Establishing a production number estimate for the Grand Seiko First
Introduction
The Grand Seiko “First”, also referred to as the “3180” (its caliber number); “J14070” (model number); or, probably most accurately simply “The Grand Seiko”; was introduced to the Japanese market on December 18th 1960.
Created by the Suwa Seikosha company – now Seiko Epson – based in the Nagano prefecture in central Japan, it was born out of a desire to create a watch that was “as precise, durable, easy to wear and as beautiful as possible”. The Grand Seiko “First” was the start of what was to become a fascinating journey of continual development and innovation to create watches worthy of competing with – and indeed, beating – the very best that the Swiss had to offer.
The new calibre “3180” was designed to be accurate to +12/-3 seconds per day, and was the first Japanese watch to receive a rating of excellence from the Bureaux Officiels de Controle de la Marche des Montres.
In a previous article that detailed the different variants of the reference that were produced over its production lifespan, we mentioned that it is generally believed no more than 36,000 units were produced of the model. Since the publication of that article, we have continued to undertake research to try to establish a better estimate of production numbers, and also try to get a more accurate handle on the relative numbers produced across the three major dial configurations of Printed, Carved, and Raised logo dials.
One result of this research is that we are extremely confident that we have now also successfully decoded the logic behind the movement serial numbering, something that has - to the best of our knowledge - never been documented before.
Heads up - lots of words, and very few pictures, to follow!
Dating a Grand Seiko “First”
As discussed in the previous article, there are two different serial numbers that we can look to in order to attempt to establish when a particular example of a Grand Seiko “First” was produced.
Case serial number
Stamped on the inside of the case back of the watches can be found the case serial number. Comprised of seven characters, we explained in the previous article how the first two characters of the serial number can be used to determine the month and year of production. Excepting production in the months of November and December (where the year and month codes are reversed), the first character of the case serial number indicates the year of production, and the second the month (January through to October are identified by the numerals 1 through 0, November and December are identified by the letters N and D).
The third digit is always a “1”, and the final four digits are a counter.
For the purpose of attempting to establish production numbers of the “First”, one might think that this counter would be all that we need, but when looking to it to provide some insight, it actually raises more questions than it answers.
We discussed in the previous article how, from an examination of many examples of the reference, we had for the first time been able to establish some logic behind the counter part of the case serial number.
We also explained that it appeared that the case number counter was increasing at a faster rate than the movement counter number, thus questioning whether or not it could be used to provide a reliable estimate of production numbers.
Movement serial number
The movement serial number can be found engraved on one of the bridges of the movement.
It is a six digit number, and examples of Grand Seiko “Firsts” have long been known to have movement numbers starting with either a “6”, “1”, or “3”.
Previously it has been suggested elsewhere that the first two digits of the movement serial number can be used to provide an indication of the year the movement was manufactured. However, as we will prove for the very first time in this article, this is not the case.
It is clear that at least some of the digits of the movement serial number are a counter, but to-date we believe that no-one has ever managed to crack exactly how that counter functions over the production life-span of the reference.
The challenge of establishing a production number estimate
Without a clear and logical explanation behind the movement serial number counter, we are reliant on the case serial number counter to estimate production numbers.
There are a few key facts that can be concluded from the many examples of Grand Seiko “Firsts” that we have recorded both serial numbers from.
As it is just four digits, the counter section of the case serial number has to reset a number of times over the lifespan of the reference, since it is very obvious more than 10,000 examples were manufactured.
The “logic” behind resetting the counter clearly changed over the production timeline.
Even once we have determined the logic behind the resetting of the counter, if we use it to provide insight into production numbers, they would appear to be significantly higher than previously believed.
Establishing when the case serial number counter was reset
Having now documented significantly more examples of Grand Seiko “First’ case and movement serial numbers since publishing our earlier article, we believe we have gained some additional insight into reset points for the counter.
We believe that the production of the Grand Seiko “First” can be split into two distinct timeframes.
The initial production of the reference - stock manufactured for the launch of the watch - took place between April and June of 1960. For this initial three month production run, the counter simply ran sequentially. There was then a three month break, with production restarting in October 1960.
The earliest example of a Grand Seiko “First” that we have come across has a case serial number of 0411347 (a reminder that the counter is the last four digits), whilst the latest example from this initial production is numbered 0615020.
We have never seen an example of a Grand Seiko “First’ produced between July and September of 1960, with the earliest example from October being numbered 0011743.
From these examples, we believe the first time the counter was reset was for when production of the reference recommenced in October 1960.
Between October 1960 and December 1960, there are some anomalies when trying to understand the logic in the counter. We have documented seven examples produced in this period, with October production counters ranging from 1,743 through to 1,807; November’s from 6,116 through 6,648; and December’s single example numbered 5,638.
A couple things to draw out from these numbers. It is not realistic to infer that the counter was reset each month, as it is beyond the realms of possibility that over 6,000 “Firsts” were being manufactured in just one month. And given that is the case, if we assume the counter was simply running sequentially from 0001, how is it that we have an example from December seemingly produced earlier than examples from November?
Once we get into 1961, it would seem possible that the counter was reset at the end of every month, as we have examples throughout the year of the lowest counter numbers in one month being lower than the highest counter number of the previous month. This pattern continues through to the end of production in August 1963.
Why we can’t rely on case serial numbers to establish production numbers
As we can see from the above, there are a number of challenges if we try to use the case serial number counter to establish production numbers. The most significant of those challenges being that there are unexplainable anomalies in how the counter increments over time. Additionally, even with a large number of examples to extrapolate from, we are still left with multiple instances of just a single production example from a particular month.
Decoding the movement serial number
In the past, the movement serial number has not been used to try to establish production numbers for a couple of reasons.
Firstly, it is well known and documented that, from a chronological perspective, the use of movements in produced watches is all over the place.
It has long been believed (correctly, as it turns out), that movement serial numbers starting with a “6” were produced in 1960. And yet, we actually have examples of these earliest movements in watches with case serial numbers indicating production in 1960, 1961, 1962, and even in the last month of production in August 1963.
Additionally, as mentioned before, no-one has managed to understand the “logic” behind the leading digit of the movement serial number, and without this insight, it is impossible to get to the bottom of how many movements were produced.
We mentioned previously how we do not believe the case serial numbering was uniquely applied to the Grand Seiko “First”, as it would imply production of close to 3,000 units per month in some instances - a figure that seems much too high for a watch believed to have production not exceeding 36,000 units over its 41 month production run.
However, what we do know for a fact is that the 3180 movement in the Grand Seiko “First” was used exclusively for that reference. If there was some way to cracking the “code” behind the movement serial numbering, maybe it could give us some insight not just into how many “Firsts” were produced, but also an insight into the relative production of the different dial variants.
Use of movements over the production timeline
If we consider the first two digits of a movement serial number, and then look at the production year (as provided by the case serial number) of watches those movements are found in, we can gain some insight into why previously people have believed the first two digits identify the year of production of the movement.
For 1960 production watches, all movements are prefixed “60”.
For 1961 production watches, movements are prefixed “60”, “10”, and “11”.
For 1962 production watches, movements are prefixed “60”, “10”, “11”, and “12”.
For 1963 production watches, movements are prefixed “60”, “11”, “12”, “30”, and “31”.
It is understandable then that people in the past have believed that the production year of a movement can be interpreted as follows -
“60” prefix indicates 1960 production.
“10” and “11” prefixes indicates 1961 production.
“12” prefix indicates 1962 production.
“30” and “31” prefixes indicate 1963 production.
One thing that is clear is that a movement must have been manufactured prior to being installed in a case. This may seem to be stating the obvious, but it is a very important detail - if we look at the manufacturing month of the watch as indicated by the case serial number, then that is the absolute latest month that the movement could have been manufactured.
Another important insight into production that can be gleaned from the many examples of “Firsts” that we have examined in detail is that the dial was not installed on the movement at the time the movement was manufactured and given its serial number. We know this because of examining the case and movement serial numbers of watches with different dial types.
Thus, you will find examples of movement serial numbers with the prefix “60” having Printed, Carved, SD, and AD dials. But you will not find any genuine examples of, for example, a watch with a dial code indicating production in 1962 or 1963 with a printed or carved dial.
The detail behind the movement serial number
We have now recorded the serial number details of a sufficient quantity of Grand Seiko “Firsts” to be able to confidently share the logic behind the serial numbering. Once this logic has been established, we can extrapolate some key insights into relative production of the main dial types.
Whilst we are not prepared to reveal the full dataset, we will share some examples of movement serial numbers and the associated date of production of the watch that they were used in.
Firstly, let's take a look at the outlier examples - that is, the earliest and latest examples of each of the two-digit prefixed movement serial numbers, and the production dates of the watches they are found in.
Movement serial code prefix 60
The lowest numbered 60 prefix serial code we have is 600413, found in a watch dating from May 1960.
The highest numbered 60 prefix serial code we have is 607879, found in a watch dating from November 1961.
The earliest watch (based on the case serial number) with a 60 serial movement dates from April 1960.
The latest watch (based on the case serial number) with a 60 serial movement dates from August 1963.
Movement serial code prefix 10
Lowest code 108230, in watch from February 1961.
Highest code 109766, in watch from June 1961.
Earliest watch with 10 prefixed movement, February 1961
Latest watch with 10 prefixed movement, July 1961
Movement serial code prefix 11
Lowest code 110057, in watch from May 1963.
Highest code 119683, in watch from April 1962.
Earliest watch with 11 prefixed movement, June 1961.
Latest watch with 11 prefixed movement, August 1963.
Movement serial code prefix 12
Lowest code 122108, in watch from August 1963.
Highest code 129118, in watch from May 1963.
Earliest watch with 12 prefixed movement, September 1962.
Latest watch with 12 prefixed movement, August 1963.
Movement serial code prefix 30/31
Lowest code 300364, in watch from May 1963.
Highest code 311331, in watch from August 1963.
Earliest watch with 30 prefixed movement, April 1963.
Latest watch with 31 prefixed movement, August 1963.
Explanation behind the logic
Those of an analytical nature should be able to spot the obvious in the data provided above, so if you want to have a crack at working it out, take a break now and have a think.
For those who just want to jump to the answers, here we go…
The prefix is not two digits - it is just one.
The counter for the movement serial number was only ever reset once during the entire production timeline of the Grand Seiko “First”.
Firstly, lets just drop the first digit of the movement serial number and list the lowest and highest codes from each of the above sets of data -
00413
07879
08230
09766
10057
19683
22108
29118
00364
11331
Apart from the last two data points (the ones with the “3” prefix), all numbers are in numerical sequence. In our view this provides irrefutable proof as to the logic behind the movement serial numbering, and hence we can draw a number of conclusions from it.
The timeline of the numbering logic is as follows -
At the start of production, use a “6” prefix (probably the original intention was to indicate the year of production), followed by a 5 digit counter.
At the beginning of 1961, change the prefix to “1”, but let the counter continue to run.
Throughout 1961, 1962, and (almost certainly) into 1963, the counter continues to run, until the movement serial number hits 129999.
Rather than let the counter roll over to 130000, change the prefix to 3, and reset the counter to 00000.
Retain the 3 prefix and let the counter run until the end of production.
It is possible that the changing of the prefix to 3 started at the beginning of 1963, with the movement serial number not actually reaching 129999, but we think based on the evidence of watches we have the details of, that would be an unlikely coincidence.
The highest 1-prefix serial number we have recorded is 129118; the significant majority of watches in the 20,000 range of production are in watches from 1963; and the earliest example of a 3-prefix serial number we have come across is in a watch from April 1963.
Addendum November 1st 2019
A dealer in Japan has shared an example of a Grand Seiko First with a serial number 13055x. This is the first time we have come across an example of a serial number starting with 13, and would indicate that the first counter wasn't reset from 129999 to 300001, but instead carried into the 13XXXX. The existence of a serial number starting 13 probably does lend credence to a resetting of the counter and introduction of the 3 prefix at the start of 1963. The following text has been edited to reflect this discovery.
Production numbers
Given the above, it is evident to us that a minimum of 41,881 of caliber 3180, and ergo, Grand Seiko “Firsts”, were produced. We have a single counter that ran from the start of production right through into 1963 that reached at least 30550; the counter was reset to 0, and ran to at least 11,331.
Given we have an example of movement number 603862 in a watch from April 1960, and the lowest serial number we have is 600413, it is reasonable to assume that production of the 3180 movement commenced a few months prior to the first casing of a watch.
Assuming production commenced in January 1960, and that we are correct in our belief that production was halted over the summer of that year, we have a total of 41 months of active production, giving and average of around 1,000 units per month coming off the production line.
Relative production numbers of the major dial types
In previous discussions with collectors and dealers, it has been suggested to, and repeated by, us that the ratio of the major dial type logos was probably something of the order of 30:1 for Raised:Carved, and 50:1 for Carved:Print.
Whilst that may “feel” somewhat right for what we see these days in the market, it does need to be addressed now that we have much better insight into production numbers and timeline.
If we work on the basis of a total of c42,000 watches produced, the above ratio would imply the following actual unit production across the major dial types (some rounding here) -
Print - 30
Carved - 1,350
Raised - 40,500
From the examples we have documented, it is clear that both print and carved logo dialed variants of the Grand Seiko First were manufactured in the first production run from April to June 1960. Then, from October 1960 through to around June or July of 1961, we see carved logo dials. After a brief period around that same time where we find the transitional dials, raised logo dial production runs through to the end of August 1963.
Given the carved logo dial variant was in production for 12-13 months, it is clearly an error to believe that only 1,350 were produced. Therefore, we need to reexamine the estimate for the ratio of Raised:Carved dials.
We know that we can set a lower-bound on movement production up to a particular month by looking at the counter for the highest numbered movement put in a watch for that month. For June 1961, the highest numbered movement we have an example of is 113096. Assuming the break in the summer of 1960, and that movement production commenced in January 1960, this would give an average of around 870 movements produced per month.
We can also glean from the data that we have, that there is probably an average lead time of around 3 months from when a movement is manufactured, and when it is put in a watch.
Based on the data and some reasonable assumptions, we therefore believe that the number of finished watches completed by June 1961 to be of the order of 9-10,000.
The overwhelming majority of these would be carved dials. Print dials as mentioned only appear in the initial production run of 3 months, and it certainly appears that within those three months very few were produced.
There are simply too few examples of Print logo dialed variants that we are aware of to be able to draw a totally objective conclusion as to how many were produced. However, now that we know there were significantly more carved logo watches made than earlier suspected, the number of print dialed variants does need to be revised upwards.
Based on our research presented here and observations of the market over the last 5 years, we now believe a more realistic breakdown of relative production numbers of the three major dial variants to be as follows -
Print - 240
Carved - 9,500
Raised - 32,000
One other way to look at the proposed number of print logo dialed watches is that if - as has been suggested - print dials were used to make up for production shortfall of carved logo dial watches due to yield issues, then the figures provided would imply a yield of around 90-95% of carved logo dialed watches in the initial 3 month production.
It must be stressed that these are estimates of production numbers, and do not reflect relative scarcity of watches that we see becoming available on the market.