A lottery with rare jackpots: that's what the gravitational-wave background looks like. Most tickets are duds, but occasionally a huge prize comes up: the merger of two supermassive black holes, sending out a powerful gravitational wave. Such waves are caught by networks of radio telescopes monitoring pulsars, and from them scientists try to map the sky.
Previously, it was thought that you could simply take the average value and judge how 'noisy' the sky is based on that. But it turns out the average is deceptive.
The authors generated 700 000 random sky realizations and found that the typical shot-noise value is 2–3 times lower than the mean, and at high frequencies the mean is completely off the charts. So the real sky will almost certainly be quieter than predicted by mean-based estimates.
This matters for future observations: if we expect a signal calculated from the mean, we might not see it and draw wrong conclusions. The lottery goes on, and now we know the odds are unevenly distributed.
The median shot noise at 1 yr⁻¹ reaches 0.05, which is only four times below the current NANOGrav upper limit of 0.20—future observations may detect background anisotropy for the first time.