jeep did that for a couple years in the 90's on some Wranglers.)

If anyone sends me info, I'd be happy to add it to my list. I don't track anything that doesn't have a VIN though, because if I did, I would never know if I'm recounting it later.

I will post again with more when I parse it out.

[fyrffytr1]

Interesting. Thanks for taking the time to do this. My car is Mariner Turquoise with dark aqua interior and black top(convertible). According to PHS it came with an off-white pin stripe but I can't make myself put it back on.

[squidtone]

fyrffytr1,
Where on billing history card/PHS docs do you see the pinstripe color? I didn't think that was tracked by the factory...

[geeteeohguy]

Very, very interesting. And from what I'm looking at from a Left Coast perspective, accurate. Out here, Signet Gold seems to be the most common color, followed by Cameo Ivory and then the Reds/blues. I can't remember seeing a Tyrol Blue '67 in the Northern part of the state in the past 30 years, though. I've seen a couple of Linden Green cars, but no Tyrol Blue. Seems like every '67 is or was Signet Gold, including mine. Funny also: Ponitac was really pushing the gold interior in '67: it was available in every single body color. Nuts! Thanks for the good work, Squid.

[george kujanski]

Squid; you can add mine to your list: gulf turquoise, aqua interior, black vinyl top. Now, coincidentally, the exterior is painted Tyrol blue, soon to be changed to some other color.

George

[squidtone]

GeeTeeOhguy,
I wonder too how color choice ranked depending on what part of the country they were sold in. But the way I'm compiling cars, I believe they are truly randomized due to decades of cars moving all over the country, and most of the cars I get data on are on nationwide sites anyway.

I spent a week this past spring traveling from SF down to Cambria, then over (through Fresno) to Yosemite NP. It was dry as a bone. I couldn't believe how dry it was. No one gave us water in restaurants; we had to buy bottled water. I felt bad for everyone. I saw pickups with big plastic (hundred gallon size?) tanks trucking water around...I thought that might be for landscaping or something.

While stopping in Fresno, we came upon a big car show in a shopping mall parking lot. I didn't have time to stop unfortunately. But anyway, I saw no muscle cars at all during my whole trip down from SF, down Rt 1, or anywhere else except that show in Fresno. I take that back, some guy had a 70 chevelle in one of the redwood forest parks, but he was from Virginia! But I did see about 5 billion new convertible mustangs and maybe an order of magnitude less convertible camaros (new ones). Saw many old pickups though. In fields and stuff.

So, anyway, California sure is a different place for this ol' yankee New England boy. I'm sure color rankings were skewed depending on left, middle or right coast!

[400 4spd.]

Squid, I can provide you information on a few more 67's. It will probably be after the 1st of the year, though. Red, Gold, Maroon, White. Three from Baltimore, one Kansas. Do you need pictures of data and VIN tags?

[Diego]

I have statistics on the convertible distribution for Mustangs and Cougars, but not colors. You'd be surprised to learn that the strongest states for convertibles were not Sun Belt but, rather, Rust Belt states.

[Gator67]

Gator,
WOW, I find that formula fascinating. What is that process called? I notice it doesn't seem to reference the actual sample size compared to the total cars built. At what point does my sample size matter less and less? What if I eventually got info on 50 percent of all 67 GTO's built...wouldn't that formula give me the same confidence numbers? Even though (in my mind) I've really got actual numbers on half the cars, and presumably better confidence in the accuracy?
Well, it'll be interesting to see how the percentages track as I collect more data. I'll keep going as long as I can.

It's just basic stuff that you can find in a college level stats textbook. I just got done teaching a PhD seminar on research methods, so it was fresh in my mind. There are other statistical approaches to the issue, but this one's straightforward. We just have to assume that the sample you have is drawn randomly from the population (81,722). it might not be true if, for example, the survival rates are higher for cars with colors preferred by older more responsible drivers. If you play with the formula with a specific color percentage, it's easy to see where the standard error levels off. With Linden Green cars, going from a sample of 100 to 300 will cut the standard error in half, approximately 4% to 2%, but you'll need a sample size 700 to get the standard error to 1%. With half the population, you'd still have a standard error of .2% (interval of about 300 cars). You might notice that you'll still have about .1% error even with the full population…this is because the formula assumes a population that is theoretical. Although I might feel uncomfortable about this remaining .1% error there would likely be errors of various sorts in the 82,000 data points, so .1% error would likely be generous.

[squidtone]

400 4spd
Sure, thanks. Send it anyway you want. If you send pix I'll transcribe and then just delete the pictures. Or if you send just text, just send the VIN, then the 2 digit color code, and the 3 or 4 digit trim code. I usually note the (usually "B") suffix if it's taken off the trim tag on the firewall.

Gator67,
Thanks for the info. For now I'm content just compiling data. As time goes on I'll keep updating and including the statistical info too.

I have to make up a spreadsheet so that it updates automatically when i add new info. Right now my spreadsheet is a bit of a mess (chart-wise).

Thanks and have a good Christmas all....


Dave

[Nicks67GTO]

I'll check mine when I get home. It was either that Sunfire Yellow or Montego Cream/ black vinyl top from the factory but I switched it to tyrol blue.....

[YNOBIL]

Dave Here's my list of cars I own, once owned, or have positive knowledge.

242177P261813; Color: F2; Trim: 219
242677P264654; Color: G2; Trim: 221
242177P298174; Color: F2; Trim: 219
242677K126223; Color: L2; Trim: 223
.................................................. ..........
I undercoated this new GTO when I worked at Kole Pontiac in 1967 / 24217K144108; Color: NN; Trim: E (taken from Kansas Plant Build Sheet)

[YNOBIL]

Dave Found another one

242677K125832; Color: F-2; Trim: 223-B

[squidtone]

Ynobil,
Got 'em. But, what do you mean by "E" on the trim code on the 5th car in your second to last reply?

[YNOBIL]

Ynobil,
Got 'em. But, what do you mean by "E" on the trim code on the 5th car in your second to last reply?

That's what the Build Sheet listed. Apparently Kansas was mainly a Chevy Plant and used Chevy Codes (Note "Chev" at the top left corner). Here's the Build Sheet.

http://i1357.photobucket.com/albums/q744/wmnawrot/1967%20GTO%20Build%20Sheet%20-%20400HO%204spd_zpsnjs3lrud.jpg (http://s1357.photobucket.com/user/wmnawrot/media/1967%20GTO%20Build%20Sheet%20-%20400HO%204spd_zpsnjs3lrud.jpg.html)

[mountaingoat]

242177z122537...HH....224
freemont plant