Most of us assume the night sky is the one thing that doesn't change. Faces age, houses get sold, people leave, but the stars are fixed points we can count on. That's mostly true, with one large exception. High in the northeast on autumn evenings there's a star that loses about two-thirds of its light in a few hours and then gets it back. It has done this every two days and twenty-one hours for longer than anyone has been watching. Almost nobody notices, because almost nobody looks twice.

The star is Algol. In Arabic it was ra's al-ghūl, "the demon's head." In Greek myth it is the eye of Medusa, whose severed head Perseus carries across the sky. Whether ancient observers named it for its changes is still argued over. What isn't argued over is that you can watch it change yourself, tonight or within the next few nights, with nothing but your eyes and some patience.

What's actually happening up there

Algol isn't one star. It's a pair (plus a more distant third companion) orbiting so close together that no telescope can separate them. The brighter one is a hot, blue-white star. The dimmer one is a cooler, larger orange subgiant. From Earth we see their orbit nearly edge-on, so once every orbit the dim star passes in front of the bright one and blocks much of its light.

Astronomers call this an eclipsing binary, and Algol is the classic example. It usually shines at about magnitude 2.1, close to the brightest stars around it. During an eclipse it fades to about magnitude 3.4. The magnitude scale is logarithmic, so that drop means losing roughly two-thirds of its light. The whole event takes about ten hours from start to finish. The deepest part lasts only a short time, and then the star spends a few hours brightening back up.

The story of how we figured this out is one of the most moving in astronomy. The Italian astronomer Geminiano Montanari recorded in the 1660s that Algol's brightness varied. More than a century later, in 1782 and 1783, John Goodricke pinned down the period and suggested that something orbiting the star was blocking its light. Goodricke was a teenager. He had been deaf since early childhood, and he worked with his neighbor Edward Pigott from a window in York. He died in 1786 at age 21, shortly after being elected to the Royal Society. His explanation turned out to be essentially correct.

Algol still holds a puzzle, too. In a pair of stars born together, the heavier one should age faster. In Algol, the lighter star is the more evolved one. This is called the Algol paradox. The current explanation is that the orange star was once the heavier of the two and has since lost much of its mass to its partner. The pair has been reshaping itself, quietly, for a very long time.

Why you can't just look and see it

This is the part that applies to more than stars.

Your visual system is poor at judging absolute brightness and very good at comparing. Look at a single star and you can't really say how bright it is in any fixed sense. Put it next to another star and you'll see the difference immediately. Perception researchers have known this for a long time. Weber's law describes how we notice changes relative to a baseline, and much of what we "see" as brightness is really contrast.

There's a related and more uncomfortable finding. In a well-known 1998 study by Daniel Simons and Daniel Levin, an experimenter stopped strangers to ask for directions. Partway through the conversation, two people carrying a door walked between them, and the experimenter was swapped for a different person. About half the participants didn't notice they were now talking to someone else. Psychologists call this change blindness. We don't keep a detailed record of what we're looking at. We keep a rough sketch and assume the rest stays the same.

That's why Algol can lose two-thirds of its light over a single evening, in plain view of millions of people, without anyone reacting. Nobody kept a reference point.

The same goes for the people you see every day. A friend who has gone a little quieter, a parent who moves more slowly than last year, your own habits shifting. Change you see constantly is the hardest kind to notice. You notice it when you have something to compare against: an old photo, a note in a journal, the version of a person you remember clearly.

How to catch Algol in the act

Variable star observers have a simple, reliable technique for this. Don't try to judge the star by itself. Compare it with neighbors whose brightness doesn't change.

First, find it. On autumn and winter evenings in the Northern Hemisphere, locate the W of Cassiopeia, then look toward the Pleiades, the small misty cluster rising in the east. Perseus sits between them as a curved chain of stars. Its brightest star is Mirfak. Algol is a little way off from Mirfak toward the Pleiades, slightly separated from the main chain, like a head held at arm's length.

Next, pick your comparison stars. Two work well:

  • Almach (Gamma Andromedae), around magnitude 2.1, which matches Algol at full brightness.
  • Epsilon Persei, around magnitude 2.9, a little fainter than Algol usually is.

When Algol looks about as bright as Almach, it's at normal brightness. When it's clearly fainter than Epsilon Persei, you're watching an eclipse.

Then check the timing. The eclipses repeat every 2 days, 20 hours, and 49 minutes, so they shift to a different time of night each cycle. Some of them happen in daylight or while Perseus is below your horizon. Predicted times of minimum are published by astronomy magazines and by organizations such as the AAVSO (American Association of Variable Star Observers). Look up the next minimum that falls during your dark hours.

Your next moves

  • Tonight, find Algol and compare it with Almach. Step outside after dark, find Cassiopeia's W, follow it toward the Pleiades, and locate Algol in Perseus. Write down one sentence: is it brighter than, fainter than, or equal to Almach?
  • Look up the next Algol minimum for your time zone. Search "Algol minima" and pick the next eclipse that falls between full darkness and midnight. Put it in your calendar with an alarm 3 hours early.
  • On eclipse night, check every 30–45 minutes. Each time, judge Algol against Almach and Epsilon Persei, and write down the time and what you saw. Five or six notes are enough to see the dip and the recovery in your own handwriting.
  • Do the human version. Find a photo of yourself or someone close to you from three to five years ago and put it next to how things are now. Pay attention to what changed that you never noticed changing.
  • Pick one personal reference point. Choose one thing about your life you want to keep track of, such as sleep, mood, or a relationship, and write one honest line about it each week. Without a baseline, slow changes go unnoticed.

The sky that changes while you watch

The first time you see Algol in the middle of an eclipse, clearly dimmer than a star it matched two nights earlier, it's a small shock. It doesn't look like what you see in space photos. It feels more like realizing a room you walk through every day has been rearranged. The sky turns out to be less fixed than it seemed, and you realize you only knew that because you compared.

The hard part of watching Algol is being sure which star you're looking at, especially the first few times, when Perseus is just a faint scatter of dots. That's where Astra helps. Point your phone at the northeast and it will name Algol, Mirfak, Almach, and the other stars around them, so the comparison you're making is the right one. After that, put the phone away and keep watching. If you want a reliable guide for that first night, Astra is here.