It just took some treetops to glisten and some children to listen for Bing Crosby to enjoy a white Christmas. Bookmakers are a more cynical bunch, so usually demand to see at least a single white flake fall during the 24 hours of 25 December. The chances of snow at Nature HQ this weekend are diminishing as Britain basks in unseasonably mild conditions — bookmakers put the odds of an official white Christmas in London at about 10 to 1.

If that seems too long a shot, then science offers a way to make the bet more attractive to punters. By combining it with a second bet on an event much more likely to happen, bookies can exploit a psychological tic called the conjunction fallacy. Odds, for example, of a cover of the Rolling Stones’ classic ‘You can’t always get what you want’ reaching the UK Christmas number-one slot are a much shorter 9/4 — it’s one of the favourites. (All odds correct as Nature went to press.) And although logic and statistics tell us that the chance of both events occurring must be lower than either of the single events alone, gamblers routinely fail to recognize that.

Study after study shows that pairing with a dead-cert makes an unlikely wager seem more — not less — likely to happen. And that makes people more willing to put money on an outsider. This logical illusion can explain much fixed-odds betting on sport, including football. Gamblers routinely think there is more chance that West Bromwich Albion will win at Arsenal on Boxing Day (9/1) if the wager is combined with a Manchester United home victory over Sunderland (2/9).

Exactly why this happens is not clear, but it seems that some gamblers play the odds off against each other in their heads, and assume — incorrectly — that the combined chance of the two is an average of the odds, that the extreme likelihood of the second option somehow tempers the outlandishness of the first. It can be an expensive mistake.

With supreme knowledge of the human condition, one might think that scientists would be immune from making rash bets. Not so. This year, astrophysicist Shrinivas Kulkarni has lost a US$1,000 wager on the origins of fast radio bursts, and another astrophysicist, David Wiltshire, has stumped up $200 for a lamp after losing a 10-year wager with a colleague on the role of the cosmological constant in dark energy.

Bookies can exploit a psychological tic called the conjunction fallacy.

Although such simple bets between researchers (sometimes friendly and sometimes not so) are a long-standing feature of science, perhaps the most lucrative are those in which scientists (just like bookmakers) pit their calculated professionalism against the optimism and emotion of those who follow a lost cause. This year has also seen climate scientists cash in on bets made with sceptics about the continued warming of the planet. Indeed, the annual meeting of the American Geophysical Union last week had a session dedicated to betting on climate change.

Bets accepted or refused can be a good way to gauge how firmly a sceptic truly believes their contrarian position, because wagers typically follow strongly and honestly held (however unlikely) opinions. (Hence, some fans of Sunderland will see the odds of 14/1 on them winning the above match as too good to turn down.)

Some events baffle punters, scientists and bookmakers alike — and 2016 has seen plenty of those. So who would dare to argue that a theoretical ‘social physics’ model — used, among other things, to predict the behaviour of plastic crystals — would do a worse job than pollsters and experts at predicting the results of political votes such as this year’s Brexit referendum and US presidential election? Physicists last month published such a mean-field model, which they say describes the dynamics of two-group conflicts on the basis of the interactions between group members, opponents and how willing people are to change their minds (H. T. Diep et al. Physica A 469, 183–199; 2017).

The model’s output shows whether each side in a political dispute will tend towards negotiation or conflict, and the often wild swings and oscillations in their attitudes towards each outcome along the way. It is not a tool of prediction, the physicists caution, but rather one of anticipation for strategic purposes. That seems a sensible approach given recent events, which have shaken faith in predictions of all sorts. So in that spirit, as 2016 draws to an end and as Bing almost sang: your days may be merry and bright, and all your Christmases may be white.