NavList:

This is another update on my experiment using three cheap, $17.00,
watches as a chronometer.

I obtained another set of data today, October 28, 2020 Z at
0602 Z, 4058 days since the start, more than ELEVEN YEARS
See attached photo taken at 06:02:28Z,

watch "A" is 16 minutes and 42 seconds fast, etc. (The seconds
are in the top windows.)
I have not had to replace any of the batteries in ELEVEN YEARS.

At the beginning of this experiment I determined the daily rate of
each watch from a 99 day sample. Watch "A" has a rate
of 0.1919 seconds fast per day; watch "B", 0.3737 fast; and watch
"C" is 0.6263 fast. Using these values I calculated what the
predicted corrections would be, the values that would be
used by the navigator to correct the chronometer time,
 and compared with the measured values.
The differences show how much of a navigational error would have
 resulted from the navigator using the predicted times.

In the format for A, B, and C in seconds: actual error;
predicted error; difference.

A = 1002/825/177 fast: B = 1483/1539/56 slow: C = 2569/2601/32 slow

Averaging these differences equals 30 seconds fast.

Using only one watch with the largest error, 177 seconds, would
resulted in a 44 minute of longitude error. Using the average of
all three watches would have produced an error of only 7.4 minutes
of longitude, not bad for fifty-one dollars worth of watches
after more than ELEVEN YEARS.
============================================
Prior update:

This is another update on my experiment using three cheap, $17.00,
watches as a chronometer.

I obtained another set of data today, May 31, 2013 Z at
0108 Z, 1351 days since the start, more than three years
and eight months. See attached photo taken at 01:08:00Z,
(WWV radio signal)
watch "A" is 5 minutes and 6 seconds fast, etc. (The seconds
are in the top windows.)
I have not had to replace any of the batteries.

At the beginning of this experiment I determined the daily rate of
each watch from a 99 day sample. Watch "A" has a rate 
of 0.1919 seconds fast per day; watch "B", 0.3737 fast; and watch
"C" is 0.6263 fast. Using these values I calculated what the 
predicted corrections would be, the values that would be
used by the navigator to
 correct the chronometer time,
 and compared with the measured values.
The differences show how much of a navigational error would have
 resulted from the navigator using the predicted times.

In the format for A, B, and C in seconds: actual error;
predicted error; difference.

A = 306/266/+40 fast: B = 528/536/ -8 slow: C = 906/846/-20 slow

Averaging these differences equals 4 seconds fast.

Using only one watch with the largest error, 40 seconds, would
resulted in a 10 minute of longitude error. Using the average of
all three watches would have produced an error of only one minute
of longitude, not bad for fifty-one dollars worth of watches
after more than three and a half years.

See my prior reports at:

https://NavList.net/m2.aspx/Watches-chronometers-LaPook-oct-2011-g17175