An Incomplete Record of 62MAS Serial Numbers and Casebacks
Jun 13, 2018 5:39:16 GMT -8
jim1963, cobrajet25, and 6 more like this
Post by rossr on Jun 13, 2018 5:39:16 GMT -8
Hello All.
I'm currently putting together an update of the price trend analysis (last done here in mid-2017) for the Seiko 62MAS (6217-800x) watch. As well as noting the prices achieved, I've also captured some other features of the watch - the condition, the authenticity, the caseback-type, and the serial number. As the price-condition trend work is a bit time-consuming, I thought that I'd get some other results out first. I thought it would be quick and simple. Not exactly.
As before, the data has been sourced from EBay, Yahoo Japan, the Watch Site, the old Network 54 SCWF, Wrist Sushi, WUS and various other forums and sites. Though the data dates back to 2003, it is not a complete record of 62MAS watches that have come up for sale, or otherwise presented. I’ve cut off the data input at 12th of June 2018.
I've not seen this information captured and presented in quite this way before. Perhaps it is of some interest.
Data cut by Month
The first chart notes the serial numbers I've been able to record from the early 6217-8000 model manufactured in April and May 1965.
Chart 1
Note that the serial number was not re-set when moving from April to May. This is not a bad record, though there are a couple of gaps - notably in the 2000s and low 3000s, and then again in the 8000s and 9000s. What can be deduced from this record? There is no serial number exceeding 12234. We could surmise that there were some watches made after this, but given that the relatively even distribution of serial numbers to this point, I would expect that this number would be fairly small. My guess? There were less than 13000 units of the small crown 6217-8000 ever released for sale. I have seen no record of any 62MAS watch with a serial number after May 1965 and before January 1966.
The second chart records the watches manufactured from January to April 1966.
Chart 2
Note that the serial number is again not re-set between January and March. The April watches show a big jump in serial numbers. Again, it seems that there were not a lot of watches being made - perhaps less than 15000 in total between January and March. The April '66 watches have very closely spaced serial numbers. I'd guess not many more than 1000 watches were made in April.
Now we're into mid-1966. Here's where it starts to get interesting.
The record looks good for the first part - an even record of serial numbers from 0 to 3000, a large gap to 11000, and then an even record to 14000. There are two very conspicuous outliers at 28000 and 45000, and nothing connecting them to the earlier batches. Though there are not many point in the record, I'd venture that there may have been 5000-8000 watches made in the first batch, but very few around the outliers.
June 1966 and September 1966, display a similar characteristic. I have no serial numbers recorded for July, October and December '66 and one lonely record for each of August '66 and November '66. I have no record from January 1967. A picture emerges.
It has been observed by others that the first two digits of the Seiko serial number - that denoting the year and month of manufacture - is stamped separately to, and most likely after, the second part of the serial number - in the case of the 62MAS, the last 5 digits.
If the May to September data is combined, we get a characteristic as shown in Chart 3, “May-June-August-September 1966”.
Chart 3
The pattern suggests that the same batch of casebacks were used over this period, save for a separate batch released in June. The outliers are shown. Special batch replacements? The August solo is a bit odd.
Why didn’t I include the previously mentioned lone November model in this consolidation? The May to September batch are all Dolphin backs. The November model is a Horseshoe - the earliest known piece to have this caseback feature.
I have no records of pieces with serial numbers from December 1966 or January 1967.
Chart 4 shows a small group of pieces with a February 1967 serial, and each has a serial number below 2000. I was tempted to consolidate them with the 1966 group, but the low serial numbers don’t really fit the pattern. They are likely part of the larger 1967 batch.
Chart 4
There is a good record of pieces with a March 1967 serial number – all with serials below 10000. Refer to Chart 5. This is the first regular appearance of the Horseshoe back.
Chart 5
Chart 6 – April 1967 - is particularly interesting. We see quite a few Horseshoe backs from Suwa. The famed Daini 62MAS models make an appearance. Note that sweet-spot of Daini cases in the centre of the monthly record. Also note that Suwa Horseshoe in the centre of all those Daini pieces. I checked the source, and this is correct – the Horseshoe Suwa was sold by Saitama – the photos were very clear.
Chart 6
Chart 7 – May 1967 – shows a batch of Horseshoe casebacks, the first two of which are Daini. The second batch with the higher serial numbers are Dolphin-backs. I need to confirm and adjust the chart to show this. There are a couple of odd outliers.
Chart 7
Chart 8 – June 1967 – has a very similar pattern to that shown in May, Horseshoes with low serials and Dolphins with higher serials. Of interest is the only Daini I have a record of with a June serial (owned by a member here).
Chart 8
Data cut by Caseback-type
Chart 9 is a consolidation of the Horseshoe casebacks I have recorded, and could check and confirm. I may have encountered other Horseshoe casebacks, though maybe not a large number. When I first collated the data, I was really chasing prices, sale dates and serial numbers. In many cases, photos were not available, and the type of caseback was not noted. Much of the source data is no longer available – either through expiry of available Ebay/YJ data, the demise of the old Network54 SCWF as a navigatable and searchable reference, and the issues with photos that had been hosted on Photobucket.
The data is split by month and source – Suwa or Daini. I can’t really see a pattern with respect to the distribution of Suwa and Daini serial numbers. There is a conspicuous outlier amongst the Horseshoes – Serial 10158 from April 1967. I have not shown it on the charts.
Chart 9
Chart 10 is a consolidation of all confirmed Dolphin-backs and unclassified casebacks (which were most likely to be Dolphins). This shows a reasonably continuous presentation of serial numbers – with a couple of unusual gaps. I have left out two outliers – 49967 and 84345 which do not fit the characteristic. I have included the strange October 1967 outlier (7026558), as the serial fits the production pattern.
Chart 10
Estimation of Total Production Number
By identifying the production runs, an approximation can be made as to the total number manufactured. The limits of each run are not known, but given the even distribution of serial numbers that appear in the record, it would seem that the number of outliers is very small, and unlikely to significantly influence the production total;
Table 1 - Estimation of Production Numbers
The exercise was a bit more involved than I originally thought, but I think that it brought out a few things that I certainly had not noticed previously. There are a couple of things that I need to chase up. If I can, I’ll try to convert some of the "Unconfirmed" casebacks into something confirmed – most likely these will be Dolphins.
Feel free to comment and make your own interpretation of the data. Serial numbers welcomed!
Regards,
Ross
I'm currently putting together an update of the price trend analysis (last done here in mid-2017) for the Seiko 62MAS (6217-800x) watch. As well as noting the prices achieved, I've also captured some other features of the watch - the condition, the authenticity, the caseback-type, and the serial number. As the price-condition trend work is a bit time-consuming, I thought that I'd get some other results out first. I thought it would be quick and simple. Not exactly.
As before, the data has been sourced from EBay, Yahoo Japan, the Watch Site, the old Network 54 SCWF, Wrist Sushi, WUS and various other forums and sites. Though the data dates back to 2003, it is not a complete record of 62MAS watches that have come up for sale, or otherwise presented. I’ve cut off the data input at 12th of June 2018.
I've not seen this information captured and presented in quite this way before. Perhaps it is of some interest.
Data cut by Month
The first chart notes the serial numbers I've been able to record from the early 6217-8000 model manufactured in April and May 1965.
Chart 1
Note that the serial number was not re-set when moving from April to May. This is not a bad record, though there are a couple of gaps - notably in the 2000s and low 3000s, and then again in the 8000s and 9000s. What can be deduced from this record? There is no serial number exceeding 12234. We could surmise that there were some watches made after this, but given that the relatively even distribution of serial numbers to this point, I would expect that this number would be fairly small. My guess? There were less than 13000 units of the small crown 6217-8000 ever released for sale. I have seen no record of any 62MAS watch with a serial number after May 1965 and before January 1966.
The second chart records the watches manufactured from January to April 1966.
Chart 2
Note that the serial number is again not re-set between January and March. The April watches show a big jump in serial numbers. Again, it seems that there were not a lot of watches being made - perhaps less than 15000 in total between January and March. The April '66 watches have very closely spaced serial numbers. I'd guess not many more than 1000 watches were made in April.
Now we're into mid-1966. Here's where it starts to get interesting.
The record looks good for the first part - an even record of serial numbers from 0 to 3000, a large gap to 11000, and then an even record to 14000. There are two very conspicuous outliers at 28000 and 45000, and nothing connecting them to the earlier batches. Though there are not many point in the record, I'd venture that there may have been 5000-8000 watches made in the first batch, but very few around the outliers.
June 1966 and September 1966, display a similar characteristic. I have no serial numbers recorded for July, October and December '66 and one lonely record for each of August '66 and November '66. I have no record from January 1967. A picture emerges.
It has been observed by others that the first two digits of the Seiko serial number - that denoting the year and month of manufacture - is stamped separately to, and most likely after, the second part of the serial number - in the case of the 62MAS, the last 5 digits.
If the May to September data is combined, we get a characteristic as shown in Chart 3, “May-June-August-September 1966”.
Chart 3
The pattern suggests that the same batch of casebacks were used over this period, save for a separate batch released in June. The outliers are shown. Special batch replacements? The August solo is a bit odd.
Why didn’t I include the previously mentioned lone November model in this consolidation? The May to September batch are all Dolphin backs. The November model is a Horseshoe - the earliest known piece to have this caseback feature.
I have no records of pieces with serial numbers from December 1966 or January 1967.
Chart 4 shows a small group of pieces with a February 1967 serial, and each has a serial number below 2000. I was tempted to consolidate them with the 1966 group, but the low serial numbers don’t really fit the pattern. They are likely part of the larger 1967 batch.
Chart 4
There is a good record of pieces with a March 1967 serial number – all with serials below 10000. Refer to Chart 5. This is the first regular appearance of the Horseshoe back.
Chart 5
Chart 6 – April 1967 - is particularly interesting. We see quite a few Horseshoe backs from Suwa. The famed Daini 62MAS models make an appearance. Note that sweet-spot of Daini cases in the centre of the monthly record. Also note that Suwa Horseshoe in the centre of all those Daini pieces. I checked the source, and this is correct – the Horseshoe Suwa was sold by Saitama – the photos were very clear.
Chart 6
Chart 7 – May 1967 – shows a batch of Horseshoe casebacks, the first two of which are Daini. The second batch with the higher serial numbers are Dolphin-backs. I need to confirm and adjust the chart to show this. There are a couple of odd outliers.
Chart 7
Chart 8 – June 1967 – has a very similar pattern to that shown in May, Horseshoes with low serials and Dolphins with higher serials. Of interest is the only Daini I have a record of with a June serial (owned by a member here).
Chart 8
Data cut by Caseback-type
Chart 9 is a consolidation of the Horseshoe casebacks I have recorded, and could check and confirm. I may have encountered other Horseshoe casebacks, though maybe not a large number. When I first collated the data, I was really chasing prices, sale dates and serial numbers. In many cases, photos were not available, and the type of caseback was not noted. Much of the source data is no longer available – either through expiry of available Ebay/YJ data, the demise of the old Network54 SCWF as a navigatable and searchable reference, and the issues with photos that had been hosted on Photobucket.
The data is split by month and source – Suwa or Daini. I can’t really see a pattern with respect to the distribution of Suwa and Daini serial numbers. There is a conspicuous outlier amongst the Horseshoes – Serial 10158 from April 1967. I have not shown it on the charts.
Chart 9
Chart 10 is a consolidation of all confirmed Dolphin-backs and unclassified casebacks (which were most likely to be Dolphins). This shows a reasonably continuous presentation of serial numbers – with a couple of unusual gaps. I have left out two outliers – 49967 and 84345 which do not fit the characteristic. I have included the strange October 1967 outlier (7026558), as the serial fits the production pattern.
Chart 10
Estimation of Total Production Number
By identifying the production runs, an approximation can be made as to the total number manufactured. The limits of each run are not known, but given the even distribution of serial numbers that appear in the record, it would seem that the number of outliers is very small, and unlikely to significantly influence the production total;
Table 1 - Estimation of Production Numbers
The exercise was a bit more involved than I originally thought, but I think that it brought out a few things that I certainly had not noticed previously. There are a couple of things that I need to chase up. If I can, I’ll try to convert some of the "Unconfirmed" casebacks into something confirmed – most likely these will be Dolphins.
Feel free to comment and make your own interpretation of the data. Serial numbers welcomed!
Regards,
Ross