If you're talking about the bone-knob on the pinkie side (unsure of the proper name)...the band should sit just behind it. You may be wearing it to low and it's interfering with your natural wrist movement. The watch will look better there as well.
Wow, I just realized that I've probably been wearing my watches incorrectly since third grade. I usually wear link bracelets too which are less forgiving and I also sleep with it on. My links go right over the knob. This week I stopped wearing a watch for the first time since I first put one on. It has been very weird feeling. Like my arm is naked. First of all I think my watch knew I was replacing it because one of the link pins broke (two per link, so it was still holding on somewhat). Then a few weeks later the thing just stopped accurately telling time. Maybe the battery was just dying, but I've never had it start and stop like this before. It was making me really late sometimes if I wasn't closely paying attention.
Anyway, so where my watch was I actually have a dent in my arm compared to my other arm. Even days later after I stopped wearing it. I didn't do a good job adjusting links over time. Even my skin has a really weird texture where the watch used to be, especially on the underside of my wrist. Since my SS 42mm loop isn't coming until June I figure I can finally let my arm repair itself. I'm even missing a lot of hair there since I wore watches well before puberty and maybe it somehow disrupted it? It's quite apparent when looking at my wrist that a watch is missing. It's pretty hilarious actually. Maybe I can finally start wearing a watch properly without it being too tight when I get my milanese loop.
The magnetic closure on mine is rock solid - doesn't budge at all without me actively pulling it open. I find it extremely comfortable as a strap and I like the infinite adjustability.
I do have to wonder if the reason some people have trouble and others don't is because of different sized wrists. Depending on the wrist size the magnetic part will be in a different position. I would hypothesize that if the magnet is on the side then it's less likely to be knocked off while working on a computer, etc. But on the bottom it's more likely to encounter friction.
