The Nextbike service is an amazing addition to the public transport in Marburg. In particular the fact that all university students can use the bikes free of charge motivates its usage. In this article, I will derive a few quantitative conclusions for people who use Nextbikes to travel around Marburg.

Now off to the article. First of all, it is relevant to know where the stations are located. The following interactive map shows the bike stations in Marburg. Click on the markers to see the station name.

When walking through Marburg, though, not only the location of the station is relevant. Most important is the question if one will find free bikes at the station of choice.

I will answer this question in the following. Let me prepare the answer by first showing the same interactive map again but this time with the marker size scaled by the average occupancy of the station - hence, the average number of bikes standing at a station is represented by the marker size. Additionally, the size of each marker is represented by its colour: Yellow denotes large values and violet denotes small values. You can click on the markers to see the station name and the average occupancy.

Apparently, the stations Hauptbahnhof (German for “train station”) and Elisabeth-Blochmann-Platz are two most occupied stations. Let’s investigate that a bit deeper with a chart that shows the average occupation numbers of each station next to each other. The following graph uses the same colour scheme as the previous interactive map.

The Hauptbahnhof is the most popular station by a large margin with well over 10 bikes on average. Behind that, the Elisabeth-Blochmann-Platz and Cafe Trauma/Affoellerwiesen with 9 and 6, respectively. The station Am Plan/Wirtschaftswissenschaften is with around 0.5 bikes on average the smallest station in Marburg.

Now that we know where the stations are in Marburg and which ones are the largest and smallest, respectively, I come to the most important question for a Nextbike cyclist:

In order to quantify this question, we look at the probability of a station to be empty over time. Again, yellow denotes large values and violet denotes small values.

Unsurprisingly, the smallest bike station in terms of average occupation, Am Plan/Wirtschaftswissenschaften, is also the one that is often empty: Since the station only offers a small number of bikes, the station easily runs out of bikes. Hence, it might be worth to increase the number of bikes to supply the demand there.

The careful reader will notice that the last two charts are almost identical but mirrored. That means that the small stations are typically more prone to running out of bikes and that larger stations, such as the Hauptbahnhof or other large stations, do not have the issue of running out of bikes.

Therefore my advice to Nextbike users in Marburg: When in dire need for a Nextbike, try to walk to one of the larger stations that are shown above as they won’t disappoint you with a lack of bikes to rent.

To test the hypothesis that larger stations run out of bikes less frequently, I plot the probability to find the station empty against its average occupancy.

The points collapse nicely on a curve which reveals that, indeed, smaller stations are more likely to run out of bikes and vice versa. The only two marginal outliers to that are the two largest stations, Hauptbahnhof and Elisabeth-Blochmann-Platz. They seem to be more prone to running out of bikes bikes than the other large stations.

In conclusion, there is quite a number of Nextbike stations spread over Marburg - 36 to be precise. I first investigated the average number of bikes at each station. Then, I asked the question how likely it is for a station to run out of bikes. This is an important question for Nextbike users. Interestingly, I find the probability that a station runs out of bikes is anti proportional to its average occupation. Hence, Nextbike users should avoid small stations when they desperately search for a bike.

---

This article belongs to a series of articles. This is the list of all articles:

• Cycling in Marburg (1/4): Project introduction and data source
• Cycling in Marburg (2/4): For Nextbike users
• Cycling in Marburg (3/4): For the city council
• Cycling in Marburg (4/4): Machine learning predictions