Media Briefings

LIMITED HUMAN ATTENTION DRIVES RISKIER BEHAVIOUR: Evidence from World Cup alpine skiing

  • Published Date: March 2015

People typically focus on the leftmost digit of a number and pay only partial attention to other digits – and in World Cup alpine ski events, this makes them liable to take far more risks. According to research by Lucas Schmid and colleagues, skiers with an opening-run time difference to the leader just below a tenth-of-a-second threshold are significantly more likely to crash in the second run.

Their study, to be presented at the Royal Economic Society’s 2015 annual conference, analyses data on 1,865 athletes in World Cup alpine skiing to explore the effects of left-digit bias when analysing the time difference to the leader after the opening run on the riskiness of their strategy for the second run.

The result suggest that athletes consider distances such as nine hundredths of a second much smaller than ten hundredths and pursue a riskier strategy, which increases their probability of crashing by up to 28.5%. Moreover, the differences of race times in the second run increase by 26.1%.

Both numbers reflect the increased risk-taking, when athletes pay limited attention to right digits. The result does not change even if only older, more experienced athletes are examined or races with large prize money.

The authors believe that the athletes’ example is useful for understanding the causes of risk-taking behaviour in general: ‘We find that irrespective of an individual's genetics, experience or what’s at stake, the way of processing information shapes behaviour under uncertainty.’

More…

This study shows that limited attention affects individual risk-taking behaviour of professionals. The authors document that athletes in the Alpine Ski World Cup suffer from a left-digit bias, an empirical regularity that people tend to focus on the leftmost digit of a number and pay only partial attention to other digits.

In particular, athletes with a seemingly small distance to the current leader are 28.5% more likely not to finish the race. This seems to be the first real-world evidence that limited attention not only affects consumer behaviour but also has consequences for individual risk-taking behaviour.

The study investigates the presence of a left-digit bias by using data on 1,865 athletes in World Cup alpine skiing over the period of 1992-2014. The empirical analysis exploits the fact that slalom and giant slalom races consist of two separate runs. After the opening run, each athlete obtains information about her own time as well as her distance to the current leader.

The researchers explore whether athletes exhibit a left-digit bias when processing this time difference to the leader. In particular, they test whether the use of heuristic thinking affects the way athletes choose their risk strategy for the second run.

In the presence of a left-digit bias, they show that athletes misinterpret distances such as nine hundredths of a second to be significantly smaller than, for example, ten hundredths of a second. This behavioural bias in turn leads to the adoption of a more risky strategy because achieving the great success (winning the race) appears to be more likely if the gap to the current leader is small than if it is large.

In the empirical analysis, the researchers exploit the fact that the right digits in athletes' time distance to the leader are quasi-randomly allocated. In other words, whether an athlete trails the current leader by, for example, nine or ten hundredths of a second is beyond her control.

The results reveal that individuals with an opening-run time difference to the leader just below a tenth-of-a-second threshold are significantly more likely to crash in the second run. Risky behaviour increases the probability of not finishing the race by up to 28.5%. Moreover, the variation of race times in the second run increases by about 26.1%. This reflects increased risk-taking, too.

As expected, the effect is present only among athletes close to the leader after the first run, thus having a plausible chance of winning the race. These results are consistent with the theoretical prediction that athletes receive a signal about their time distance to the leader and pay only limited attention to right digits.

In contrast to previous research, this behavioural bias does not disappear when restricting the sample to older, more experienced athletes. On the contrary, the adoption of risky behaviour is large and significant even among athletes aged 25 and older.

Furthermore, there is no evidence that the left-digit bias is smaller in races with particularly large prize money. The researchers argue that their results can be explained by the concept of ego-depletion. The particularly high stakes in World Cup competitions cause athletes to exert extreme effort during the race, making them vulnerable to behavioural biases afterwards.

The findings have implications beyond alpine skiing because they help to understand the causes of individual risk-taking behaviour. Irrespective of an individual's genetics, experience or what is at stake, the way of processing information shapes behaviour under uncertainty. Hence, limited attention is likely to have implications more far reaching that suggested by previous research.

ENDS


Reto Föllmi, Stefan Legge and Lukas Schmid
email: lukas.schmid@unisg.ch
Phone: +41 788 982 273