Guaranteed accuracy on the wrist

[Der Amf]

I was wondering about COSC specs, how it's the movement that is tested, and also about how autos behave when worn for only a day, etc. Etc etc. All getting murky and muddled in my head.

Let us say I bought a watch with COSC certified movement.

Each week I would put it on on Monday morning and wear it all day everyday until Friday evening.

Where would be the threshold of inaccuracy where I could say to the manufacturer, "excuse me, but...."?

I know that some people get lucky, and find that over the week the watch gains just 10 seconds or something. I'm wondering how unlucky you have to be before you get to ask for the dice to be rerolled.

I imagine answers will want to make reference to specific manufacturers

[happyscrappyheropup]

If I bought a new watch that was marketed as a COSC chronometer and its timekeeping drifted outside COSC specs during the warranty period I would be tempted to have the manufacturer service it.

However, like cars, I don't buy new watches.

[geoffbot]

Cosc mechanical (quartz is tighter) is -4/+6 average daily rate, so after 5 days +30 seconds would be in spec I believe? Though there are other stats like average variation between daily rate. I guess that most manufactureres would regulate a COSC for free within warranty if it doesn't perform within this. Anything more would be a courtesy I think?

[Der Amf]

So are we saying that although COSC specs don't refer to the performance on the wrist, it would be reasonable to use those figures for opening a conversation with the manufacturer?

(For what it's worth, my basic 2824 would be +45 by the end of Friday; were I to pay considerably more for a certified movement, I would want a bigger leap in performance) (maybe this just tells me how good a standard 2824 is)

[CFR]

Very hard to tell.

We tend to over simplify and accept that COSC means a watch should perform within the -4s/+6s (mean daily rate) interval; that's not what COSC certificates. You may want to google cosc to find out all the procedures involved.

Generally speaking a wristwatch should perform at its best in two main positions: dial up and crown down (assuming you are wearing the watch on your left arm and it rests dial up during the night) because these are the positions it will face most of the time. Isochronism, temperature change, etc. are not so easy to atest at home. Manufacturers are usually vague, stating something like -1/+11s or anything like that, but - again, generally speaking - they are not very explicit about the conditions their limits are achieved.

So, unless the watch is running pretty bad or way out the standards the manufacturer finds acceptable, it's hard to make a solid case...

[CFR]

So are we saying that although COSC specs don't refer to the performance on the wrist, it would be reasonable to use those figures for opening a conversation with the manufacturer?

Yes, if you find a comprehensible manufacturer. I'm not being ironic here, it's possible; I've already did it...with success!

[Der Amf]

Since C Ward a reputation for being fair, I had a look at their website. Even in the manual for the new Trident containing their own in-house COSC movement avoids making any promises. However the website does contain a glossary, and the entry for COSC says:

Abbreviation of "Contrôle Officiel Suisse des Chronomètres" the most important Swiss institution responsible for the functioning and precision tests of movements of chronometers. Tests are performed on each individual watch at different temperatures and in different positions before a functioning bulletin and a chronometer certificate are issued, for which a maximum gap of -4 to +4 seconds per day is tolerated. With fewer than 3% of all watches receiving this level of certification, it is easy to understand why chronometers are so highly valued by watch aficionados and collectors.

So from this, I would feel happy opening a conversation with C Ward if one of their COSC watches was performing, when brand new, worse than +20 or -20 seconds over the four and a half days of consecutive wear (and that's better than twice as good as my basic 2824)

[CFR]

Since C Ward a reputation for being fair, I had a look at their website. Even in the manual for the new Trident containing their own in-house COSC movement avoids making any promises. However the website does contain a glossary, and the entry for COSC says:



So from this, I would feel happy opening a conversation with C Ward if one of their COSC watches was performing, when brand new, worse than +20 or -20 seconds over the four and a half days of consecutive wear (and that's better than twice as good as my basic 2824)

Yes, I believe you could.

[geoffbot]

How much more would you pay for a watch with a cosc 2824 vs a top 2824? No markings on the dial or other differences.

[Domo]

Yeah it's pretty much up to the good will of whoever you buy it from or the manufacturer I'd say. A COSC itself is not a guarantee of actual performance, only a stringent set of criteria for an uncased movement. Manufacturers may or may not add their own tighter allowances on the actual watch itself, for bare COSC up to 10 seconds a day maximum drift is technically allowed.