Women’s Soccer Powers

By | February 4, 2019

Reading about the dominant women’s soccer team in In a League of Their Own!: The Dick, Kerr Ladies Football Team made me consider what makes certain countries strong in women’s soccer. At the time, it seemed that England had a huge advantage over the US in women’s soccer and probably over everyone in the world. However, now the US Women’s Soccer team is now considered the preeminent power in the world. I am certain that there are innumerable factors that went into why the US is dominant and why other nations that one might expect to be dominant are not. There are certainly thousands of decisions and pivotal moments of which I am completely ignorant. However, I attempted to identify three main quantifiable factors that I would think could influence the development and power of women’s soccer in a country. I came up with the level of rights granted to women in the country, the size of the country and the skill and power of men’s soccer in the country. I then chose a metric to simplify each of these factors.

 

First, to quantify how strong a women’s national team is, I used their FIFA ranking. Due to time constraints, I chose to only look at the top 50. To attempt to quantify the degree to which women are empowered in each country, I found a US News and World Report list that ranked the top 80 countries for women to live. I understand that this is highly subjective and at best a crude way to quantify this complex concept, but I think that it at least gives an indication as to where women are more empowered. To quantify the size of the country, I used the rank of the country in terms of population. Again, this is a crude approximation as the 40th country is not necessarily twice as big as the 80th country, but it at least gives an indication about the relative size of the countries and how it might correlate with women’s soccer skill. Lastly, to quantify how powerful the men’s soccer team is I used the FIFA men’s rankings. While the rankings are transient and do not necessarily convey how much of a “soccer country” a country is, they at least give an indication of which countries have a richer history in soccer (France and Brazil are near the top; Canada and Australia are not).

 

I then plotted each of these rankings against the FIFA women’s rankings to see how well they correlated.

 

 

As is shown, the ranking for empowerment of women correlates most strongly with the strength of the women’s national soccer team. I calculated r2 values to help quantify how closely each of these factors correlated with the strength of the women’s national team. Empowerment of women in the country posted the strongest r2 at .439. Rank of the men’s team was next at .069, followed by country population at .057. This should not come as a shock, as some of the more progressive countries also boast strong women’s soccer teams, even if they don’t have strong men’s programs: USA, Canada, Australia, Japan, Norway and South Korea to name a few.

 

I recognize that there are many limitations to my analysis. The metrics are not foolproof indicators for what I was attempting to measure. Also, I only looked at the top 50 teams rather than all women’s national teams. And my statistical analysis skills are not particularly robust so it is probable that there are some errors or false assumptions. Additionally, I had to throw out nine of the countries because I could not find data for them in at least one of the categories, and decided that rather than risk them skewing one of the metrics, I would throw them out altogether.

 

However, despite the potential shortcomings of the findings, I still think it is interesting to think about and attempt to prove which factors contribute to success in soccer. Interestingly, if you combined all three factors together evenly, it would create a stronger correlation than any individually at .505. I am sure that using some simple computer science, someone could find the optimal combination of all three factors. I would love to hear people’s comments about this, particularly if anyone disagrees with what I did or how I did it!

 

5 thoughts on “Women’s Soccer Powers

  1. Patrick Donley Post author

    Matt- I’d definitely be interested in working on some more analysis on this topic or something similar! It did get a little technical for me but it sounds interesting.

    Reply
  2. Laurent Dubois

    This is a really interesting discussion, and it would be great if you were interested in teaming up to deepen this analysis. It’s true that there is not really enough research on this topic out there so you could produce something very insightful, especially if you are able to think through all the complexities in terms of using statistical analysis to analyze such a complex phenomenon. Great work!

    Reply
  3. Matthew Farrell

    Hey Patrick – very insightful post. I too find sports and data science extremely interesting and will hopefully be able to add similar posts in the future. After looking at Kaggle and other dataset providers, however, I found very little publicly available datasets on women’s soccer, so I really like your approach on attempting to find explanatory variables that predict how good national women’s soccer teams are by using there FIFA rankings. As our classmate, Armin, noted in his comment to your post, Matthew Yeaton from Columbia has written a paper on this very exact idea, and your results confirm what he, and the others in the literature have found.

    One aspect of this study that I would love to be developed upon is looking at factors that affect these rankings in a regression type framework. Yeaton using principal components analysis (PCA), and without getting too technical, is a statistical method used to find the most power explanatory factors that affect a certain dependent variable. However, a major limitation of PCA is that it does not specify exactly what the factors are (think about finding an x variable that explains y very well but without having a label and therefore knowing what this x variable is). Yeaton uses PCA to circumnavigate multicollinearity issues with many of the explanatory factors. For example, one would expect women equality to be positively correlated with more advance economies and therefore higher GDP. This multicollinearity then dampens the statistical significance of many of the coefficients used in the analysis. However, what could be very interesting is running a multivariate regression, and instead of trying to prove the statistical significance of each explanatory variable, to use a joint F-test to find the joint statistical significance of the model as a whole.
    For example, if one runs the regression:

    Y_i=β_0+ βX_i+ ε_i

    Where Xi is a vector of explanatory variables (i.e. country size, men’s ranking, gender equality proxy, etc.) and Yi is the FIFA women’s ranking. By then looking at the joint significance of these variables, we could develop a strong theoretical model that accurately explains the variation in women’s soccer rankings.

    I know this comment has gotten a little bit technical, but I would absolutely be willing to work with you and this data in the future to develop more conclusions on what really affects women’s soccer rankings. There doesn’t seem to be too much academic literature on this topic so most of the contributions here would be relatively novel. Overall, super great and interesting post that I could easily see turning into a very effective and compelling research question.

    Reply
  4. Jacqueline Allain

    Patrick and Richard–thank you for doing so much outside research on this topic. You have given us a lot to think about. I wonder if these conclusions hold true for other sports.

    Reply
  5. Armin Ameri

    Reading through this post, I became interested in looking up additional research on the subject. Specifically, I find it interesting how academics utilize big data to better understand seemingly commonplace entertainment like sports. I came to find one study that particularly interested me: Gender Inequality and Women’s Soccer Success: Utilizing Principal Component Analysis to Isolate Inequality by Matthew Yeaton.

    He begins by asserting the importance of studying sports in economic analyses. Sports are a study of a country’s culture, and they measure seemingly invisible phenomena that GDP analyses cannot by themselves measure. Yeaton writes, “by showing that gender inequality has an impact in determining success in soccer, or any other sport, we are showing that gender inequality is real, and that it has an impact in the culture as a whole” (2). He points out that soccer presents a unique opportunity for global analysis given how universally it’s accepted and how standardized the rules are globally. One of the issues Yeaton notes is that much of the academic literature on soccer is centered on men’s soccer, not just because it is the more popular of the two, but also because there’s a lack of data on women’s soccer.

    He then conducts a historical analysis of women’s soccer. Interestingly, much like Gail Newsham’s In a League of their Own: The Dick, Kerr Ladies 1917-1965, Yeaton also focuses on the history of soccer before the first women’s World Cup in 1991. He notes that from the 1930s onwards, women’s games were drawing crowds of up to 92,000 people, clearly demonstrating the interest people have in the sport. This shows that, from an economic perspective, there is a lot of money to be made in women’s soccer, and that the sport is less popular than men’s soccer is sign of missed opportunity. Yeaton uses the following data sources for his analysis: FIFA men’s and women’s historical rankings, gender-inequality data, macroeconomic country-level development data, and environmental data, such as average temperature in a country, toxins, and others.

    Patrick’s intuition and analysis are corroborated by Yeaton’s study. Yeaton finds that gender inequality is very important to women’s soccer and that the success of a team very much entangled with the level of liberty and rights afforded to women in a given country. These results are rigorously checked for robustness, demonstrating their validity. Ultimately, the relationship between women’s rights and soccer success is a strong one that should not be ignored.

    Yeaton, Matthew. “Gender Inequality and Women’s Soccer Success: Utilizing Principal Component Analysis to Isolate Inequality.” (2012).

    Reply

Leave a Reply to Jacqueline Allain Cancel reply

Your email address will not be published. Required fields are marked *