COSC and Accuracy

[Roggie Rog]

I own a 16710 GMT Master and would like to know more about what constitutes acceptable COSC accuracy, and how to understand and interpret the -4/+6 spec.

Case in point: I set my 16710 to the time on my Macbook Pro, which pings Apple's server for correct time, three weeks ago. Today my watch is around 3 minutes fast. It was a minute fast after one week and 2 minutes fast at the 2-week mark.

Is this considered out of spec? Thanks in advance for your help.

[SearChart]

It's just a little faster than cosc rating, still fine though IMO.

[Tools]

Originally Posted by Roggie Rog

I own a 16710 GMT Master and would like to know more about what constitutes acceptable COSC accuracy, and how to understand and interpret the -4/+6 spec.

Case in point: I set my 16710 to the time on my Macbook Pro, which pings Apple's server for correct time, three weeks ago. Today my watch is around 3 minutes fast. It was a minute fast after one week and 2 minutes fast at the 2-week mark.

Is this considered out of spec? Thanks in advance for your help.

COSC is a certification agency. They use a variance of a 10 second swing from a known baseline (-4/+6) during the 15 day test of a movement. (as well as other criteria) If a watch comes to them with a 3 second variance, the -4/+6 is based on that 3 second variance as a baseline. If it passes the test it is certified as a Chronometer and is never subjected to the test again.

It has nothing to do with how your watch performs on your wrist, only it's capabilities, and has been wrongly interpreted by enthusiasts for years.

Your Mac is not a good source since it will correct to any variance the server adjusts to.

Check your variance over several days using a known standard source then divide by the number of days. If the variance is not to your satisfaction, have it regulated. I would never let my newer watches get to the 5 seconds off mark, they are capable of better than that.

If I had your watch I would have it regulated. 60 seconds over 7 days = 8.6 seconds per day variance.

[Rock]

RR - all of us have different standards for the amount of variance we consider acceptable. Some folks don't worry about accuracy much and others are obsessive.
I monitor mine daily (I am retired) and the average daily variation is about +2 -3 secs but varies slightly from month to month.
If the average daily variation taken over a month was more than 5 it would be off to the RSC for adjustment - for the same reason that Tools mentions.

[Roggie Rog]

Thanks for your input guys. Since my watch is in year 2 of a 3-year warranty, I sent it in for adjustment/regulation. I'm waiting to get it back.

[skprd13]

My TT Black K series no service history over the last 70 days I have been monitoring her and she is consistent at +6 with a variance of +4.1. I am happy with her performance however plan on sending her in to Vanesa in the near future.

[BrianMay]

Having just traded a Quartz Breitling for a second-hand Datejust, I was VERY interested in accuracy. My ethos was I didn't really need accuracy of that degree since I gave up flying for a living.

BUT . . . my 2001 Datejust 16203 has lost 10 seconds in three weeks. Reckon I'm quite happy with that. My Rolex qualified watchmaker said 'Don't touch it' !!!

I'll take note and feel very grateful.

[cop414]

Originally Posted by BrianMay

Having just traded a Quartz Breitling for a second-hand Datejust, I was VERY interested in accuracy. My ethos was I didn't really need accuracy of that degree since I gave up flying for a living.

BUT . . . my 2001 Datejust 16203 has lost 10 seconds in three weeks. Reckon I'm quite happy with that. My Rolex qualified watchmaker said 'Don't touch it' !!!

I'll take note and feel very grateful.

That is fabulous accuracy Brian!
My Submariner will loose about 16 seconds over three weeks and I'm delighted with that. Would I like it to be +.75 sec per day rather than -.75 sec per day, sure but that's not something that I'm going to mess with.

[gontomt]

My new (purchased November, 2014) ceramic No Date Sub was gaining 4-5 seconds a day. I realized that this was within COSC specs, and, as such, not a warranty issue. Having had some much better accuracy in two other Rolex watches, I was disappointed. I thought I didn't like the watch, and wore it less. even thought of flipping it I live the majority of the time (soon to be ALL of the time) in a 100-year-old cabin at 6,400 ft elevation - no internet, rare cell signal. I needed to "survive" without an external source of exact time. I kept trying the Sub, and checked the time whenever in range of a reliable Internet signal (NIST). It was getting worse. Finally, when I realized that the power reserve was only 13 hours, and arguably a warranty issue, I got my AD to send it to RSC Dallas. Five weeks, but power reserve is 48 hrs (Spec), and it's about 1/2 second off after 10 days. Actually, it was off 1/2 second after the first day, and hasn't varied since. I wear it 24/7 except when I shower or do dishes - not a moisture thing, just a soap scum thing. I couldn't be happier, save for that this is my new standard. Regulation is everything - I would have been happy to wait for it and pay for regulation, but it was covered under warranty.