Caliber 440 movement serial number dating

Vintage Gruen Watch Discussion Forum
Message
Author
User avatar
MikeTheWatchGuy
Posts: 348
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Caliber 440 movement serial number dating

#1 Post by MikeTheWatchGuy » Mon Jul 03, 2017 11:34 am

Out of all of the movements that I hoped to decode the caliber 440 Curvex movement was near the end of my list. The reason is I [erroneously] thought the value was without meaning. What I was ignoring was the letter prefix that I discovered some time back. It moved the dating using 440 serial numbers from an impossible task to a trivial one.

The 440 Caliber movements serial number format differ from all of the other "modern", post 1940 Gruen movements. It is a 4-digit number with an optional single letter code A, B, C (or none).

Let's take a quick look at a 440 movement.
440 A Labelled.png
440 A Labelled.png (1.84 MiB) Viewed 1387 times
Prefixes on serial numbers have several potential meanings. It could mean a particular "factory" manufactured the movement or it can signify "time" or "batch". It's clear that the Letter Prefix signifies a "batch" of 10,000 watches made beginning at a particular time versus signifying the factory the movement was manufactured. The problem we all saw with 440 movements is the 4-digit serial number can ID 9999 movements. Thus, the Prefix steps in and saves the day by tacking on another digit to the serial number.

You can validly read the prefix as a 10,000 adder. Prefix A = add 10,000. Prefix B = add 20,000. Prefix C = add 30,000.

To create this Cal 440 Manufacture Date Table, I examined caliber 440 movements that was in Wadsworth cases. This gave me the ability to date the case. The theory being that the case and movement were likely manufactured NEAR each other in time. The delta of time between the manufacture date of the case and the manufacture date of the movement can be debated. It could be 1-3 years difference in either direction. It's not surgically precise, but it's a whole lot better than having no other data.

I'll be posting more information and updating this table over on the GruenWristwatches site on the Serial Number page.

OK, so here goes.... an attempt to decode the 440 movements. An unofficial caliber 440 manufacture date table!
Serial Date
0000 1940
- 1941
4,100 1942
5,000 1943
7,400 1944
A 0000 1945
- 1946
- 1947
B 0000 1948
C 0000 19??



If you happen to be poking around in a 440 watch soon, check it against this table... please provide feedback!!!!

----------------------------

This is where I would normally insert text warning of the dangers of dating using movement serial numbers but I'll spare you and instead say that this date is but one of MANY date you should be collecting about your watch as you research it. It should be one data-point among as many as you have the patience to collect which should be equal to the maximum data points you can possibly collect :-)

----------------------------
Image

Tags:

User avatar
thojil
Posts: 88
Joined: Tue Apr 16, 2013 12:05 pm
Location: South Africa

Re: Caliber 440 movement serial number dating

#2 Post by thojil » Mon Jul 03, 2017 11:52 am

Great piece of analysis as usual Mike.
Here are the serial numbers of the six 440's that I have.

8551 in a "G3" case
A1299 in a "G4" case
A4321 in a "G7" case
B1298 in a "G8" case
B3006 in a "G9" case
C2430 in a Falcon Rist-Side case (C27761 APEX)

It seems that my serial numbers follow your dating assumption.

User avatar
MikeTheWatchGuy
Posts: 348
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#3 Post by MikeTheWatchGuy » Tue Jul 04, 2017 10:20 am

Thanks for those! Fantastic values to encounter. Not sure I could ask for better.

I'm wondering if we can agree on the approximate chance the 440 movement in a vintage Gruen is original. Take not all watches ever made by Gruen, but just the 440s.

With those, could you answer this question with any degree of certainty?
X out of every 10 Vintage 440 Curvexes are likely original to the case

There are certainly factors that we can throw in that will raise or lower that number. Who you're buying it from matters if it's for sale. If it really does look "New Old Stock". Your own experiences and senses should be brought to the question for sure.

Or, if you can't answer that directly, can you rule out certain stats?

I think about 50% is a generalized approximation if you do not take into consideration anything at all about the source. Some eBay sellers, for example, have one a very consistent basis, Gruens that are clearly been well cared for and are higher end watches. When I pull data from these listings I have an overall better confidence in the data.

Some sellers of movements for parts or low end multi-owner Gruens maybe it's 2 in 10 have the original movement?

I don't think it's an obscene number of movements have been swapped to the point that this information is useless. It's a tad "junk science" feeling when talking about the movement manufacture date in a vacuum, but like in Thojil's post, the Wadsworth case dates bolster the confidence.

I would like to put some words with the table that informs the reader better as to how much confidence to put in their data. We're seeing that the date tables do cross-check, so I've got confidence in the growing numbers of tables. The lack of confidence comes not with the tables, but with the watches being examined.

Thoughts on ways to take the knowledge we as a group have built by looking carefully at the interiors of thousands of watches?
Image

User avatar
MikeTheWatchGuy
Posts: 348
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#4 Post by MikeTheWatchGuy » Tue Jul 04, 2017 11:35 am

Having this serial number decoded it gives us a glimpse in the "uptake" of the Curvex! At least through the 1940s as the 440 was the only movement for the majority of that time.

The numbers seem to say....
The FIRST 10,000 Curvex 440s were sold over 6 years
The SECOND 10,000 of them over the next 3

I don't recall ever having any kind of hard numbers when it comes to Curvex sales. This gives us not just ballpark numbers, but we get the seat number in the ballpark.

It's far fewer than I would have guessed, but the earth's population was smaller then too.

Hey, let's compare today's Curvex sales!

Must do that back of the envelope! 10,000 over 6 years is 1,666 per year. Or 4.57 a day
eBay claimed 219 sold over the past 90 days or 2.43 a day.

The ebay numbers are not limited 440. That's limited to ebay sales. Who knows how Gruens are sold on OTHER sites. Maybe just as many??

You can make some astonishing statements from those numbers :-) Like, today worldwide sales of Vintage Gruen Curvexes are pretty close to the same as when it was launched in the 1940s. Even if off by many orders of magnitude, the numbers are still relatively SMALL in the grand scheme of things.
Image

User avatar
thojil
Posts: 88
Joined: Tue Apr 16, 2013 12:05 pm
Location: South Africa

Re: Caliber 440 movement serial number dating

#5 Post by thojil » Wed Jul 05, 2017 3:36 am

MikeTheWatchGuy wrote:
Tue Jul 04, 2017 10:20 am
I'm wondering if we can agree on the approximate chance the 440 movement in a vintage Gruen is original. Take not all watches ever made by Gruen, but just the 440s.
With those, could you answer this question with any degree of certainty?
X out of every 10 Vintage 440 Curvexes are likely original to the case
That's an interesting question which I guess will be answered over time collecting enough data points. I recall from all the quantitative market intelligence surveys that I have been involved in that any sample size lower than 70 should be considered insignificant and therefore not reliable.
Given the frequency in which 440's are available on the swamp it should not take too much time to collect enough data points. That of course is assuming one has the time to spend to do this :roll:

The point you make on swapping movements a few thoughts from my side.

I'm not "in the know" how common it is for sellers and watch makers to swap movements. I can imagine that back in the day when all watches were still mechanical and parts widely available that repair was the norm instead of swapping complete movements?
In more recent times with the web providing an easy source of complete movements I agree with you that it is more economical to swap movements instead of repairing them. Especially with brands that do not exist anymore.

I agree with you that for the well established sellers (eBay or other) you should be able to trust they have a professional honor to keep the watch they are selling original. Usually this is also reflected in the price they ask or what people are willing to pay when buying from such a seller.

The danger is with the ones that are "specialized" in selling watches without really having knowledge or emotional interest in them. Obviously it is not easy to identify them.

I personally put trust in the "real junkers" that are widely offered as normally there would be no interest from the seller side to waist any time on this for an unknown brand like Gruen. And I trust the ones even more that are not in running condition. Also I always check what else the seller is selling. If they are not specializing in watches and/or watch parts the chance that at least this seller has not touched to movement is bigger. Obviously there is no way in knowing wheter the movement wasn't swapped at some point in its history. But over time you create an eye for identifying the good ones, which is often as you say a gut feeling that is difficult to explain :ugeek:

Probably what I'm saying here is that I have no clue what would be a reasonable assumption to take as a percentage for non original movements and your guess is as good as any guess...I guess :lol:
However I think assuming that 2 in 10 movements is original to the case may be a bit low when I follow my own assessment experience. Maybe 5 out of 10 is a save assumption. What has been your experience when you worked on dating the 405 calibre family, which I assume was based on the Wadsworth case serial number knowledge?

But like you I'm a great believer in data. Gather enough data and one will see the pattern emerging.

User avatar
MikeTheWatchGuy
Posts: 348
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#6 Post by MikeTheWatchGuy » Wed Jul 05, 2017 8:32 am

Oh, I didn't do the trailblazing movement serial numbering of the 405 series. The 405 series was done by Jack Wood.

You're correct about needing a fair number of data points. I used 48 ebay listings of 440 Curvexes from Holbens as the core data set.
Even though I've got scads of data from prior years, because that 440 Serial Prefix is easy to miss, I tossed all the old data and started fresh for this exercise.f
Some of those were clearly movement swaps so they had to be tossed to.

Here's a graph of the remaining data that shows a strong correlation :)
Caliber 440 Curvex Serial Numbers.jpg
Caliber 440 Curvex Serial Numbers.jpg (61.12 KiB) Viewed 1342 times

The more data the better of course... and I'll be adding more as time goes on now that I've got this serial pealed off from all the others. They made it particularly easy to work with using a 4 digit serial and a single letter prefix. I don't know if it was by design so that every could easily see by picking up a Curvex watch, how many of them had been made so far.

In 1943, presumably one could pick up a watch being timed and accurately see on the movement the total number of 440 movements made to date.

As the Movement Catalog is just getting rolling, I am unsure how many other movements share a similar ease of reading. I recall some of the movements having short serials, but don't remember which nor how short off the top of me head. I'll certainly publish more as I learn. Adding a new field to the movement Catalog "Serial Number Format" to help in identification and also finding which movements are perhaps in the same "Family" from a serial number point of view. This is what Jack, no doubt, needed to do to find the family of movements in the 405 serial numbers.
Image

C. Hurt
Posts: 434
Joined: Thu Jun 02, 2011 8:54 pm

Re: Caliber 440 movement serial number dating

#7 Post by C. Hurt » Wed Jul 05, 2017 6:53 pm

I think you're onto something, but as with most things Gruen, it's rarely simple or straightforward. This one quite clearly is marked with a "2". This pre-supposes there must be some marked with a "1". Where would these fit in?

Image
Attachments
image.jpg
image.jpg (686.62 KiB) Viewed 1328 times

User avatar
thojil
Posts: 88
Joined: Tue Apr 16, 2013 12:05 pm
Location: South Africa

Re: Caliber 440 movement serial number dating

#8 Post by thojil » Wed Jul 05, 2017 10:42 pm

The exception confirming the rule maybe ... :lol:

Do you happen to have a case to go with that movement?

User avatar
MikeTheWatchGuy
Posts: 348
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#9 Post by MikeTheWatchGuy » Thu Jul 06, 2017 11:58 am

I've not seen a 1 nor a 2 that I can recall.

If we look at how easy it is to read these 440s compared to all other Gruen serial number series, it's perhaps the EASIEST to follow, decode, and date, if the table works out in the long run.

As I think is being implied by "where the beef?" err case, is perhaps movements marked with 1 or 2 were issued as replacement/parts movements.

If it is a marking indicating a replacement movement then the 440 may be the most collector friendly serial numbering system of the movements. No doubt that Wadsworth currently has the "Ease of dating" record on case serial numbers.

And this ability to read how many Gruen 440 Curvexes have shipped to date that spelled the downfall for using it with any other movements. Too many trade secrets would be easy to find????

I could see Gruen wanting to be kinda cagey with their specific sales numbers and if the only thing a competitor needed to do to find the sales numbers all they need to do is open one of the newest Curvex arrivals from Gruen and read the number from the 440.
Image

User avatar
JackW
Posts: 278
Joined: Mon Mar 26, 2012 12:20 pm
Location: Denver-metro area

Re: Caliber 440 movement serial number dating

#10 Post by JackW » Thu Jul 06, 2017 1:26 pm

Hi Mike,

What I'd be curious to see in your graph is the R^2 value for your data and an error or standard deviation calculation. You said you got your data from one single reliable source.

However, I wanted to see if you are on to something here. I took data that I had from several years ago but also went through eBay to find cased movements with pictures of the case serial and the movement serial. Those that I could not discern I ignored. There were some with dates engraved on the case too. This is the red line (with n=18). There is no question on if I got the extra part of the serial that is on the barrel bridge. I did not see anything other than 'blank', 'A' or 'B' no other letters or numbers in my small sample.

I took the larger data set (n= 34) including the older research (blue line), but I seem to have noted in the spread sheet if there was a mark 'A' or 'B' on the barrel bridge. There were some that were marked as questionable and these were left off. But I am not 100% sure. I want to use these because of the number of inscribed dates from case backs. These are from the Wadsworth project.

Here is my graph of the two data sets.
graph.jpg
graph.jpg (54.71 KiB) Viewed 1297 times
The R^2 for the blue is 0.591 and the red is 0.741. That isn't too bad; especially since I have a bit more confidence in the data collected for the red-graph line that I caught the extra digit. The R value from the red line is pretty good. I bet Mike, that your data will give one even higher.

I need to do some more work to figure out what the error will be. The statistical analysis I did for the Wadsworth cases likely won't work for this type of data. Once the deviation is figured out, it likely can be used to ID swapped movements. There are wrinkles such as what Cary shows that need to be addressed.

I too think you are on the right trail. Great work Mike!

More to come...
All I know is based on hard work & writing by others. I can only aspire to augment this body of knowledge. If I am wrong it is because of my own failings. -me
"If I have seen farther it is by standing on the shoulders of giants." - Newton

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests