From Image of the Day: Evidence of a Past Universe? Circular Patterns in the Cosmic Microwave Background (June 10, 2012), we learn:
The circular patterns within the cosmic microwave background shown above suggest that space and time did not come into being at the Big Bang but that our universe in fact continually cycles through a series of “aeons,” according to University of Oxford theoretical physicist Roger Penrose, who says that data collected by NASA’s WMAP satellite supports his idea of “conformal cyclic cosmology”.
Penrose made the sensational claim that he had glimpsed a signal originating from before the Big Bang working with Vahe Gurzadyn of the Yerevan Physics Institute in Armenia. Penrose came to this conclusion after analyzing maps from the Wilkinson Anisotropy Probe.
If these people are so convinced that there are other universes, why does it seem to them like a “sensational claim”?
These maps reveal the cosmic microwave background, believed to have been created just 300,000 years after the Big Bang and offering clues to the conditions at that time. Penrose’s finding runs directly counter to the widely accepted inflationary model of cosmology which states that the universe started from a point of infinite density known as the Big Bang about 13.7 billion years ago, expanded extremely rapidly for a fraction of a second and has continued to expand much more slowly ever since, during which time stars, planets and ultimately humans have emerged.
There’s been some back-and-forth about this in the journals:
An astronomer friend offers UD this comment,
One of the main problems is that Gurzadyan & Penrose don’t do their probability calculations right. The low variance circles seem to be there in the data, and two independent analyses now agree with that. They might be artifacts, but that’s not obvious. Assuming they’re real, it’s important to know how significant they are. Do the circles they’ve found have “anomalously low” variance as claimed (and predicted by Penrose’s cosmological model)?
G&P err in not taking into account their probabilistic resources. They search > 10,000 different centers around the sky and several radii at each center. They have to factor this into their statistics, but they don’t seem to. It’s analogous to saying that you got 20 heads in a row while flipping a coin and claiming that this is not very likely but failing to take into account that you flipped the coin 10^5 times, in which case it’s not terribly unlikely.
In their rejoinder, G&P disagree with the analyses of the two response papers. Even if you grant them their point (and I’ve not thought carefully enough about it yet to know what I think), they still need to put numbers on their results that take into account their search method.
In short it sounds like the typical probability scambo that ID theorists address all the time.