Sleep is not homeostatic: every night is not the same because every day is not the same. You may have noticed this when you look at your sleep using the Weekly Sleep Report - no matter how good your sleep hygeine habits are.
Notice how Andi@Zeo's Deep sleep changed from night to night during a given week.
That said, there are also larger patterns. For example, weekend sleep for most of us has a different profile because we tend to be a little sleep restricted during the work week, leading to some (but not always enough) catch up on weekends. Also, sleep varies seasonally, as many people sleep more during the winter than they do when the days are long.
Some lifestyle habits can also come into play. A night out at the bar (or even a nightcap) can result in sleep fragmentation, which leads to more awakenings and a reduced ability to get restorative sleep, including Deep sleep. Caffeine consumption can also drastically reduce the amount of Deep sleep you get. Anxiety, room temperature, noise, and diet can increase sleep fragmentation as well, having an indirect effect on time in Deep sleep.
Also, let's not forget jet lag - which causes you to sleep according to your inner clock instead of your new, external clock. If you shift time zones too quickly before allowing yourself time to readjust (rule of thumb says it takes one day per hour of time change), your sleep stages could be thrown off.
But it can be tough to tell what’s going on from one night of “strange” data.
This is where keeping up with the Sleep Journal can really make a difference.
By recording information about your lifestyle habits, you can later make queries by comparing Deep sleep levels with specific conditions. Patterns emerge when more data is available. It may turn out that a combination of factors are involved, such as a few drinks at night as well as not getting any exercise the day after a long run.
Or was it the full moon?
At any rate, diving into your data and making an effort to keep a journal of your daily habits allows you to look for larger patterns in your sleep data. This can then allow you to see what is merely normal fluxuation - and what merits greater attention.