How is mechanical watch accuracy measured? Compare the time shown by the watch with a dependable reference at known moments, then calculate how the difference changes over a measured interval. The result is commonly stated as seconds gained or lost per day. A useful measurement describes the method and conditions as well as the number; one quick glance alone cannot tell the whole story.
What does a mechanical watch’s rate tell you?
A mechanical watch keeps time through a powered train of wheels regulated by its balance and escapement. Its rate describes how fast or slow it runs compared with a reference over time. If its displayed time moves 9 seconds ahead of the reference in one day, its measured rate for that interval is +9 seconds per day. If it falls 9 seconds behind, the rate is −9 seconds per day.
The rate is different from the watch’s offset. Offset is the difference at one particular check. A watch could be 40 seconds fast right now because it was set 40 seconds fast and has kept good time since. Conversely, it might be exactly aligned now after gaining time and later being reset. To measure rate, compare how the offset changes between two or more checks.
Rate is also not the same as consistency. Imagine Watch A gains about 8 seconds each day for a week. Watch B gains 20 seconds on some days and loses 12 on others, averaging near +8 over the week. Both might show a similar average, but Watch A has behaved more consistently. For practical evaluation, keep both the overall average and the pattern of individual readings.
A measurement applies to the particular watch, interval, and conditions you tested. Position, winding level, activity, temperature, and how precisely you read the reference can all affect the result. The figure is evidence about observed timekeeping, not automatically a diagnosis or a guarantee of future performance.
What do you need for a useful home measurement?
You do not need specialized equipment to make a practical comparison. You need a dependable time source, a repeatable way to read it, and a record of what you did. The method should be simple enough that you can repeat it without changing the rules partway through.
A stable reference: Choose a clock or device whose time is synchronized and whose seconds are easy to read. The quality of the conclusion cannot exceed the quality of the reference and comparison.
A consistent reading point: Decide how you will compare the watch and reference. For example, note both at the start of a chosen minute, rather than trying to compare unrelated moments.
A log: Record date, time, watch display, reference display, and any relevant changes such as winding, stopping, or a different overnight position.
A meaningful interval: Use several days when practical. A difference of a few seconds is difficult to estimate accurately if the entire test lasts only a few minutes.
A synchronized phone or computer can be convenient, but consider what its display lets you observe. A display that shows only whole seconds does not support claims to fractions of a second. If you use a clock with a seconds display, note the instant you compare it and use that same method every time. A radio-controlled clock or other synchronized source may also work. The goal is not to make the setup look technical; it is to keep it stable.
Before starting, note whether the watch is running, whether it has recently been wound, and whether you know its previous setting. Do not open the case to inspect the movement. If you are evaluating a specific watch before making a decision, review the seller’s own product details separately from your timing test. For example, the PRO Tour I product information describes the watch and its seller-published movement details; those details are not a substitute for an independent timing measurement.
How do you measure the rate step by step?
There are two practical starting methods. You can set the watch to the reference and measure the difference that accumulates, or you can leave it as it is and record its initial offset. Both can work. The second method is helpful if you do not want to reset the watch; it just requires signed arithmetic so that gains and losses are handled correctly.
Choose one reference. Use the same clock or device for the start and end checks. Decide in advance whether you will reset the watch or record its current offset.
Make a starting record. Write the date and exact time, then note what the watch displays and what the reference displays. If you are comparing seconds, record which is ahead and by how much.
Keep a consistent routine. Wear the watch as you normally would, or leave it in a chosen position if that is what you intend to study. Note unusual changes instead of quietly changing the test conditions.
Make the next comparison after a known interval. A full day or a number of days is easier to interpret than a quick check. Record the ending offset in the same format as the starting offset.
Find the change in offset. Subtract the starting offset from the ending offset. Keep gains positive and losses negative.
Divide by elapsed days. The result is the observed rate per day for that interval.
Repeat the measurement. Several intervals show whether the watch follows a fairly steady pattern or changes from one period to another.
Worked example: At the first check, a watch is 4 seconds fast. Three days later, it is 22 seconds fast. The offset changed from +4 to +22, a gain of 18 seconds. Divide 18 by 3 days: the measured rate for the interval is +6 seconds per day. Subtracting the initial offset matters. The watch did not gain 22 seconds during those three days; it ended 22 seconds fast in total.
Here is a second example with a sign change. A watch begins 3 seconds slow, which is an offset of −3 seconds. At the next check it is 9 seconds fast, an offset of +9 seconds. The change is +9 − (−3), or +12 seconds. If that interval lasted two days, the measured rate is +6 seconds per day. Looking only at the magnitudes, 3 and 9, and subtracting them would miss the fact that the watch crossed zero.
If you reset the watch precisely against the reference at the beginning, the starting offset is zero. At the end of four days, if it is 28 seconds fast, the measured rate is +7 seconds per day. In real home checks, the reset and reading may not be exact. Record the actual starting comparison when possible rather than treating an approximate alignment as perfect.
How can you keep a timing log that is easy to interpret?
A useful log is more than a list of final numbers. It lets you reconstruct the comparison and see whether a surprising result followed a change in use or conditions. A notebook or spreadsheet is fine. Use one line for each check, and do not erase an unusual reading; make a note beside it and decide later whether the test should be repeated.
A practical entry can include:
Date and reference time: Keep a record precise enough to identify the interval.
Watch offset: Write “+12 seconds” or “−5 seconds,” not just “ahead” or “behind.”
Elapsed time: Note how many hours or days passed between comparisons. If the interval is not a whole number of days, use elapsed hours and convert consistently.
Use and position: For example, worn during the day and left dial-up overnight, or not worn and stored crown-up.
Interventions: Note a reset, stop, winding change, or any interruption. If the watch stopped, the interval no longer represents uninterrupted running.
Context: Record an unusual impact, a large change in routine, or another event you think could be relevant without assuming it caused the result.
Consider this sample log: Monday, 8:00 a.m., +2 seconds; Wednesday, 8:00 a.m., +14 seconds; Friday, 8:00 a.m., +27 seconds. The first two-day interval changed by +12 seconds, or +6 seconds per day. The next changed by +13 seconds in two days, or +6.5 seconds per day. Across the full four days, the change was +25 seconds, or +6.25 seconds per day. The full-span average and the two shorter readings tell a similar story, while still showing ordinary variation in the observations.
Now consider a different log: +2 seconds Monday, +20 seconds Tuesday, and +1 second Wednesday. The two-day overall change is −1 second, an average rate close to zero. But that average hides a sharp swing between checks. Do not report only the average as if the watch were reliably keeping time near zero. Repeat the comparison, make sure the reference and reading method were consistent, and note the conditions around each check.
Why should you test across more than one day?
Short intervals can magnify small observation errors. If the watch’s displayed seconds and the reference are each read a second late, the apparent difference can shift even when the watch has not suddenly changed its behavior. Over a longer interval, the accumulated change is easier to distinguish from the uncertainty of a single comparison.
Longer testing is not automatically better if the conditions change without a record. If you wear the watch on Monday and Tuesday, leave it unworn Wednesday, reset it Thursday, and compare on Friday, the resulting number combines different routines and an intervention. That may resemble real ownership, but it is not a controlled comparison. State the pattern honestly and, when the aim is a baseline, restart a clean interval after a reset or interruption.
Use both a broad view and a close view. The overall rate across several days answers, “What was the average change over this period?” Individual intervals answer, “Did the behavior stay similar from one interval to the next?” If the results differ substantially, repeat the test before drawing a conclusion. First check for a recording or arithmetic mistake; then check whether winding, wear, position, or interruption differed.
How can you plan a repeatable three-part test?
A simple test plan prevents the comparison from turning into several different experiments. Think of it in three parts: establish a baseline, observe ordinary use, and repeat the baseline if anything unusual happens. The goal is not to control every detail of daily life. It is to know enough about the test to explain what the result means.
Part one: establish your starting point
Choose a day when you can make an unhurried starting comparison. Use the same reference you expect to use at the end. If the watch is already running and you prefer not to reset it, record its offset precisely enough for your method. If you do reset, record that fact and the time. Note the watch’s usual state: whether you have been wearing it, whether it has just been wound, and where it will rest when it is off your wrist. You do not need to create a perfect laboratory condition; you need a baseline that you can describe.
For a straightforward everyday test, choose a check time you can repeat, such as before starting your morning routine. Decide whether you are comparing the displayed seconds directly or using the minute change as a coarser check. Keep the method consistent. If your reference displays only whole seconds and your eyes cannot reliably judge a fraction, record whole seconds and accept that limitation rather than writing a falsely precise decimal.
Part two: observe the usual routine
For the next several days, wear and rest the watch as you normally would. Make a note if the routine changes materially. For example, your baseline may be a watch worn through the day and left dial-up overnight. If you leave it unworn for one full day, note that. Do not label the day “bad data” automatically; it can still describe what happened. The note simply helps you decide whether it belongs in an average meant to represent normal wear.
At each check, write down the offset rather than trying to remember how far ahead it looked. If Monday is +5 seconds and Tuesday is +11, the one-day change is +6 seconds. If Wednesday is +18, the next change is +7 seconds. Those adjacent results are close, so the short log suggests a fairly steady pattern, although more days would provide a stronger basis. If instead Wednesday is +2, the second interval is −9 seconds. Verify that the reference was read correctly and consider whether the watch’s position or use changed before deciding that the watch itself behaved differently.
Part three: repeat or narrow the question
When the first run produces a surprising number, repeat the same routine before changing anything. A second run that produces a similar result gives you more confidence that you observed a real pattern. If the second run differs, compare the notes. Perhaps one run began after a reset, one included a stop, or the overnight position changed. These differences do not prove a cause, but they suggest what to keep consistent in a later comparison.
Only after you have a repeatable everyday baseline should you try a narrower question, such as whether one overnight position changes the morning reading. Change that one condition while keeping check times and the broader routine as similar as practical. Write down the question before the test. For instance: “Does my morning offset differ when the watch rests crown-up rather than dial-up?” That is more useful than “Which position is best?” because it identifies the conditions compared and does not assume one result applies to every watch or owner.
This sequence also helps when you talk to a professional. You can distinguish a stable everyday pattern from a one-off reading, show how the number was calculated, and explain whether a position experiment was part of the record. A clear test plan does not replace inspection, but it makes your observations easier to understand.
How do position and daily wear affect the measurement?
A mechanical movement can show different rates in different positions. Dial-up, dial-down, crown-up, and other resting orientations place the movement differently relative to gravity. The watch also experiences motion on the wrist, changes in temperature, and variation in how fully it is wound. These are reasons to record conditions, not reasons to dismiss every measurement as meaningless.
If your question is how the watch performs in normal ownership, wear it in your normal routine. Do not deliberately place it in a special position simply because you heard that position may make it gain or lose less. Your actual daily pattern is the relevant test. Keep notes such as “worn through the day; left dial-up overnight” so that another week of results can be compared fairly.
If you want to explore positional variation, change one variable at a time. For example, keep the daytime routine and winding pattern similar, then compare several overnight periods in one resting position with several in another. Use the same reference and check time each morning. One night in each position is a weak comparison because a small reading error or a different winding level can dominate the outcome. This home exercise can indicate a pattern, but it does not reproduce a watchmaker’s full test.
For a watch worn inconsistently, make the record reflect reality. Note the approximate periods it was worn, left resting, or stopped. A simple average over a period when the watch was not running cannot describe its running rate during stopped hours. Separate the intervals when it was running from periods when it had stopped, and avoid dividing by calendar days as if the movement had been operating continuously.
What does a timegrapher measure?
A timegrapher listens to the sound of a watch’s escapement and displays readings such as estimated rate, amplitude, and beat error. The rate is an estimate of gain or loss derived from the timing of the beats. Amplitude concerns the balance’s swing. Beat error describes the difference in timing between alternating beats. These measures can help an experienced operator assess a movement, but each needs context.
A timegrapher is a snapshot under a particular setup, not a week-long record of how the complete watch behaves on your wrist. The watch must be held and positioned appropriately, and the operator needs suitable settings for the movement. Background noise, poor sensor contact, winding state, or incorrect setup can make the display misleading. Several digits on a screen do not make the result more trustworthy if the procedure is wrong.
The home log and a timegrapher answer related but different questions. Your log describes observed performance over time, under your recorded routine. The instrument can show clues about movement behavior in selected positions during a test. A qualified watchmaker can combine readings with symptoms, inspection, and service history. A single timegrapher number does not establish the cause of a timing change or prove that service is needed.
How should you interpret the result?
Start with the specific maker’s published guidance for the watch or movement, if available. Do not assume that every mechanical watch shares one accuracy tolerance. Different models and makers may state expectations in different ways, and a general rule of thumb is not a promise for an individual watch.
Next, compare repeated results. A steady rate and a pattern that jumps between gains and losses are not the same, even if their averages match. Report the interval and the conditions: “It gained 24 seconds over four days while worn in my usual routine” gives someone more to evaluate than “It runs fast.” If you have daily figures, include the range or explain that one interval differed.
Finally, consider what you need from the watch. Some owners are comfortable correcting the display occasionally; others want closer agreement with a reference. That preference is personal. It does not change the watch’s published specification, but it can help you decide whether to seek advice. Avoid treating the price, appearance, or category of a watch as proof of a particular rate.
When considering an automatic watch with a strong design or artwork focus, separate the questions you are asking. Product information explains what the seller says is included; a timing record describes how a particular watch performed during your test; and a watchmaker’s inspection addresses technical condition. For the golf-art PRO Tour I, Bart Forbes Chrono describes a Swiss-made, hand-assembled automatic watch with an SW200-1 Aig2 movement. Those are seller-published claims, not independent timing-test results. You can review the PRO Tour I product details and ask the seller for clarification about any performance information that matters to you.
What should you do if the measured rate changes?
A changed result does not by itself identify a cause. The watch may have been tested in a different position, wound differently, worn for a different amount of time, or checked with a less consistent reference. Begin by repeating the same measurement before deciding that the movement has changed. If the pattern continues, preserve your notes so that you can explain what you observed.
If the watch stops unexpectedly, develops a much larger gain or loss than your earlier record, or shows another change that concerns you, contact the seller or an appropriate watch service professional. Describe the time span, the number of seconds gained or lost, the use pattern, and whether the watch stopped or was reset. This is more useful than trying to diagnose a cause from a number alone.
Do not open the case or attempt an internal adjustment just to improve a home measurement. Regulation and repair require appropriate tools and knowledge. A change made without a reliable diagnosis can complicate the problem. Keep the test focused on observation, and leave internal work to a qualified professional.
Common measurement mistakes to avoid
Confusing total offset with daily rate: Being 30 seconds fast at a check does not mean the watch gains 30 seconds every day.
Ignoring the starting offset: If you did not reset the watch, calculate the change between its starting and ending offsets.
Dropping the sign: A move from −3 seconds to +9 seconds is a 12-second gain, not a 6-second difference.
Testing for only a few minutes: A tiny interval makes comparison and reading error loom large relative to the change measured.
Using different references: A change in clocks or methods can look like a change in the watch.
Changing several conditions at once: Different positions, winding, and wear patterns make it difficult to tell what the result represents.
Leaving out interruptions: A reset or stopped watch changes the meaning of the interval. Record it or begin a fresh test.
Treating an instrument reading as a complete verdict: A timegrapher reading describes one setup and needs interpretation.
Applying another model’s tolerance: Check the guidance for the specific watch instead of borrowing a number from another one.
Frequently Asked Questions
How many days should I track a mechanical watch?
Several days of consistent checks are more informative than a brief comparison, especially when you want to estimate a daily rate. If readings vary, extend the record and note the conditions rather than relying on a single average.
Should I reset the watch before measuring?
You can reset it to the reference or record the initial offset and subtract it from the final offset. Either method works if you note the start accurately, use the same reference, and handle gains and losses with their correct signs.
Can a phone be used as a reference?
A synchronized phone can be a convenient reference if its display makes the comparison clear. The precision of the result is limited by the seconds display and how reliably you can read it. Use the same device and method throughout, and do not claim finer precision than your observations support.
Does a timegrapher tell me whether the watch needs service?
Not by itself. It provides readings that may help an experienced operator assess movement behavior, but setup, the watch’s condition, symptoms, and service history matter. A rate number alone does not diagnose a fault.
Explore the PRO Tour I product information
Final perspective
To measure mechanical-watch accuracy, compare the watch with a dependable reference over a known interval, calculate the change in offset per day, and repeat under conditions you have recorded. The result is most useful when you distinguish rate from offset and describe what the watch was doing, rather than treating one number as a specification or diagnosis.