Improve your watch accuracy by resting it in a different position - How to do it. 

Watches that gain seconds are far more convenient than watches that lose seconds.. if the watch loses a minute, u have to set the watch, pull the crown, if it hacks, right on 12, then set the minute hand forward. With a watch that gains, just pull the crown on second hand 12, wait for the atomic clock to catch up, and as it hits in sync, push the crown in and ur good to go.

One solution is to have at least 2 automatic watches and rotate them every 2-3 days....personally I enjoy setting the time, date and seeing it come to life....my little pleasures in life...

I figured this out shortly after buying my Omega. It likes to be dial down. I always think, "why doesn't somebody do a video about this or that." Well here we are!

Swiss watch makers: Buy our $300,000 tourbillon Dave: leave your watch upside down

I use this method, and my Orient Bambino 38mm is off by +5 seconds for two months. I never reset the time until I change the date every two months from 31 to 1. Using this method makes my automatic watch more accurate than my quartz watches.

Very interesting. Never really thought about it nor did I think the orientation would have an effect on how fast or slow my watch runs. The results are surprising. Good video idea.

Thank you for addressing this topic, Dave! My most inaccurate automatic watch is the Seiko SKX007, today it's running +15s. It's old 7s26 has never been regulated by my watchmaker. I sat down my watches face up in a watch box.

I got a seiko 4r35 movement watch,which runs about 8spd on the wrist or whatever position I left it overnight,except one position which I found out after some experiments,when I left face down it overnight it runs about 8spd slower,by keeping that way,now I hve a accuracy of - + 2 spd which I think is great,guess I’m getting Rolex accuracy here 😎

Wow..I've been using this method on an NH35 movt. In 10 weeks the watch has gained 2 secs. Worn on my right wrist it loses 2 secs a day. When wresting at night , dial up it gains any lost time. I'm staggered by this level of accuracy. Thank's for confirming this, very interesting.

You’re the first person I’ve heard talk about this. I only have one mechanical watch, a Pagani with a Seiko NH35A. I noticed (very un-scientifically) that it ran about +4s/d when I wear it and leave it face up overnight, and about -2s/d when I wear it and leave it crown down overnight. When I take it off I compare it to atomic time (a Casio GW-M5610U that sits next to it) and choose how I set it down accordingly. I managed to keep it going for two months without unscrewing the crown (I eventually had to to reset the date at the end of a short month) and it never strayed more than 5 seconds from atomic time. Of course this doesn’t mean it’s as accurate as a quartz that would keep that kind of time without any intervention (not to mention that my system requires that I have access to atomic time to correct it towards), but it’s still cool.

Thank you Dave. I really never thought of doing that. I appreciate you sharing this.

This is really informative. Great content! Will definitely put my watches to the test.

Really informative and useful video. Great job.

A watch that gains and hacks is easier to reset without losing the position of the second hand.

Really interesting and informative video. Wouldn't have thought to try this but I certainly will. I have 2 automatics that I'm currently tracking the accuracy of and this will be another parameter to throw into the mix and see if it makes a difference.

I like your thoughtful presentation. I am one of those people that has gotten into a lengthy back and forth discussion on this topic. A person claimed that his mechanical watch was more accurate than a typical quartz watch, using some extremely convoluted logic. First he used +- 30 seconds per month as typical quartz accuracy, when it is really worst case accuracy. (Out of 80+ quartz watches, I only have 4 that are worse than +-20 spm. Most are under +- 10 spm and some are under +- 1 spm.) Since most quartz watches have minimal positional variation, each one tends to be fast or slow by a consistent amount every day/week/month. In his reasoning, quartz watches (left alone) could be off by 30 spm, but since mechanical watches might run fast or slow every day they would average out to better than 30 spm. Since he had figured out what positions to place his watch in each night to correct the time, he reasoned that his mechanical watch was accurate to 0 spm. I really like my mechanical watches, but I have no delusions about their accuracy. Unless I am performing a long-term accuracy test, I always check the time on a watch before wearing it. For typical quartz watches, I will correct anything over 1 second off. For my 262 kHz Bulovas, 0.5 seconds off is the limit. For mechanicals, 5 seconds off is my limit, but I will usually set them to the correct time every day I wear one.

Great video, and in general I'm impressed how often you find creative and interesting watch topics that other channels don't consider.

Dave, I really enjoy your posts; and the idea of collecting 'affordable' watches. I possess 5 seikos, from a 1966 sportsmatic that I found in a gutter(!) in 1976, a cheap but pretty snk 615, an snzh53 everyday and a sarb 065 that I can't stop looking at. Thanks for promoting stuff that you can enjoy without being a millionaire!

Thanks for the experiment! Great knowledge to have.

What an excellent video Dave! This is a very interesting and fun project to do! I think what makes ALL the difference is the app, as it takes away all the difficulty from keeping track of time on all your watches. I really love good useful tips videos as this; it sparkled a new activity for my watch collection during this lockdown :)

Interesting, and very useful knowledge. Thank you for sharing.

Real interesting, thanks for the info.

Great video, Dave!

Hi, Dave, I do this for my SNK 807 and it seems to do reasonably well to where I don't have to adjust the time as often (mine usually runs slow so it spends nights with the dial up). I think for your 12 o' clock up or down comment you can probably just use your watch pillow and position it in the box however you want it. Thanks for an informative video. Take care.

Great subject!

good tips! 👏🏻👏🏻

Good info! Never knew that! :) Thanks!

Positional variance in watches is an interesting subject, but you really need a time grapher to see the difference in positions easily. Also there is a difference between regulating a watch, which is changing the overall rate & adjustment, which is making the watch run at the same rate in different positions. Rolexs for example run accurately in 5 positions by the adjustment of microsteller screws on a free sprung balance wheel. On a watch balance without this or a similar arrangement the only way to make the balance run true would be for a watch maker to remove small amounts of material with a poising tool.

That’s a great observation, & I’ve got to experiment with my 3 autos. I did discover that one of mine is slightly magnetized...when I put a compass up to it, the compass points to it. I’ve got a $14 demagnetizer on its way!

Excellent tip! Just bought a Fossil Bronson automatic and found it to be 8 seconds fast, downloaded an app and found a resting position of 3 seconds slow to counteract the 8 secs.. not bad really for a 250 dollar watch. I'm a subscriber now!

Had a chat with my local watchmaker the other day and he was nice enough to put my watch on the time-grapher. I knew about the accuracy-issues due to orientation because I monitor that watch with the Watchcheck-app. But it was very interesting to get a more detailed and accurate glimpse at all the parameters behind that "daily deviation"-value. In my case (Sellita SW330 elaboré movement) the differences were: lying, dial up: +5.4s/d, Amplitude of 339, beat-error: 0.0ms hanging, 3oc/crown up: +6.2s/d, Amplitude 320, beat-error: 0.1ms hanging, 12oc up: -1.6s/d, Amplitude: 308, beat-error: 0.0ms I'm a bit puzzled about the amplitudes, since some are way outside the specs that Sellita give for this movement (max for "dial up" is 315° according to them). The watchmaker did set the angle incorrectly (52 instead of 51°), so perhaps that messed up the results a bit. On the wrist, the watch runs around +5s/d, so at night, I position it hanging, 12oc up where it runs at around -3s/d to -1.5s/day. Which, overall, gives me a daily deviation of +2.5s/d. Which I find highly acceptable for a mass produced, non-COSC movement. :)

Totally agreed. I have a Steinhart Ocean Two Premium with ETA 2892 movement and it benefits from resting it at different positions to "self correct" it's accuracy. I wear the watch when I leave for work at 6:30am and take it off when I get home from work at 4:30pm. I rest the watch dial face up until 9pm and the watch gains about 4 seconds since 6:30am. I then stand the watch crown up from 9pm to next morning at 6:30am when I put it on my wrist. At 6:30am the watch runs at exactly plus or minus zero second. Throughout the 10 hours wearing day I check the accuracy with my iPhone and the watch runs pretty much dead on or maybe gain a second toward late afternoon. By using this routine I have not needed to pull the crown out and reset the time to exact zero for over a month now. Of course some super expensive movements that are adjusted for 5 positions are super accurate but at a price level that I can't afford.

I have fantastic Seiko 4205, it runs +5mins per day, by this logic I'm better off wearing it on my wrist crown down. Love the channel.

What a good video. After 2 weeks of playing with position of watch (Citizen promaster diver) I can easily keep it inside +/- 3 sec per day (average accuracy in 14 day period is 0.7 sec/day (!)).

Great video.

Regarding accuracy, this "rest thing" is what i'm very curious about. Can't wait for my modest SNK803 to arrive so i can see what's up with automatics.

Does it work the same way for different watches or is it different for different watches?

Even doing this, I still felt compelled to check my mechanical watches twice a day against my phone. I started wearing a quartz and now I have 1 less thing to worry about in life :)

Just bought my first mechanical watch and noticed this after two days, was experimenting then just saw this! Glad to see my suspicions are confirmed

A great subject! A good channel. I have a watch winder which rotates my watches in all different positions, so I would just check my accuracies just based on 24-hour times, irreguardless of position.

Very useful, Cheers Dave! :)

I have a PD explorer 2, and leaving the watch face down at night, it works at +0 / +1 sec per day. It's amazing the effect of the resting position in the accuracy.

Good video. That Swiss Watch Company dive watch is nice. I've never seen one of them before.

Sieko 5 809, i kept playing with the balance wheel adjustment and got quite accurate like my quartz. Matches my daily routine, but won't necessarily be as accurate for anyone else to wear.

I have an ETA 2824-2. I started this method a few weeks back. Mine will slow it down crown down as well.

I do the same thing with my seiko srpb 43 and a few other watches with seiko movements.

Can you adjust the spinnaker to slow it down?

This subject is somewhat known, but nobody talks about the average watch temperature. A watch worn almost 24 hours a day will be more consistent than a watch worn part of the day. This affects most quartz watches as well, but they need to be timed over a week to a month (for a very good watch). The exception is the $7000 Citizen quartz watch that may be off by ONE SECOND after 12 months of setting the time.

Dors this work with a Seiko Presage movement 4r35?

I'll be honest, this is far more impactful than I would have expected! Especially for those who wear a watch daily like you mentioned. It would be interesting to take 5 or so of the same movement and see if the variations in time are consistent dependent on the position vs another movement. Also sorry I forgot to get back to you on the positions they are regulated at! This is why I needed to wake Jeremy up and ask him the moment you reached out Haha.

An elabore grade (ETA or Sellita) movement is adjusted to three positions: dial up, 6:00 up, 9:00 up. Maximum difference in rate between positions is 20 s/d.

I have a seiko 7s26 that varies wildly. When I wear it during the day, it will usually end around +20 to +30 but then when I take it off at night, resting on its side, in the morning, its down to about +10. Makes it easy to keep it accurate within a minute for a long time.

My Glycine Combat sub, which I think runs a Sellita SW200, gains 3.5 seconds per day, but loses just about the same at night when lying in a crown up position. This morning it was dead on +0s, and now (7 hours or so afterwards) it's one second fast, which I think is pretty impressive...

I’ve been experimenting with this very idea the past few months. I have a SW200 and it runs slow on the wrist and have been trying out setting it down in different positions to gain time overnight. I think I got it to a place I want it

This is interesting, i will be testing this with my both powematic 80 Hamilton with H-10 & H-40 movements. H-10 movements has been running 5 s/pd currently while H40 been running 3s/pd daily

Very very interesting. If I could get it down to just 1 or two or even three watches as a collection then I'd definitely test them. As it is most of my watches are stopped by the time it gets put on next. Anyone else have an SKX that runs slow on a winder? Like really slow.

Hey, talking about diver watches, can you please mention / review Citizen NY0040 aka Lefty?

For fun, try a time grapher. You can find the watches errors in any position in seconds. I got one from Amazon for CDN $200. approx.

I have been doing this for long time.

I've got an old seiko I'm going to try it out on. A staggering -120 s a day at the moment. I'll try your method tonight.... probably regulate it tomorrow:)

Have a few quartz. The Casios are particularly temperature sensitive. If I leave them off, they get cold and run faster by a few secs a week. Seems that when I wear them, the body’s temp makes them operate more accurately. These are the 5400 Gshock, a Solar analog, and a Casio solar/Waveceptor. Obviously the Casio quartz are totally more accurate than the mechanicals. Randy-IL

I just got an Orient Kamasu and it's running almost a minute fast after a few days. Do you know what position these watches require to be rested in order to slow it down a bit?

I noticed that about my Seiko 5SNK615 it would be off about 6s a day ,so I put it on its side and it would get that 6s back when I woke up.

If I use this method, would hacking the watch the next day cancel it out? Or would it continue with that accuracy even when I reset the time?

Oh, I have so much to say about this topic and video.... where to start? I could start a watch channel and have a lot of material! Ha ha! But then, I'm a watch guy so...Great channel, Dave!

I have my first automatic arriving this week, I will give this a try. I have an interesting watch from when I was a Apple software developer in Japan in the early 1990's. I put a #377 battery in it and it works. Any ideas how I could get an accurate idea of the value? It has the old apple logo on the dial with a Japanese quartz movement, very unique.

"the ORIENT-ation of the watch"...🤣 pun intended? @2:52

2:27 - which Laco watch is that please?

Another great video Dave! Love how you come up with creative video ideas!

My 1995 Omega Speedmaster Automatic is so ridiculously accurate. I still can't believe it's more accurate than a 10-year-old Cosmograph Daytona

@6:06 where'd you buy that watch strap? Looks amazing!

My longines watch looses time when worn on wrist but keeps time when it's in watchwinder . Any ideas why

Why does the watch check app need access to your location and media file including videos and photos?

Thanks for the great content in this review! Very interesting. One of the things that I now look at, and I find it by looking at the technical specifications that are offered online, is the accepted factory tolerances/variances of Watches in different positions. A lot of times the quick look spec sheets on a watch will state Best results of a particular position(varies)...For instance the new Sieko presage(Cocktail time)...I believe they state the tolerance as +20/minus 30. If you look at the technical manual that is if the watch is in the upright(12 facing upward) position. I don’t consider that very tight tolerances but if you look closely the full technical manual states the factory tolerances are actually much bigger(worse) In other positions...The worst being horizontal positioning(Watch laying flat/parallel to the ground). That stated as +30/-45. I consider that awful for a watch that costs $500! By the way it’s 14.6 mm thick... supposed to be a dress watch...???So I used the positional tolerances to determine I wasn’t going to buy the watch! Cheers!

I've been having quite a varied experience with my new Seiko 5 automatic, after buying it for myself as a Christmas gift. I heard about this "watch position" dynamic months ago, and learned very quickly that while resting the watch on its side would slow it down immediately, resting it overnight (dial down, since it has a metallic wristband) did much less. Despite this, I found that it would still lose about 3.5 or more seconds per day . . . until quite recently, when it suddenly started running FAST, to the tune of the 7 seconds or so that you mention. I find myself wondering if it's going through some elaborately long "breaking-in" period, as it decides how to run.

Been using this technique on my Sellita sw200 and miota 9015. Works great. I'm able to keep both at zero seconds per day months on end.

Great videos. Can I have my skx007 regulated and would it stay accurate? Thx

I have been doing this on my own for long time, the problem is that usually it works to slow down watches that run faster, while I have never get great results on watches that are slower (basically I can't find a position to accelerate them)

Victorinox Alliance Automatic... anyone? I love this watch simplicity and casual elegance. Wish could fond more watches like this... closest is Omega DeVille

Hi Dave , great test , very interesting . Just wondering what kind of accuracy you are getting from your Gaungin GJ16034 ? , I've only had mine about a week and over the past 6 days i've only been winding it and leaving it in my watch box and it's only running at +0.8spd , which i think is crazy for a 40 euro watch :) ..... My Boldr Expedition with the SW200-1 running at -11spd , but now i'll be changing resting positions to see if it runs better , thanks man

is this just for automatic watches or does it apply to manual wind-up watches?

I have two watches I rotate during the work week and have figured them out such that I can keep them close to spot on indefinitely. If it wasn't for daylight savings time I'd never have to reset them.

I have noticed my seikos arent affected much by the position u store them in they are a little but not much at all especially the new 4r36 movements but my miyota movements are. Never tried with my ETA 2836.

In the 60’s, within the Rolex guarantee they would tell you what position to leave your watch to either gain or lose seconds per day. With my 69 DJ, gains about 2sec a day and after 3days I place it crown up and it loses those 6 second gained 👍🏽

Does this work with Ming world timer?

I tried this with my new Seiko 5 sports watch a month ago....it is running about 8-10 seconds slow/day, which is annoying. I'd rather it run fast. Each night I would put the watch in a different position on my dresser. Crystal side up, crystal side down, crown up, crown down....the positions didn't affect the watch at all unfortunately.

I've known about this for quite some time. Could this be harmful in any way to the watch? What position do watchmakers recommend when you take the watch off? I imagine dial up.

Watch speeds 15 seconds after a couple of hours of drum playing. ¿Is this normal? So, I was in the middle of trying out this thing, after measuring the precission of my diver with a NH35 movement, and figuring out that the majority of the deviation was happening at night when I took my watch off. From 9.00 am, when i checked the accuracy to 12.00 pm when i took it off, it had only sped between 1 and 2 seconds, and during the night time it would speed about 5 seconds more, giving a total of 6-7 seconds per day. So I left it with the crown facing down for a couple of nights, it sped less than a second each of the nights. Yesterday, I checked the accuracy after playing on the drums, and it had sped 15 seconds in a couple of hours... So, i suppose I could keep it quite accurate by not wearing the watch while I play the drums and leaving it sideways at night. I'm still investigating the fact, and found this video, which was interesting and instructive.

Hello I do have a question. If i leave my watch in the dial down position isn't there any chance to do anny damage to the watch or movement?

Do watch regulating services exist? If so who offers this? Thanks.

My Orient bambino generation 2 version 3 runs -15 seconds per day.

My king turtle runs almost on time to -2sec a day when I wear it to bed. It’s +4 laying down face up.

If you alternate arms you wear it on, shouldn't that work too?

That's a old trick !! and it's work as you say 😀

Wow, well I tested it face down nothing face up nothing crown up slows it substantially, crown down neutral makes it accurate doesn’t lose a second. But I can’t wear it that way lol Movement nh35

Subscribed. Rest my watch crown 👑 up. Cuz it's the easiest way for me (growing up with G-Shock 😄)

Hi. I'm not a trol... Genuine question : so you're recommending that app to check the accuracy but isn't holding your phone next to the watch going to magnetise your watch and ironically affect the way your movement work? Could you check with the compass what happens if you put a compass next to these watches?

I have recently begun to take advantage of this phenomenon but my worry is that this adds extra wear and tear on the movement.

your segesstion is one of many, and it is a very good idea, but it still demanding from you attention, for some people it might be a trouble, but this is it when you buying wathces under 500$ if people buying hand winding watches they are also need to adjust to routine,

Awesome vid :D.. But you can set the watch 12 up on the pillow in the open watchcase :)