Why walkers sometimes turn left

A study conducted in Spain and Japan finds an average bias towards the left in walking experiments. It links this to individual tendencies, without identifying their origin or presenting it as a universal rule.

When volunteers walk freely in an open space or an enclosure, their paths are more likely to rotate anticlockwise. A study published in “Nature Communications” in 2026 locates the origin of this bias in individuals rather than in crowd interactions. It identifies a statistical tendency, not a rule dictating which way each person walks — and its precise origin remains unknown.

A rotation that persists in different settings

The finding stems from a simple question: if a group of pedestrians ends up walking in a circle, does the movement come from interactions between its members, walls that constrain them, or an individual preference? The researchers conducted five sets of experiments in Spain and Japan, under various conditions. They tracked groups in enclosures, teenagers in a courtyard without comparable physical boundaries, and people walking alone.

In the group experiments, the average rotation tended towards the left, that is, anticlockwise. The bias also appeared in the Spanish courtyard, despite the absence of an enclosure, and in the trials in Japan. This latter setting made it possible to test an explanation linked to local habits: when walking in opposite directions, Japanese pedestrians tend to keep to the left, unlike the practice observed in the Spanish experiments. Yet the direction in which the groups rotated remained the same.

These comparisons weaken several immediate explanations. Walls alone cannot produce the phenomenon, since it also appears outdoors. Nor does the side chosen to avoid another walker seem to be its main cause. The authors also examined turning preferences, hand, foot or eye dominance, and participants’ sex: they found no statistically significant differences associated with these characteristics.

The crowd does not behave like a single body

To describe the movement, the researchers calculated a measure of rotation at two levels. At the level of an individual, it indicates whether their path curves more anticlockwise or clockwise around a centre. At group level, the individual values are combined into an average. A positive average here corresponds to a collective anticlockwise tendency.

This distinction changes the interpretation. A group average favouring one direction does not mean that everyone walks that way, or that every path is circular. The study finds considerable variability: some people regularly turn in one direction, others in the opposite direction, and still others alternate between turns and straighter stretches. The collective finding mainly reflects a statistical imbalance between these behaviours, not perfect synchronisation.

The individual experiment is particularly useful for separating the hypotheses. More than 200 people were recruited to walk alone in a confined space for 60 seconds. For 156 participants, the data analysed show a median bias towards anticlockwise rotation. The statistical test used by the authors indicates that this tendency is difficult to attribute to chance in the sample studied. But a positive median cannot determine which way a particular person walks: it summarises a distribution in which paths remain varied.

The researchers also compared subgroups. The results show no statistically significant difference according to dominant hand, dominant foot, dominant eye or sex. This does not prove that these factors never have any effect: some subgroups, particularly left-handed participants and those with a dominant left foot, were small. The reasonable conclusion is more limited: in these experiments, the factors tested do not explain the observed bias.

An individual bias, but not yet its mechanism

The study’s strength lies in comparing several experimental settings and analysing individual paths. The group experiments show that anticlockwise movement recurs; those conducted with people walking alone indicate that one need not assume pedestrians coordinate with one another for it to appear. The authors therefore interpret the bias as being mainly linked to individual locomotor tendencies.

“Mainly” matters. The observations do not reveal the biological or neurological mechanism that might incline some paths in one direction. The authors suggest, as a possible explanation, small sensory or balance-related asymmetries, but do not demonstrate them. A direction of walking measured in an experiment is not enough to identify the underlying cause.

The participants were healthy children, teenagers or young adults. The results therefore do not say whether the same tendency is found among older people or those with mobility difficulties. Nor do they describe every everyday situation: a dense crowd, an emergency exit, signs, obstacles or a specific goal may alter people’s paths. The authors themselves stress that their finding is not a universal law of human movement.

The experiment involving young children also illustrates the importance of context. The researchers reanalysed earlier data collected during a free-running activity in a nursery school. Anticlockwise movement was particularly pronounced there, but the children also seemed to imitate one another and line up more. This case is therefore not equivalent to the evidence for individual bias obtained in the solo walking trials: it shows that social influences can combine with a directional tendency.

Possible applications, not yet validated

Understanding how an individual movement can contribute to collective circulation is of interest to researchers studying pedestrian dynamics. The authors suggest that these findings could help inform the design of railway stations, museums or other busy spaces. If an anticlockwise tendency persists in a given place, the layout of routes could take it into account.

This scenario remains an application to be tested, not a benefit demonstrated by the study. The controlled experiments do not measure whether orienting a museum or railway station in this direction actually improves comfort, flow or safety. Behaviour would need to be checked in real spaces, with their constraints, intersecting flows and visual cues. For now, the most robust contribution is more fundamental: the apparent direction of a collective movement can result from a slight imbalance between individual paths, without the entire crowd moving as one.