The past, present and future of prediction

Our US readers will be aware that April is Mathematics Awareness Month (MAM) – and this year’s event has as its theme “The Future of Prediction”. Making predictions is tough, as Yogi Berra once observed.* But sometimes they pay off, especially when predictions are grounded in robust theory and mathematics.

Gravitational waves are a case in point. First predicted a century ago by Albert Einstein as part of his general theory of relativity, these ripples in spacetime were detected last year, and their discovery announced to the world in February.

Confirming Einstein’s predictions was largely the job of physicists and technicians. However, statisticians did their bit too. As discussed by Renate Meyer and Nelson Christensen, it was through Bayesian inference that we were able to learn more about the phenomena that created these gravitational waves.

As a demonstration of the power of prediction, the confirmation of this discovery at this time was no doubt welcomed by MAM’s organisers (among them the American Statistical Association): the news will hopefully lend support to their ongoing efforts to increase public understanding of, and appreciation for, the power of mathematics.

Here at Significance, we also wanted to make our own contribution to these efforts. This month’s cover story takes a broad look at the science of prediction, how it has evolved and where it might be heading. Throughout April, the article will be free to read via the MAM website, so please direct friends, family and colleagues to mathaware.org if you are not quite ready to hand over your copy.

And why would you be? Elsewhere this issue we delve into such weighty issues as flawed forensic science, the sidelining of statistical evidence in civil rights cases, and the gender data gap.

Elsewhere, Nassim Nicholas Taleb explains more about his work to understand the risk of future violent conflicts. And, as if to prove that we are not entirely future-focused, our In Practice interview this month features statistician and historian Stephen Stigler discussing his latest work, The Seven Pillars of Statistical Wisdom.

We hope you find plenty to enjoy, and that there is plenty here to inspire those young statisticians who are thinking of entering this year’s writing competition. Details are online here and the deadline is 28 May 2016.


*Or did he? According to the MAM website: "Yogi Berra, paraphrasing Niels Bohr, said: 'It's tough to make predictions, especially about the future.'" There is some dispute over whether Yogi Berra actually uttered these words, or similar, while the Quote Investigator website says that: "It is possible that Bohr employed the saying, but it is very unlikely that he coined it." Quote Investigator traces the origin of the quote to a 1948 Danish language book, while the earliest appearance in English, located by QI, is in the Journal of the Royal Statistical Society Series A, dated 1956.