One year from today, those of us in North America will have an extraordinary opportunity to view a total eclipse of the Sun that blasts through major cities like Dallax TX. This had me thinking about total solar eclipses, and how unlikely they are to occur…

We don’t get total solar eclipses simply because there are three balls in the sky that occasionally line up. We get solar eclipses because of an astronomically incredible coincidence that probably doesn’t happen anywhere else in our galaxy, even if we assume there are millions of planets and tens of millions of moons.

Here’s how I made the scale of the improbability meaningful for myself:

First, I started off by finding the biggest ball in my home: a 54cm diameter yoga ball. This became my standard of measurement. One yoga ball equals one Sun. Thus, one inch equals 66,666km.

Now, I had to find something to represent the Earth. At this scale, the Earth’s diameter would be 5mm, exactly the size of a juniper tree berry (abundant here in New Mexico.) If you don’t know what a juniper berry is, imagine a middling to small blueberry.

Lastly, I had to find something to represent the size of the Moon. After discounting coriander seeds (too big) and poppy seeds (too small) I found that a mustard seed is a perfect fit for the requisite 1.25mm.

So now we have our three balls to place in the sky: A giant yoga ball, a juniper berry, and a mustard seed.

Let’s put together our model of the Solar System. We place the yoga ball (Sun) on the ground. Now we walk 56 meters away and hold our mustard seed (the Moon) 15 cm away from our nose. At our eye level, we’ve got the juniper berry (Earth).

A total solar eclipse is only possible because from the vantage point of the juniper berry, the mustard seed (at 15 cm) appears to be the EXACT SAME SIZE as the yoga ball (56 meters away).

That is a honking crazy coincidence. And in fact, it’s a short lived coincidence that also happens to coincide with this period on Earth in which humans are around to witness it. In about 600 million years, there won’t be any more total solar eclipses. Every year, the Moon pulls away from us tiniest bit. Remember how your mustard seed (Moon) is just 15 cm away from the juniper berry (Earth)? In 600 million years, the mustard seed will have drifted another 7cm away from the juniper berry. This minuscule deviation is sufficient such that the mustard seed’s profile will no longer cover up the yoga ball.

So, what are the chances of this happening? Specifically: what are the chances that for any given star system, a planet’s moon will be the exact right size and distance from its central star to completely eclipse its central star? I don’t have the math skills to calculate that out, but I would guess that it’s pretty damn unlikely and almost certainly unique within our galaxy. How lucky that life evolved here so we could witness it!

How do we explain this statistical unlikelihood? My only suspicion is an unsettling one. To me this bewildering coincidence is the best evidence we have that what we call reality is in fact a simulation being run by computers. I imagine that one of the programmers of our reality thought to itself, “Let’s make life more interesting and beautiful for our simulated Earthlings by providing them with a spectacular total solar eclipse every 18 months. All we have to do is make a moon that’s exactly the right size and distance from the Sun, and make sure that civilization is happening during the brief window in which the Moon stays in just the right place.” It’s an unsettling concept, but a serious one first proposed by Hans Moravec.