How Would Your Watch Perform in The Observatory Trials?
Part 1: Scoring your watch in the Concours International de Chronométrie
Now demonstrated over a number of years, chronometry is back. Even COSC updated itself. As a precision freak I’ve always wondered how our modern watches would compare to Observatory Trial competitors both historical (1940s to 1960s) and recent. The point of this piece is to start to calibrate modern watch ability against recognised competitive frameworks.
The Concours International de Chronométrie (CIC), revived in 2009 by the Musée d’Horlogerie du Locle after a gap of more than three decades from the historic Observatory Trials, ran as a biennial competition through four editions before falling effectively silent after 2015. (After a gap in 2017, the 2019 format sadly collapsed because only one watch survived the first eliminatory round.)
Rather than revive the historic observatory-trial protocols directly, the CIC was built around the familiar ISO 3159 chronometer test framework, with an additional post-stress remeasurement phase intended to test stability after magnetism and shocks.
The test structure remained consistent across all editions:
ISO 3159 series of measurements at the Besançon Observatory (N1),
ISO 3159 series of COSC measurements at Bienne (N2)
Then to Le Locle where the watches were subjected to shock and magnetism tests carried out at the Haute Ecole Arc. The tests for resistance to magnetic fields are carried out in accordance with the ISO 764 standard (undisclosed field, but I assume 4800 A/m), while those for shock resistance are the subject of the school’s own special tests, which are carried out using a modified robot that exposes the watch head to 150 shocks at an intensity of 150g as follows: 50 shocks on the 3 o’clock to 9 o’clock axis, 50 on the 6 o’clock to 12 o’clock axis and 50 shocks perpendicular to the dial (25 in each direction). This is really the greatest difference between the historical observatory trials and these trials – adding shocks (at a pretty low level of shock of 150g considering production movements can survive 3000g plus and magnetism of only 4800 A/m (60 Gauss)).
and a final ISO 3159 re-measurement at COSC Bienne (N3)
with the final score (N) calculated as an equal one-third weighting of all three series, out of a theoretical maximum of 1,000 points. I think the original plan was for the results to be weighted 40% for the first two COSC sessions and the third to be weighted 20% in the results (to offset for the stress testing if it had a large effect). In the end, the simple average ‘N’ was used.
Critically, watches were not returned to their makers at any point during testing; they either survived the protocol on their own terms or failed.
How was N calculated?
In practice, CIC scoring was built from the same underlying timing parameters familiar from ISO 3159 chronometer testing. That is what makes the comparison interesting: if you have serious COSC data, you can begin to estimate how a watch might have fared under CIC rules.
To remind you of those:
M or Mmoy: Average rate across 5 different positions.
V or Vmoy: How much the rate changes in those 5 positions (average)
Vmax: The biggest change in rate in one of those positions.
D: Difference in rate when the watch is vertical versus horizontal.
P = Greatest difference between mean and daily rate
C = How sensitive the watch is to temperature changes (per °C)
R = Difference between first and last daily rate
And for the CIC, an extra mysterious parameter Vmoyc which we will get into below and see its effect
Here’s how the formula works for the trials 2011 to 2015:
N = 1000 – 500*│C│– 33.3*│D│ – 55*Vmoy – 45*Vmoyc – 10*P – 20*Vmax – 10*│R│ – 12.5*│M│
│x│ means taking the “absolute value” so that if a number is negative, we ignore the negative sign.
for the 2009 competition:
N = 1000 – 500*│C│– 33.3*│D│ – 100*Vmoy – 10*P – 20*Vmax – 10*│R│ – 12.5│M│
A theoretically perfect watch would score 1,000 points. More revealing, though, is not the perfect score but where the formula removes points most aggressively.
Key CIC results
Here are the results from the short lived trials.
By 2011, Greubel Forsey’s Double Tourbillon 30° Technique set what would stand as the highest named competitive score, recording 915 points in the tourbillon category.
I am not going to go over the history of this as others have already done it. Suffice to say it sadly died out as few of the large houses wanted to take on the reputational risk of either failing, not doing very well, and /or being beaten by indies or schools.
Where the formula bites
The real question is how those points relate to actual watch performance. A score only matters if we understand what kind of imperfection it punishes most heavily. So now here are
a table showing how the points come out relative to continuous halving of the limits of COSC.
a sensitivity analysis of each term in the formula from COSC
modern movement and surprising performances
how the small rule change from 2011 had a large influence
the unofficial record