Left-handedness is associated with greater fighting success in humans

Left-handedness is a cross-culturally universal, heritable phenotype in humans1 that is thought to be associated with fitness costs2,3 (reviewed in4, but see5). Typically around 11% of the population favour their left-hand6 and though exact numbers vary with culture7, left-handers are always a minority. Since left-handedness is under direct negative selection, its persistence in humans is an evolutionary puzzle.

One explanation for the persistence of left-handedness is the fighting hypothesis8. This argues that the polymorphism in human handedness is maintained due to a negative frequency-dependent advantage that left-handedness confers to males in combat (see9 for theoretical support, and10 for a review of empirical evidence as well as alternatives). According to this theory, right-handed males lack experience fighting rare left-handed males, while left-handed males accumulate plenty of experience fighting right-handed males, putting them at a selective advantage. Combined with the intrinsic fitness costs of left-handedness, this would explain the universal pattern of low but stable levels of left-handers in all studied populations. There is mounting evidence that intrasexual contest competition such as fighting has been a key component of sexual selection on human males (11,12 reviewed in13,14). Modern males may possess adaptations to assist them in fighting and assessing opponents’ fighting ability15. Handedness could therefore be considered a sexually selected trait in males, and may be expressed in females a by-product9.

Consistent with the fighting hypothesis, there is a wealth of evidence that left-handers are overrepresented in combat sports. Sports are particularly relevant systems for testing theories based on intrasexual competition, as they are thought to have evolved culturally as a display for males to advertise fighting and competitive ability16. Overrepresentation of left-handers has been seen in boxing17,18,19, mixed martial arts or MMA20,21,22,23, wrestling24, Judo25, and Karate and Taekwondo26. Left-handers are also overrepresented in sports such as tennis, where they show an advantage over right handed players27,28. Crucially, this pattern has been observed only sports requiring direct interaction with an opponent29,30. As they are rare, left-handers may gain an advantage because their actions are more difficult to predict31,32,33, perhaps due to attentional biases towards the right hand of an opponent34, which in combat sports is typically used for power strikes by a right handed fighter. If left-handed men are disproportionately successful in combat sports when they are rare, it is not unreasonable to assume they would also be successful in ancestral environments where physical violence and competition were likely much more common than today14.

Studies of the fighting hypothesis in martial artists typically do not find that left-handed fighters are more likely to win fights (e.g.20, but see19). However, previous studies have often used small sample sizes (e.g.17) or only assessed the very best members of a particular sport (e.g.,19,24). Any advantages are likely to be small as a large advantage would lead to an increase in the frequency of left-handed fighters until the advantage exactly offsets the costs of being left-handed, which may be small in populations with access to modern healthcare2. Thus, detecting the effect of left-handedness on fighting success may require very large sample sizes. Likewise, top fighters by definition have little variance in fight success, making detecting relationships in these datasets difficult. Top fighters may also have encountered enough left-handed opponents that any advantages due to unfamiliarity would be diminished. Evidence for whether left-handed fighters perform better than right-handed fighters is thus inconclusive. The present studies tested whether left-handed fighters are better than right-handed fighters in 3 large samples consisting of professional fighters at a variety of ability levels. In particular, one of our samples comprised the majority of boxers professionally active at the time of writing.

Previous studies also used win percentage records, number of wins, or ranking from a single tournament as proxies of fighting ability. These may fail to capture long term fighting performance, particularly for fighters with 0 losses, (which gives a win percentage of 1 regardless of the number of fights). These metrics also do not weight wins by quality of opponent, and fail to include how fighters beat their opponent. For example, winning a boxing match by having a better judges’ score after 10 rounds may indicate less formidability than a win by knockout in the first round. In our samples we excluded fighters who had few fights, and additionally compared left and right-handed boxers using their BoxRec score. The Boxrec score is a comprehensive measure of fighting ability, where the points awarded for winning fights are weighted by the swiftness and manner of victory (e.g. knockout or judges’ decision) as well as the skill of the opponent. This addresses the concerns with use of win percentage as a measure of fighting ability. See http://boxrec.com/media/index.php/BoxRec_Ratings_Description for a description of how a BoxRec score is calculated.

The fighting hypothesis for the evolution of left-handedness is based on male-male contest competition, but there is no reason to expect the frequency-dependent advantage of left-handedness in combat to be confined to males. However, there have been almost no studies of the success of left-handed female fighters. To remedy this, one of our samples consisted exclusively of female professional boxers and our sample of MMA fighters included women as well as men. Additionally, comparison of the left-hand advantage in male and female fighters allows us to investigate negative frequency-dependence. If there are fewer left-handed female fighters than male ones, the fighting hypothesis might predict left-handed female fighters would have a larger advantage.

Lastly, a previous study by Dochtermann et al.22 demonstrated that left-handed MMA fighters show greater variance in probability of winning a fight than right-handed fighters. They argue that this is because the advantage left-handed fighters possess increases the probability that they will reach professional level compared to right-handers even if they are less skilled. We attempted to replicate this finding in our samples.

In summary, we investigated representation and fighting success of left-handers in 3 of the largest samples tested thus far, consisting of professional male and female boxers and MMA fighters of varying abilities. For boxers, we also tested the difference between left and right-handers in BoxRec scores, a holistic measure of fighting ability. Our study provides one of the most powerful tests of the fighting hypothesis attempted to date.