01 · The physicsWhy position changes the rate.
The balance wheel is supported by two pivots that rotate inside small jewel bearings (synthetic ruby). When the watch is horizontal (dial up or down), the weight of the balance wheel bears vertically on the jewel, and friction is minimal and uniform. When the watch is vertical (crown right, left, up or down), the weight bears sideways on the jewel — greater friction, slightly damped oscillation, and small changes in oscillation period.
It's not only about friction: the hairspring also behaves differently. In a vertical position, gravity acts on the spring coils themselves, which can no longer distribute their mass symmetrically. The result is a very slight change in oscillation period — and therefore in rate — compared to the horizontal position.
This variation is normal and expected: even COSC-certified movements are tested in five different positions precisely because every mechanical watch, however well regulated, shows a positional response. The goal isn't to eliminate it — it's to learn it and use it.
02 · The 5 positionsHow each position affects the rate.
ISO 3159 (and the COSC protocol) defines five canonical positions. The codes are French abbreviations used internationally — you'll also find them in COSC certificates and watchmaker's reports. The effect shown is a typical estimate: exact variations depend on the specific movement.
During waking hours, the watch on the wrist moves constantly from one position to another: vertical when writing, horizontal when resting your arm on a table, in between in countless variations. The rate you measure on the wrist is a weighted average of all these positions. That's why testing each position separately with WatchScope is the most informative way to understand your watch's character.
03 · The bedside tableThe overnight position trick.
If your watch gains +8 s/d on the wrist during the day but spends 7-8 hours on the bedside table dial-up (CH), those resting hours barely affect the daily average — CH is usually the fastest position, so it doesn't help compensate for daytime gain.
If instead you leave the watch overnight in a position that tends to slow it down (for example Crown right, 6H, which typically gives −8 s/d), those 8 hours of "passive correction" can bring the daily average towards zero. This isn't a trick for know-it-alls — experienced collectors and watchmakers themselves do it regularly.
The optimal bedside position varies from watch to watch: it depends on the caliber, the balance wheel poising, the wear. There's no universal answer. But there's a free tool that gives you the answer in five minutes: WatchScope.
Watch that gains +10 s/d on the wrist and whose 6H position gives −9 s/d: if you sleep 8 hours at 6H and spend 16 hours on the wrist, the weighted average is (+10 × 16 + (−9) × 8) / 24 = +3.7 s/d. On your own, without visiting a watchmaker, accuracy improves from +10 to +4. With a professional regulation (bringing the wrist rate to +2-3) you'd be near zero.
04 · WatchScopeHow to find the optimal position in 5 minutes.
Open WatchScope and use the Quick Test function (30-60 seconds per position). You just need a stable surface and the phone held still nearby. The protocol:
- CH — Dial up: lay the watch flat with the dial facing the ceiling. Test 60 seconds. Note the rate. This is your reference value.
- FH — Dial down: flip the watch over, dial facing the table. Test 60 seconds. Note the difference vs CH.
- 9H / 3H / 6H — Vertical positions: lean the watch on its side with the crown left, down and right. 60 seconds each. A bent paperclip, a book, or the rim of a glass can serve as support.
- Identify the slowest position: which of the five gives the lowest (or most negative) rate? That's the candidate position for the bedside.
- Verify with WatchScope for one week: sleep with the watch in that position, test the daily rate every morning. If the average moves towards zero, that's the right position for your watch.
Always test the positions at the same state of wind — ideally after wearing the watch on the wrist for a few hours. The balance wheel amplitude varies with wind, and so do the positional rates slightly. Comparing a test with a full mainspring to one with a depleted mainspring would give misleading results.
05 · Reading the resultsWhat to do with your data.
Once you've mapped the 5 positions, you have a positional profile of your watch. Here's how to interpret it:
- Total positional variation < 15 s/d (between CH and the slowest position): excellent. The movement is well poised. The bedside trick can refine accuracy without any intervention.
- Positional variation 15-30 s/d: normal for most production movements. The right bedside position can compensate much of the daytime gain.
- Positional variation > 30 s/d (one position gives +20 and another −15): indicates an unbalanced balance wheel, high beat error, or a movement that needs regulation. In these cases the bedside trick helps little — it starts from professional regulation.
- All five positions similar (variation < 5 s/d): high-quality movement or one with a silicon hairspring. Nothing to compensate because there's no variation to leverage — but overall accuracy is already excellent.
WatchScope's COSC Test function runs exactly this protocol automatically over the 5 positions, computes ADR, Δmax and the H-V delta, and gives a pass/fail verdict against the ISO 3159 standard. It's the most complete way to perform this analysis without workshop equipment.
Find your perfect
overnight position.
WatchScope measures rate in all 5 positions in a few minutes. Free on Android — no equipment needed, just your phone and a flat surface.