Marathon Demographics of the Top 2600 Finishers at the 2018

Posted on Jul 29, 2019

Project GitHub | LinkedIn:   Niki   Moritz   Hao-Wei   Matthew   Oren

The skills we demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.


The New York City Marathon is an annual event, which draws hundreds of thousands of marathoners from across the world. It is world’s largest marathon, which courses through all five of New York City’s boroughs (Wikipedia), and is a major event on the city’s calendar. People train for years to participate in it, and the most recent marathon was held in November 2018 and nearly 100,000 people applied to participate, and over 52,000 people finished the race (NYRR website, Wikipedia).

For this project, I scraped the official results for the first 2639 finishers[1] from the marathon’s website, and analyzed the demographics of this group to understand the trends and patterns. The results provide information for each finisher on the following:

  • Age
  • Sex
  • Country of origin
  • The finishing position in the race
  • The time they took to finish (in hours, minutes and seconds)
  • Their pace

I extracted all the information provided, except the results on pace.

I used the Python package Selenium to scrape the information from the website. This was a more appropriate choice than Scrapy, since the webpage has infinite scrolling, and the button to show more records needs to be clicked repeatedly in order to load new records to the page. All data manipulation was done using Pandas, and all graphs were generated using Seaborn and Matplotlib.

Data cleaning and recodes:

Some amount of data cleaning was inevitable, since the country of origin information for US runners also included their city and state. For all other participants, the country of origin was provided as a three-letter country code such as ITA (for Italy), GBR (for Great Britain), ETH (for Ethiopia) etc.[2] I therefore changed all fields that had a length greater than three to ‘USA’. Any duplicates that were included in the dataset after the scraping process was completed were also removed. There were no missing values in this dataset, and therefore did not need to be handled.

I also selected subsamples for the top 50, top 25, top 10 and top 5 finishers to understand the demographics of this elite group of finishers.  Other variables such as the total time in hours (called ‘duration’ in this analysis), taken by each runner to complete the race, were created as needed by combining and recoding the variables that were scraped.


Total participation:

People from 71 different countries finished the race in the top 2,600. The finishers ranged in age from 18 to 63, while the median age was 37 years. The United States had the largest group of finishers in this category with 1,154 finishers (Fig. 1), reflecting the fact that it had the largest group of participants in the race to begin with – the marathon’s website indicates that there were 28,310 Americans who ran the marathon and finished it.

After the United States, Italy had the next highest number of participants in the race, as well as among the finishers in the top 2600 (Figs. 1 and 2). Many countries had only 1 participant, and due to the long tail in the distribution of finishers by country, I also created a graph to look at the 20 countries with the largest number of finishers in the top 2600 (Fig. 3).

East Africa rules:

Interestingly, though not perhaps surprisingly, relative to the total number of participants in the marathon, participants from three East African countries – Ethiopia, Kenya, and Uganda, outperformed everyone else. Ethiopia had a total of 19 people[3] participating in the marathon and 16 were in the top 2600. Kenya had 25 total participants and 8 were in the top 2600, while Uganda had only 2 participants, and 1 of them was in the top 2600 (Fig. 4). However, not only were the East African runners in the top 2600,

  • About 32% or 16 finishers in the top 50 were from East Africa (Fig. 5). This meant that 64% of all East African finishers in the top 2600 were in the top 50.
  • Among the top 25, this percent went up to 52% (about 13 finishers) (not shown)
  • Among the top 10, five finishers or 60% of all finishers were from East Africa (Fig. 6)
  • Among the top 5, all five, that is 100% were from East Africa (Fig.7)

The boxplot showing the distribution of time taken by East African runners to finish the race, compared with runners from other countries shows the difference in performance between the East Africans and the others (Fig. 8). While the median time taken by East African runners in the top 2600 to finish the marathon was 2.34 hours, for others it was 3.02 hours.  

Gender disparity in the top 2600:

There were fewer than 300 women among the top 2600 finishers (Fig. 9). While the proportions of women runners varied by country, Lithuania and Belarus had the highest percent women in the top 2600 (100%) (Fig. 10). However, since this was due to the overall low number of participants from these two countries, it is also important to examine the numbers of women finishers by country. This shows that the US had the highest number of finishers among women (188), followed by Great Britain with 10 women (Fig. 11 and 12).

However, the results by time taken to complete the marathon are more encouraging. While the median amount of time taken by men in the top 2600 to complete the marathon was about 3 hours, the median for women was 3.1 hours. The boxplot shows many outliers for men, while there was less dispersion for women (Fig. 13).

Questions for the future:

The results and trends presented in this post only skim the surface of the trends that could potentially be uncovered by analyzing the data from the NYC Marathon. Since data are also available for previous years, it would be interesting to examine trends over time along various dimensions, but especially for women’s participation in the marathon, the numbers of women who finished in the top 2600, and changes, if any, in the median time taken for completing the marathon.  

[1] The first 2639 finishers were selected purely for convenience since scraping the results for all 52,000 finishers would have been a time-intensive process.

[2] Since the three-letter country codes used by the organizers of the NYC Marathon are identical to the ones used by the Olympics Federation, they can be looked up online in the event of any ambiguity.


About Author

Leave a Comment

Gregory Lepore October 18, 2019
Any chance you could post your Python code for scraping the results? I would like to use the code to help extract runners from our local running club who run New York.

View Posts by Categories

Our Recent Popular Posts

View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI