In a discussion in the UG on Ebay thread the following was said by @bgrisso and me:

For me one of the more intruiging things about UG is the above and the fact that there hardly any hard numbers about the production of certain references. As some of you know I collected a database with references, serials and movement numbers (and only if these 3 are known). Based on this info it may be possible (or not) ot draw some conclusions. For this thread I used reference 22279 because this is - I think - a reference with a relatively high production and a relatively high number of survivors.

Based on my research this reference 22279 was produced in a couple of batches: (numbers are serials)
- 972k: most likely very small batch. Only 1 example in my database
- 1.067k: most likely very small batch. Only 1 example in my database
- 1.155/1.156: see below for the examples found but batch may be 1.250-1.500 pieces

Later batches exist but they don't have movement serials.

One of the things debated is the configuration in which the watches exited the factory. As stated above I think that they left the factory in a 'standard' configuration which could be changed at wish by the customer. Hereby the 18 examples I found: (serial, movement number - if known, if not known '?', if unnumbered 'none', link)
1) 1.155.xxx none https://www.bernardinimilano.com/products/universal-geneve-tri-compax-17
2) 1.155.160 ? https://omegaforums.net/threads/tri-compax.46419/#post-553358
3) 1.155.403 ? https://www.chrono24.com.au/all/miscellaneous-tri-compax--id12057800.htm
4) 1.155.444 223.925 https://catalog.antiquorum.swiss/en/lots/lot-81-74?browse_all=1&page=5&q=universal
5) 1.155.608 ? http://www.2diveornot2dive.com/2019/07/01/universal-geneve-tri-compax/
6) 1.155.609 none https://www.shucktheoyster.com/universal-geneve-tri-compax-calendar-chronograph-22219/
7) 1.155.791 ? https://www.sothebys.com/en/auctions/ecatalogue/2013/important-watches-ge1304/lot.236.html
8) 1.155.793 none https://awadwatches.com/product/194...-tri-compax-ref-22279-cal-481-untouched-dial/
9) 1.155.863 252.677 https://www.chrono24.nl/universalgenve/1940s-tri-compax-ss-moonphase-chronograph--id11426433.htm
10) 1.155.871 252.739 https://omegaforums.net/threads/universal-geneve-tri-compax.82469/
11) 1.156.078 263.922 https://www.phillips.com/detail/universal/CH080121/195
12) 1.156.160 none https://www.catawiki.com/l/41664657...chronograph-tri-compax-22279-unisex-1970-1979
13) 1.156.171 ? https://www.cambiaste.com/IT/asta-0368/universal-geneve-tricompax-no-1156171-ref-.asp
14) 1.156.206 ? https://catalog.antiquorum.swiss/en/lots/lot-54-160?browse_all=1&page=1&q=22279
15) 1.156.257 ? https://www.chrono24.nl/universalge...ETLANG=nl_NL&SETCURR=EUR&goal_change_domain=1
16) 1.156.279 ? https://catalog.antiquorum.swiss/en/lots/lot-266-147?browse_all=1&page=1&q=22279
17) 1.156.301 235.462 https://www.liveauctioneers.com/ite...geneve-tri-compax-moonphase-chronograph-watch
18) 1.156.376 ? https://catalog.antiquorum.swiss/en/lots/universal-ref-22279-lot-306-28?browse_all=1&page=1&q=22279

Tank number: 1.367. This is an estmation of the total number produced. Formula: https://www.wolframalpha.com/input/?i=(6376+-+5160)+*+(1+++(2+/+(17+-+1)))+-1 (see https://en.wikipedia.org/wiki/German_tank_problem for more info)

 

Amazing research! thanks for sharing.

 

So what do we have?
Dial:
- 2-4-8-10 & dots (printed): 11x
- 2-4-8-10 & dots (imposed): 1x
- 1-2-4-5-7-8-10-11 (printed): 3x
- 1-2-4-5-7-8-10-11 (imposed): 1x
- 2-4-8-10 & triangles: 1x
- Triangles: 1x

Hands:
- Sword: 10x
- Pencil (steel): 2x
- Pencil (blue/black): 2x
- Black (steel): 1x (=5.793)
- Military: 1x (=6.171)
- Gold (later UG(?)): 1x
- Gold (UG(?)): 1x (=6.257)

 

Well done!

 

Anyone having more examples please share and I'll update the table.

 

Sorry Mark, but I don't get the point. You state that there's a relation (of course speaking about the same reference) between serial number and dial/overall configuration, but for these examples we have (thank you for your work!) we see that approximately 30% of pieces have a different configuration.
Moreover, almost every pieces of the 11 with the principal configuration have the so-called "forniture dial" without dash between the words "Tri" and "Compax".
I mean, to my eyes, this is exactly the demonstration of the fact that in UG there are no specific dials for specific serials. There could be some dials more common than others, but no rules, in the end.

Happy to read a reply from you, and thank you again for your work!

 

Thanks for pulling this information together. I'm also struggling to draw any conclusions, except maybe what we already presumed; chaos

As @vujen already stated, the fact that a huge majority of the above examples have what we've been calling "service" dials complicates these examples.

In terms of serials, I'm not surprised to see some smaller runs earlier, with a big run later. Tri Compax was supposedly introduced at baselworld 1944 to celebrate the 50th anniversary of UG, and was a big hit, so it's not hard to imagine some small runs pre-introduction, and then some large runs post baselworld. It seems likely to see this same pattern across a broad range of case references.

I don't know what conclusions to draw, if any, about movement numbers. Are you simply grabbing this information if it exists, because it may be useful down the road, or do you have any specific theories in mind regarding movement numbers?

In any event, I commend your dedication on this research. Gathering info is often the first step to getting a better understanding of things.

I think that @LouS was also compiling spreadsheets of case reference and serials, across all models. I haven't seen him active here in a long time, not sure if he might be open to pooling data with you at some point.

 

I state nothing. I (still) think that the watches will have come in a certain configuration from the factory. If not, they would have been more or less kit watches which would have been assembled at the dealer. If that would have been the case why would UG then have bothered to produce so much different references.

Having said that: you certainly have a point about the dials. The elephant in the room for me is something else: the movements. Why do some of them have unnumbered movements?

 

Well it started because I acquired a 224102 and wanted to find other examples of that reference. It ended up in the spreadsheet I now have

My first thought was that it may be possible to say more about the production numbers or production numbers in certain years but that was (obviously) not correct. One thing I found out that the difference between serial and movement number is around 600k with serials between 700-800k and that this difference gradually becomes higher and ends up well over 1 mio at serials 1.3-1.4m. What does that tell us? I don't know but at least that the proportion of watches with unnumbered movements became higher.

Later on I became aware of the batch effect. I also compiled info on UG sold as other brand and there is very clear. For instance: all Berthoud ref 6309's have a serial around 69x.xxx and the batch size is 4-5k I think.

 

And some more:
19) 1.155.7x3 ? https://orologi.forumfree.it/?t=75401097
20) 1.155.757 ? https://www.instagram.com/p/CARr8myHQa6/
21) 1.156.xxx https://watchnet.co.jp/en/pickup/past-sale-item/48369/
22) 1.156.008 https://www.watchprosite.com/horolo...-tri-compax-some-pictures-/17.963293.6699433/
23) 1.156.126 ? https://onlineonly.christies.com/s/...tage-tri-compax-100/77722?ldp_breadcrumb=back
24) 1.156.204 ? https://www.liveauctioneers.com/item/49074324_universal-geneve-compax-chronograph-wrist-watch