(This is Part 4 of a series on whether the universe can be cyclic. Read Part 1, Part 2, and Part 3 first.)

So we left off with a beautiful idea. A universe that always existed, bouncing eternally between higher-dimensional sheets, no singularity, no inflation, entropy neatly reset on every cycle. A genuinely good idea.

Good ideas, unfortunately, are not the same thing as right ideas. And ekpyrotic theory has a list of troubles, starting small and ending with a problem it simply cannot talk its way out of.

Let's begin with the singularity, the very thing ekpyrotic was supposed to abolish. When the two branes collide, the math describing that instant gets extremely uncomfortable. The theory's defense is essentially a hopeful one: that the exotic machinery of string theory will smooth everything over at the moment of the bounce and keep the dreaded infinities from rearing up. Maybe it does. But "we're fairly confident the string math will work out" is a promissory note, not a proof, and an awful lot is riding on it.

Next, that elegant trick with dark energy. Remember that the whole cyclic mechanism hinges on dark energy eventually shutting off, so the branes can stop drifting apart and start falling back together. We have not the faintest evidence that dark energy does any such thing. As far as anyone can actually measure, dark energy is a steady, constant feature of the cosmos with no expiration date printed anywhere on the label. Ekpyrotic theory needs it to quit. The universe has made no such promise.

Then there are the branes themselves. String theory, left to its own devices, has a real appetite for branes. Like zombies, once you allow a couple of them onto the stage, it becomes very hard to stop more from shambling in. Many versions of the ekpyrotic scenario end up populated with whole stacks of branes, which drags the theory straight back into the very multiverse swamp that inflation got bogged down in. The thing it was supposed to do better, it ends up doing about the same.

Now, you can hand-wave your way around all of that. Cross your fingers on the singularity, assume dark energy behaves itself, wave off the extra branes. Cosmologists are nothing if not resourceful. But there is one obstacle no amount of clever hand-waving gets you past, and it happens to be the same standard that crowned inflation in the first place: what we can actually observe.

Inflation, recall, makes a very specific prediction about the seeds of cosmic structure. Not just that the seeds exist, but their detailed statistical properties, the precise mix of large and small ripples sprinkled across the early universe. We can't watch inflation happen directly, but we can read those seeds straight off the cosmic microwave background, the oldest light in existence and the earliest baby picture of the cosmos we will ever get.

Ekpyrotic theory has to play the same game. You cannot call yourself a serious cosmology without making a prediction for those same statistical properties. So ekpyrotic, too, says something concrete about what the microwave background should look like. And the original version of the theory, worked out in the early 2000s, was off. Not a little off. It was "how did you even find my office to tell me this" off, badly out of step with what we observe. Defenders found ways to patch it, the kind of adjustments a less generous person might call hacks, twisting the model until it produced the right sort of fluctuations.

And here is the killer. Even the patched-up version doesn't survive.

Ekpyrotic theory was originally tuned to match the rough, blurry measurements of the microwave background that were available back in the early 2000s. It was built and refined to fit the data of its day. Then we launched Planck, a spacecraft devoted to mapping that ancient light in exquisite detail, and it handed us measurements vastly sharper than anything we'd had before. Those measurements lined up beautifully with what inflation predicted. They did not line up with ekpyrotic theory. The sharper our picture of the infant universe became, the worse ekpyrotic looked and the better inflation held up.

There are not many escape routes from a result like that. The theory isn't dead, exactly. But it is marginalized. Few people work on it now. It carries too much baggage: the math is brutal to wrangle, it doesn't fully or cleanly deliver on its grand promises, and it stumbles on the one observational test that matters most. It is a good idea. But good ideas are not guaranteed to be right ones. That call belongs to nature, not to us, and nature, so far, has voted against it.

So the theory putters along, with a few stubborn researchers poking at it here and there. Maybe because it is still a genuinely fun idea. Maybe because it might yet bear some unexpected fruit down the line. Stranger things have happened in physics.

But for now, the Big Bang and inflation agree with every observation we can throw at them, and the dream of an eternal, cycling, self-renewing cosmos remains exactly that. A dream.

All the evidence points the same hard direction. This one universe, this single shot, is all we get.

So make the best of it.