Relative Age Effect Among Olympic Medalists: Evidence from Ten Summer and Winter Olympic Games held between 2000 and 2018 

ABSTRACT

This study examines the Relative Age Effect (RAE) among 4,453 individual Olympic medalists from ten Olympic Games (five Summer and five Winter) held between 2000 and 2018. We analyze athletes’ birth quarters and ages at the time of competition to assess patterns by gender, event type, and medal outcome. Using descriptive statistics, regression analysis, a Pearson 𝜒2 test, and a logit model, we find that athletes in judged and combat events tend to be younger, while those in skill and endurance events tend to be older. Gold medalists are, on average, younger than bronze medalists and more likely to be born in the first half of the year. These results confirm the presence of RAE at the highest level of sport and suggest that early developmental advantages persist among Olympic medalists. The findings have implications for athlete development systems and elite sport selection criteria. 

Key Words: Athlete Development; Birth Quarter; Elite Sport, Logit Analysis, Pearson 𝝌𝟐 test 

INTRODUCTION

The Relative Age Effect (RAE) refers to the phenomenon in which individuals born earlier in a selected period, typically a calendar year, tend to benefit from developmental advantages over their younger peers within the same cohort.  These advantages may include earlier physical growth, cognitive maturity, and better access to competitive opportunities.  This concept was described by Barnsley and Thompson (3) in Canadian youth hockey, where players born in the first half of the year were disproportionately over-represented.  RAE has since been documented across various sports, including professional baseball (Thompson, Barnsley, and Stebelsky (14)), elite youth soccer (Glamser and Vincent (7)), youth swimming (Costa et al. (5)) and basketball (Werneck et al. (17)).  Extensive empirical evidence over the last three decades has confirmed its presence in multiple athletic and academic domains (Musch and Grondin (11); Patiño et al. (12)). Researchers have also explored alternative approaches to identifying RAEs by comparing athletes’ relative ages at the time of competition (Zetaruk (18) and Longo et al. (10)). Yet little is known about whether RAE endures at the pinnacle of sports performance. 

Many past studies have focused on youth and amateur athletes, where selection systems, age-based groupings, and physical maturation exert considerable influence.  However, less is known about whether RAE persists at the highest levels of athletic achievement.  The Olympic Games, which represent peak international competition, provide a valuable lens to explore whether early developmental advantages have long-term consequences that extend into elite performance.   

The Olympic context introduces additional layers of complexity.  Events vary widely in physical demands, skill development, and peak performance age.  For instance, judged events such as gymnastics and ice skating often feature younger athletes (Zetaruk (18) and Cummins (6)) while skill and endurance events, such as archery, cross-country skiing, and the marathon typically feature older athletes (Longo et al. (10)).  Seasonal differences between Summer and Winter Games, and gender specific trajectories, also warrant attention. 

Although prior research has examined RAE in Olympic contexts, findings have been mixed.  Baker et al. (2) find evidence of the RAE in skiing, snowboarding, and Nordic combined, find no evidence for figure skaters, and report an atypical pattern in gymnastics.  Joyner et al. (9) find evidence of RAE across multiple sports but note variation by gender and season.  Raschner et al. (13) analyzed data from the first Winter Youth Olympic Games and found evidence of RAE in both genders and across strength, endurance, and technique-related sports.  This study differs by focusing exclusively on Olympic medalists – those who reached the highest level in their sport – to determine whether RAE persists not just in participation, but in podium success. 

This study analyzes 4,453 individual medalists from ten Olympic games (five Summer and five Winter) between 2000 and 2018. We classify events into six categories (timed, judged, skill, endurance, strength, and combat), and examine both the athletes’ age at the time of competition and their birth quarter. The central research questions are (1) Are Olympic medalists disproportionately born in the earlier quarters of the calendar year? (2) Does the probability of winning a gold medal vary by birth quarter? and (3) Are athletes’ ages at the time of competition systematically associated with event type, gender, or Olympic season? This study expands the literature by analyzing RAE by event type among Olympic medalists across both Summer and Winter Games. 

METHODS

This study examines 4,453 medalists (gold, silver, and bronze) from ten Olympic Games held between 2000 and 2018 – five Summer Games (Sydney 2000, Athens 2004, Beijing 2008, London 2012, Rio de Janeiro 2016) and five Winter Games (Salt Lake City 2002, Turin 2006, Vancouver 2010, Sochi 2014, PyeongChang 2018).  Data were compiled from official Olympic databases during 2019.  Athlete biographies were consulted to ensure accuracy regarding birthdates, event categories, and medal results.  Medalists disqualified as of December 2019 due to doping violations were excluded from this analysis.  

Athletes were categorized by type of event into six mutually exclusive groups: timed/weight/measured, judged, skill, endurance, strength, and combat. Hilmer and Hilmer (8) apply these same categories to investigate the presence of confirmation bias in judged events at the Olympic Games.  The first category is timed/weight/measured, where competitors start together and medal winners are determined by that individual competition (henceforth referred to, for lack of a better term, as “timed events”), such as the 100-meter dash, canoe, and downhill skiing.  Judged events rely on subjective scoring either fully (ie, figure skating) or partially (ie, mogul skiing).  The next category is skill events such as archery, shooting, and table tennis.  The fourth category is endurance events that take a relatively long time to complete, such as biathlon, cross-country skiing, and the marathon.  Strength is the fifth category of event, which includes weightlifting, shot put, and hammer throw.  The final category of events is combat, which includes boxing, judo, taekwondo, and wrestling.  Team sports were not included in this analysis because we are interested in an individual’s age and birth quarter at the time of competition.  A team is comprised of a variety of individuals with various birth dates, which makes it difficult to isolate the impact of birth quarter and age at the time of competition.  Thus, team events such as soccer, softball, basketball, and relays are excluded from this analysis. Age was calculated in days at the time of competition, and birth quarters were based on the calendar year: Q1 (January-March), Q2 (April-June), Q3 (July-September), and Q4 (October-December). 

Table 1 presents the breakdown of the medal winners for each of the Olympic Games held between 2000 and 2018.  The Summer Olympics have the bulk of the athletes, with 78% of the medal winners, while 22% of the medal winners compete in the Winter Games.  The number of athletes winning individual medals has increased steadily over the years.  Individual sports added to the Olympic Games during this time were skeleton in 2002, BMX racing in 2008, and golf in 2016. 

The dependent variables are either type of medal, gold, silver or bronze, and how old the athlete is in days at the time of competition.  The independent variables are quarter of birth (Q1 = Jan-Mar, Q2 = Apr – June, Q3 = Jul – Sept, Q4 = Oct – Dec), gender, season, and event type (timed, judged, skill, endurance, strength, combat). Table 2 presents the percentage of competitors in the types of events, medals earned, and quarter of birth, broken down by male and female medal winners and Summer and Winter Games.  As evident from Table 2, the timed category has the most competitors with 45% of the medal winners, ranging from 40% in the Summer Games to 60% in the Winter Games.  Skill, Strength, and Combat award all of their medals in the Summer Games.  Judged events comprise 10% of the medals, while skill has 11% of the medals.  The endurance category has 7% of the medals overall but it is an important component of the Winter Games, with almost a quarter of the medals earned falling within this category.  

Under random distribution, one would expect medals to be evenly divided among the three categories. According to Table 2, bronze medals account for 36% of the overall awards.  Similarly, we would expect the athletes’ birth quarters to be split evenly, with each having 25% of the medal winners if there is no presence of RAE. The first quarter has the most medal winners at 26%, while the last quarter has the least amount of medal winners at 23%, which is a statistically significant difference with a z-score of 3.07 and a p-value of 0.0022. 

Table 3 provides means and standard deviations for how many days old the medalists were when they competed in their event.  The average age of a medalist is 26.3 years old with a standard deviation of 4.8 years, with men at an average of 26.57 and with women at 25.94.  This is similar to the finding of Longo et al. (10), who analyzed all competitors from the 2012 Summer Olympics and found men were an average of 27 years old and women were an average of 26.2 years old.  Awosoga and Chow (1) find that the peak age for a track and field athlete is just under 27 years old, that finalists were on average 16 months older than the average competitor, and medalists were just one month older than the average participant. On average, the youngest medalists are those who compete in judged events, while the oldest medalists compete in skill and endurance events.  This holds across males and females and for the Summer and Winter Games. The age of the medalists is distributed fairly consistently between gold, silver, and bronze medals with the gold medalists being around 100 days younger than either silver or bronze medalists for the entire sample.  Males are older than females by 228 days while Winter medalists are older than Summer medalists by 241 days.   

Figure 1 is a kernel density function that depicts the age in days of the medalist by the type of event.  A kernel density function is a non-parametric method for visually representing the distribution of the data. Unlike a histogram, it is a smooth representation of the probability distribution function (Weglarczyk (16)) and is more informative than summary statistics because it shows the entire distribution of the data.  Judged events have the youngest athletes with the mass of the distribution primarily in the lower end of the age distribution.  Endurance has the bulk of its mass to the right of all of the other distributions, while skill events exceeds all of the other events at the very top of the age distribution.  Figure 2 compares the distributions for males and females.  Females have more medalists at the lower end of the distribution but the distributions are nearly identical at the top end of the age distribution.  Figure 3 is a kernel density function for the Winter and Summer Games.  The distribution for the Summer Games lies to the left of that for the Winter Games, suggesting that Summer medalists are younger than Winter medalists.  

RESULTS

Table 4 provides our first look into the presence of an RAE within Olympic medal winners with a two-way table between birth quarter and type of medal.  The Pearson 𝜒2 test statistic for differences among the categories is 14.12 with a p-value of 0.028.  The Cramér’s V p-value of 0.0398 suggests that the observed association between birth quarter and medal type is unlikely to occur by chance.  Taken together, these results suggest that there is a statistical relationship between birth-quarter and type of medal.  The expected count is in parentheses and suggests that gold medal winners are over-represented for the first and second quarters of the year.  All statistical analysis for this paper is performed in STATA.    

Another option for analyzing the birth quarter of a medalist is to empirically assess whether it impacts their probability of winning a gold medal.  To accomplish this, we estimate a logit model of the form 

                      

 

(1) where gold is 1 if athlete i received a gold medal and 0 if they earned a silver or bronze medal, Q1, Q2, and Q3 are the quarter of their birth of individual i, with the fourth quarter as the omitted category, and ε_i is the error term.  The marginal effects are the change in the probability of the athlete winning a gold medal relative to the omitted category 

Table 5 presents the marginal effects from the logit model in equation (1).  Athletes who are born in the second quarter are 4.3% more likely to win a gold medal relative to those born in the fourth quarter at a 5% significance level.  Athletes born in the first and third quarters are not statistically more likely to win a gold medal than those born in the fourth quarter.   

In addition to examining how birth quarter impacts the medal received, we perform an empirical analysis to assess if the age of the athlete, measured in how many days old they were when they competed in their event, statistically differs for gender, type of Games, category of events, and medal type.  The most inclusive model takes the form: 

+  ε_i                            (2)

where ε_i is the error term. Each of the explanatory variables is binary with the value being 1 if the individual has the characteristic in the named variable and 0 otherwise.  For example, the variable male will equal 1 if the athlete is male and 0 if the athlete is female.  The omitted categories for this model are female, Winter, timed events, and bronze medal.  This model is estimated using multiple linear regression with robust standard errors. Because all of the independent variables are binary, this regression model tests for differences in means between the explanatory variables, holding the other included variables constant. 

The first column in Table 6 presents the results for the general model. These results suggest that, on average, males are older than females by 262 days, while Summer medalists are an average of 230 days younger than Winter medalists.  Judged medalists are on average younger than timed medalists by 1090 days, skill medalists are older than timed medalists by 1002 days, endurance medalists are older than timed medalists by 848 days, and combat medalists are younger than timed medalists by 245 days. Gold medalists are an average of 151 days younger than bronze medalists and silver medalists are not statistically different in age than bronze medalists.

The results found in the initial model generally hold for models that estimate male and females separately. The statistical significance for event type for the model with only males is similar to the general model, but the magnitudes differ.  For example, skill medalists are an average of 1,369 days older than timed medalists for the male-only model, while the difference was 1002 days for the full model. The other difference is that gold and silver medalists are not statistically different in age than bronze medalists.  In the female-only model, athletes who medal in judged events are an average of 1,374 days younger than those who medal in timed events, while in the full model the difference was 1090 days.  Female skill medalists are an average of 560 days older than female timed medalists while endurance medalists are 891 days older than timed medalists.  Strength and combat medalists are not statistically different than timed medalists in age.  For females, gold medalists are an average of 225 days younger than bronze medalists. 

Summer and Winter Games models estimated separately follow a similar pattern to the general model in the first column.  In both the Summer and Winter Games, males are statistically older than females, judged medalists are statistically younger than timed medalists, and endurance athletes are statistically older than timed athletes.  In the Summer Games, skill medalists are statistically older than timed medalists and combat medalists are statistically younger than timed medalists.  Summer athletes who win a gold medal are an average of 158 days younger than those athletes who win bronze medals.  Together, these results suggest that the results are generally consistent across males and females as well as Summer and Winter Games.    

Discussion

Our findings affirm the presence of the RAE among Olympic medalists in terms of both birth quarter and competition age.  A Pearson 𝜒2 test for a difference between birth quarter and medals found a statistically significant relationship between the two variables.  We also found that athletes born in Q2 are more likely to win a gold medal relative to those born in Q4.  This echoes patterns identified in youth and elite-level sports by previous researchers (Joyner et. al., 2017; Musch and Grondin, 2001).  These results suggest that the developmental advantages conferred by earlier birth within a competitive cohort persist even at the highest levels of sport. 

The variation in age across event types aligns with existing literature suggesting that events with aesthetic or acrobatic elements, like gymnastics or figure skating, tend to feature younger athletes (Zetaruk, (18) and Cummins (6)), while events requiring cumulative physical or technical development, such as endurance or skill-based events are dominated by older competitors (Longo et. al (10)).  This supports evidence of distinct developmental trajectories across Olympic disciplines.  These findings contribute to a broader understanding of how structural factors such as age-grouping policies and youth sport calendars may contribute to influence athlete development long after initial talent identification.  This finding may support a revision of the youth categorization system and selectors to mitigate the effects of RAE.

We can interpret these patterns using the Developmental Systems Model (Wattie et al., 2015), which posits that RAE arises from interacting individual (e.g., birthdate, maturation), task (e.g. sport type), and environmental (e.g. selection policies) constraints.  Our findings reflect all three of these inputs. From the individual perspective, older athletes may possess more maturity and resilience.  From the task perspective, certain disciplines favor youth, such as gymnastics and figure skating, while other disciplines favor experience, such as equestrian and long-distance running.  From the environmental perspective, qualification systems often reinforce early selection biases that persist all the way up to the Olympic Games.   

This study has several limitations.  Our data only includes athletes who received medals at the Olympic Games, allowing us to examine RAE for those who have achieved the highest pinnacle of their sport.  The broader population of Olympic participants may not exhibit the same patterns as medalists.  Another caveat is that team events and relays were omitted, despite the possibility that such formats may dilute or amplify RAE effects due to different selection or substitution dynamics.  Finally, the analysis does not account for cross-national or cultural variation in athlete development systems, which could meaningfully shape RAE patterns.  Future research should address these gaps by examining a more comprehensive athlete pool, including non-medalists, and incorporating institutional and cultural context.

CONCLUSIONS

This study provides evidence that the RAE persists among Olympic medalists in the Summer and Winter Games held between 2000 and 2018.  Medalists in judged and combat events tend to be younger, while those in skill and endurance events tend to be older, confirming widely held beliefs about athlete development pathways.  Additionally, athletes born in the second quarter of the year are statistically more likely to win a gold medal than those born later in the year, reinforcing the influence of birth timing, even at the elite level.

Our results demonstrate that the effects of age-based selection advantages are not confined to youth or amateur competition but may have enduring implications for performance outcomes at the pinnacle of sport.  These insights underscore the importance of re-evaluating current age-grouping structures in sport development systems.  Policymakers, coaches, and sporting organizations should consider how age-based selection mechanisms might inadvertently limit long-term talent development by favoring relatively older athletes.  By acknowledging and addressing these structural biases, it may be possible to create more equitable opportunities for younger athletes within a given cohort, ultimately enhancing both inclusivity and performance sustainability. 

APPLICATIONS IN SPORT

To mitigate the impact of RAE, sporting bodies and youth development programs should consider pilot programs that rotate cutoff dates or cluster athletes by biological age rather than birthdate alone (see Wattie et al. (15) and Cobley et al. (4)).  Musch and Grondin (11) suggest varying cutoff dates for different sports, allowing youth participants to choose the sport with the most favorable cutoff date for them.  Raschner et al. (13) suggest a limit on the number of participants by each birth year across two-year age groups. Future research could explore how the dynamics of RAE evolve over an athlete’s career trajectory and examine whether similar effects are observable in non-medalists or team events.