Caliber 440 movement serial number dating

Vintage Gruen Watch Discussion Forum
Message
Author
User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#11 Post by MikeTheWatchGuy » Thu Jul 06, 2017 1:56 pm

Here's a link to the full spreadsheet with 26 values in it loaded into Tableau...

https://public.tableau.com/profile/mike ... ublish=yes

I made a crude Wadsworth serial to month/year formula so that the case serial number went from being a 1-year resolution to being a monthly resolution. It helps show the linear nature to the data.

You can easily "exclude" values that are clearly out of range by simply clicking on a value and choosing "exclude". You can also download the workbook that it's based on.

I pulled this initial data from HolbenFineWatches ebay watch sales over the past few years. He's not a big movement swapper. I wrote down data from at least 50 watches, but of those, only 26 were Wadsworth cases.

I've got more data that I'll add that has prefix values. My old data that does not begin with A, B, or C I just don't trust as I don't know when I started looking for and recording the prefix.
Image

Tags:

User avatar
Barney Green
Posts: 226
Joined: Mon Nov 23, 2015 2:30 pm
Location: Germany

Re: Caliber 440 movement serial number dating

#12 Post by Barney Green » Thu Jul 06, 2017 11:13 pm

I doubt that production was linear over the year. There were peak seasons like XMAS or graduation time just to name the two highest peaks. So I believe that in about April and October the production numbers should have been at least doubling or tripling the figures of let's say January production.Taking into account that the movements came by ship the movements for christmas should have been built in August / September.

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#13 Post by MikeTheWatchGuy » Tue Jul 11, 2017 10:46 am

I'm trying to capture and communicate to collectors that there are 3 "Fundamental dates" that you can attach to a Gruen watch, or most Gruen watches.

Here's one attempt to show these dates and to put them into a useful context...
Attachments
Gruen Watch  Dates - A Visual Guide.jpg
Gruen Watch Dates - A Visual Guide.jpg (457.2 KiB) Viewed 1450 times
Image

User avatar
JackW
Posts: 282
Joined: Mon Mar 26, 2012 12:20 pm
Location: Denver-metro area

Re: Caliber 440 movement serial number dating

#14 Post by JackW » Tue Jul 11, 2017 1:44 pm

Barney Green wrote:
Thu Jul 06, 2017 11:13 pm
I doubt that production was linear over the year. There were peak seasons like XMAS or graduation time just to name the two highest peaks. So I believe that in about April and October the production numbers should have been at least doubling or tripling the figures of let's say January production.Taking into account that the movements came by ship the movements for christmas should have been built in August / September.
No doubt in my mind that you are correct. I think it will be hard to pin down months of production given what we know presently.
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

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#15 Post by MikeTheWatchGuy » Tue Jul 11, 2017 3:00 pm

JackW wrote:
Tue Jul 11, 2017 1:44 pm
Barney Green wrote:
Thu Jul 06, 2017 11:13 pm
I doubt that production was linear over the year. There were peak seasons like XMAS or graduation time just to name the two highest peaks. So I believe that in about April and October the production numbers should have been at least doubling or tripling the figures of let's say January production.Taking into account that the movements came by ship the movements for christmas should have been built in August / September.
No doubt in my mind that you are correct. I think it will be hard to pin down months of production given what we know presently.
Yea, I'm graphing the data in a non-date way now. I'm mapping 2 serials against each other to get as much resolution as possible to see if there's a pattern that will emerge. I'll post the 440 my data shortly. I need to redo a bit of it.
Image

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#16 Post by MikeTheWatchGuy » Sat Jul 15, 2017 1:41 pm

I have freshly collected 440 serial numbers and Wadsworth case numbers. I am adding to it.

This graph is of 24 cases that I verified recently.

I've got another 100 saved yet to be mined which is what I'm doing now.

You can see and manipulate the data from this address:
https://public.tableau.com/profile/mike ... ublish=yes

This is the view I published.
440 Wadsworth Serial Number plot.jpg
440 Wadsworth Serial Number plot.jpg (212.62 KiB) Viewed 1397 times


The best way to view this graph is like the Milky Way. It won't be a line, but rather a "cloud".

I did not add a trend line on purpose. I think your eye will do it for you.

You can also make observations using this graph. Like... "All but one "A" or "B" series are above the 500,000 mark." It looks like A flipped over at Wadsworth G6.

And "B" is perhaps G8 and beyond.

You can click on the datapoints in Tableau and exclude them if you wish.


Try it! Click on the outlying datapoints and exclude them on the Tableau site. The graph will auto update, etc. Cool tool!
Image

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#17 Post by MikeTheWatchGuy » Sun Jul 16, 2017 10:31 am

Made more updates to the dataset and posted the update to the Tableau workbook that you can find online.

I'm frankly STUNNED that Tableau created a "Public" version of their normally $3,000 tool. They repriced everything such that individuals can get a copy for a few $100's, but even better, the FULLY FUNCTIONAL program is available to use for FREE as long as you make your data public :-) That's the catch and it's not a bad catch at all!

Create your data in Excel, then EXPLORE it using Tableau.

This scatter plot with the Prefixed movements having different colors, shows BANDS when movements were in use.
440 Serial Numbers.jpg
440 Serial Numbers.jpg (226.12 KiB) Viewed 1386 times
We can make these statements about this data using the colors alone
Of the 36 data points in this sample, all of the blank prefix movements were pre-G6
All "A" Prefix were no latter than G8
All "B" Prefix were G8 or G9
All "C" Prefix were L2 and L3

Here is a screenshot of the DESKTOP version of Tableau. It's identical to the "Public" desktop version. I'm including it to show you that with a few drag and drops, you can get the fancy graph shown.
440 Serials in Full Tableau.jpg
440 Serials in Full Tableau.jpg (358.04 KiB) Viewed 1386 times
This was created by dragging and dropping values from my spreadsheet to represent a "Scatter plot".
I "simply" (lol) made the 440 serial numbers one long sequence from 0 to 39,999. A 0000 = 10,000. B = 20,000. C 9999 =39,999. That "440 Extended Serial Number became my X Axis.
Then for the Y-axis choose the Wadsworth Case Serial number that I also made into a longer sequence. G0 = 0. G 999999 = 999,999 L0 = 1,000,000

I'll admit that this tool can be frustrating and time consuming as it's often trial and error to get the kind of visualisation you want because there are so many to choose from. But it is the most powerful visualization and analytical tool I know of and the fact that you can get it for FREE is astonishing.
Image

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#18 Post by MikeTheWatchGuy » Fri Jul 21, 2017 9:18 am

I've EMBEDDED the scatter-plot of the 440 movements and wadsworth case on the serial number page. I hope that it doesn't slow things down too much. It's such a great tool for showing this data.

The best way to view the graph is "full screen" on a PC. This will scale the graph so that all the data is visible. The view is flawed at the moment because you have to scroll to see all the data. That's a mega-no no for displaying data and I'm sorry about that. But, it's still very useful.

I liberally applied the "roughly 50% of the watches do not have their original movements" and removed "rogue data points". In other words, I fudged the data in a manner that fits my theory... how is THAT good statistics???

Maybe it's not, but it's all in an effort to simplify and try to find SOME order to the chaos, a dim flashlight in a dark cave is better than no flashlight :-)

Here's a before and after. I used the interactive nature of the plugin to do it. In other words, you can do it too... just click on a datapoint and choose "Exclude". TRY IT!! It's fun. See what you can find in this data!

I'm updating the table on a periodic basis as I collect and verify more entries. If you sent me serial numbers, thank you very much! I hope to have a system someday where anyone can enter the data directly in to a borg-Gruen-Mind-Share thing

Here's the original dataset
Gruen Cal 440 cs Wadworth - Full data set.jpg
Gruen Cal 440 cs Wadworth - Full data set.jpg (377.89 KiB) Viewed 1334 times


Then I tossed roughly 1/2 of the Greens and Reds. This leaves a trend that can be more easily segmented to form the actual dating table, the goal of this exercise.

I wished I had taken more statistics classes in college.
Gruen Cal 440 cs Wadworth - filtered data set.jpg
Gruen Cal 440 cs Wadworth - filtered data set.jpg (345.84 KiB) Viewed 1334 times
Image

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#19 Post by MikeTheWatchGuy » Sat Jan 27, 2018 4:23 pm

I'm taking a completely fresh look at the serial numbers, this time using the cleanest data I can get.

I've collected 32 data points so far for just the 440 movements cased in Wadsworth cases. This should put a pin in the 440 dating table once all the stats are closely examined.

Here's a plot of just these new data points.
Gruen 440 Cal Serials.jpg
Latest cal 440 movement data
Gruen 440 Cal Serials.jpg (246.29 KiB) Viewed 744 times
Not surprising it has a similar upwards trend when mapping the known Wadsworth case serials against the 440 caliber serial numbers.

Hopefully something a little more official will get completed in the next couple of weeks. I'll publish a final version of the 440 dating table at that time.

2018 should be a good year for the serial numbers. I expect a number of them to fall including the 435 and Guild-era movements. Cases are the bigger priority of them all as they determine the actual 'birth date' for a watch since it's the case that is the determining factor in dating Gruens. The movement dates help fill in the case dates and vice-versa.

I'm anxious to get the Quadrons done next as there are not enough Guild-era tools.
Image

User avatar
MikeTheWatchGuy
Posts: 349
Joined: Mon Mar 05, 2012 11:55 am
Contact:

Re: Caliber 440 movement serial number dating

#20 Post by MikeTheWatchGuy » Fri Feb 02, 2018 6:43 am

DONE!

After looking at 121 Caliber 440 movements cased in Wadsworth cases, I'm calling this version of the 440 Serial Number Dating Table complete!

I tried to use sources that have a history of selling Gruens that are in excellent condition and thus have a less chance of the movement being swapped at some point.

Code: Select all

440
Serial   Year
-------------
0        1940
2000     1941
3000     1942
6000     1943
9000     1944
A 0000   1945
A 4000   1946
A 7000   1947
B 0000   1948
B 6000   1949
C 0000   1950
C 2000   1951
C 5000   1952
C 9000   1953
If you have suggestions for changes, by all means post'em.
Image

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests