Can a watch's movement, or timing, slow over a period of time?

[chamen]

I'm hoping that no one throws rotten vegetables at me, or this post...

I recently purchased a new Rolex SubC. Taking it out of the box, I synchronized it to the clock on my computer (which in turn is synchronized with a network time protocol server). I wore the watch daily for the first 2 weeks and gradually saw a +2/day drift (I think the math came to +2.25ish). Very acceptable according to COSC, correct?

This is the point where I start to question my logic, sanity, or perhaps just my general knowledge of watch mechanics. Near the end of the 3rd week, I noticed that the drift seemed to actually be closing. by the 21st day, the drift was around +1.5.

So, I took a second look at the clock on my computer and I compared to this nifty little widget from NIST:

http://www.time.gov/widget.html

Sure enough, they were in sync. So, I decided to reset my watch, on June 1st. Here I am a week in to the month and the SubC has not drifted a second.

I've read posts where people have claimed that a watch runs "hot" at the start of its life after being dormant for an extended period of time. Is there any truth to this? And if so, what causes it? I feel that my daily routine didn't change much over the course of any of the 4 weeks I've worn the watch, so I don't believe that it was getting a zealous winding at the beginning and I've since slowed down.

Can someone provide some insight and rationalize what I'm seeing?

Thank you!

[Rikki]

You're seeing the proper functioning of a highly precise mechanical device. I could explain the nuances of positional errors and how some watches run better on some people and not others but sounds like you have a great watch and enjoy it. Rikki