My running stats

The beginnings

Let's have a deeper look to the running stats. The good thing about analyzing this sport is the amount of available data I have. I used to run normally in summer periods in order to get in shape for the football preaseason. As someone who has always played with and against older people, one of the aspects to work was my physcal condition.

It was not until end of 2019 though that I started to regularly go for runs. In autumn, I remeber joining a 10 km race (4:28min/km) in San Sebastian and a half marathon a month later in Vitoria Gasteiz (4:14min/km). Any amateur young runner may notice that these paces are not outstanding, but back then they felt great! I have not participated in many races since then, but we will talk about them in future posts.

My first race(s)

It was not until end of 2019 that I started to regularly go for runs. In autumn, I remeber joining spontaneously a 10 km race (4:28min/km) in San Sebastian. A month later, I ran a half marathon in Vitoria Gasteiz (4:14min/km, see the activity here). Any amateur young runner may notice that these paces are not outstanding, but they felt great back then! I quickly improved my endurance with few trainings that made me stick to the habit!

Exploratory data analysis (EDA)

This is why all the analysis I will present has the data starting from the beginning of 2020. Basically when I came to Grenoble. Here, I included some trail running in my running habits and therefore the elevation gain metric will be present during the analysis. To begin with, let's create pair plots to see the relationships between single variables. I chose the distance, elevation gain, the pace and the time.

The diagonal terms are smoothed histograms for each variable, while the scattered plots link different ones. We can see for instance that the longer I go, the lower the pace is. Similar to the elevation gain: my faster workouts are those without hills. Also, we notice that the vast mojority of my runs are lower than 2h.

My running distance

Cumulative, distribution and frequency

Let's start looking at how my running distance evolves during a year. This is called a cumulative plot, where we see how the kms add up. The variations in the slopes refer to the more/less running I have done.

For instance, we clearly see the flat part in 2020 related to the lockdown. Moreover, we can notice that both in 2021 and 2022 there is a small plateau. This happend when I twisted my ankles. Up to now in 2024, I am have run less than in 2023 but still keeping quite a bit of running volume. For instance, 2023 was not a great winter in terms of snow, which made me load the running kms.

This is a boxplot that shows how the distances of my runs have evolved during the years. We can clarly see how the length of the runs has decreased. For 2024, there is only the data until May, which can be the cause of the lack of longer runs. Moreover, the back and forth runs to the gym are recorded and included (5-6 km in distance).

Let's put some numbers here. Until 2023, the median distance of my runs has been 10.4 km, while it has dropped to 6.8 km during 2024. It is true this data is only counting the first months of 2024, but the change in my habits is considerable.

We can also compare how the monthly running distance has evolved with the years. The variations are often linked to seasons or biking trips that I may do.

More globally, we can also analyze how the number of runs per month has evolved during the years. We can notice a clear overlap between the frequency and the distance, except in 2024. This is related to the shorter but more frequent sessions that I have done so far. This is in line with the boxplot I have shown above, where we could see how median of my runs has decreased.

Soon more!

Page updated

Google Sites

Report abuse