The 2025 Chicago Marathon
Summarizing the 2025 Chicago Marathon
The "2025 Chicago Marathon" was an absolute blast! 53,000 people from around the world competed in one of the fastest courses in the world! I completed it in a graceful 2 hours and 48 minutes, with splitting halves of 1:20 and 1:28, the latter half stopped by the infamous wall! Featured Instragam pic and Strava embed below.
Data Analysis - Distribution
With Python and the requests module - we scraped the entire sets of results. GitHub repo with the code located here.
First - the entire dataset including all male and female runners:
Next - only the males:
Finally the females whose distribution appears more symmetric than the males. Another observation is that while the mean, min and max are reasonably different (measures of location) - the measure of spread in the standard deviation is almost identical:
Data Analysis - Marathon Splitting
Some marathoners attribute their run as a "positive split" or "negative split". If the first half of the marathon was faster than the second - then the run is a positive split. Else, it is a negative split. Intuitively - a marathon is so long that one would expect the second half to be slower than the first. If a runner does the opposite then there is a clear strategy to conserve energy and finish strong. Here we investigate the success of this strategy among males. Define $$S = \text{Second Half Marathon Time}-\text{First Half Marathon Time}.$$Then if $S$ is positive - the second half of the marathon was longer than the first. Below is a histogram of $S$. An expected result, though, illustrates that a negative split is not common, making up about 10% of all cases.
Next we investigate the linear association between $S$ and marathon finishing time. A scatterplot and the regression results below. There appears to be a positive correlation between spread and marathon finishing time, implying, the smaller the spread (in the $\mathbb{R}$ space) the smaller the finishing time! Therefore, the best runners are those who are bringing $S$ down. What we can also see is that the best runners are solving for $ -5 \leq S \leq +5$.
Looking forward to the next marathon!