What do you actually do when two of your devices disagree about the same night

Two wearables on the same body, same night, and they report different deep sleep. Sometimes an hour apart. Most people I talk to have quietly picked a favourite device and stopped opening the other one.

I want to know what you actually do, because I think the common answers are all defensible and I cannot tell which one is right.

Three things I believe and would like argued with.

First, deep sleep is the least comparable number across devices. The staging is inferred, every vendor infers it differently, and there is no shared definition to reconcile them against. Comparing two vendors on deep sleep is closer to comparing two opinions than two measurements.

Second, the disagreement matters less than the direction. Two devices that sit an hour apart on the absolute number can still agree that last night was worse than your last two weeks. The agreement is in the trend, and the trend is the part you can act on.

Third, a population range cannot settle it. Both readings can sit inside normal and still be telling you different things about you. Normal for a population is not the same question as normal for you.

So when yours disagree, do you pick one device and commit, average them, throw away the absolute number and keep only the trend, or drop the metric entirely and stop looking at it.