Citi Bikeshare Analysis


Overall Ride Patterns

The graphs below show the total amount of rides that were catalogued by Citi Bikeshare in March through June 2019 and for March through June 2020.


The number of rides picks up as the time period goes on in 2019, likely due to improving weather patterns more suitable for biking. Compared to this 2020 line-graph of the same time period, we can see that the overall number of rides is much lower than the previous year, particularly during April which traditionally is a popular time to ride bikes in DC. Compared to this 2020 line-graph of the same time period, we can see that the overall number of rides is much lower than the previous year.

If we compare riders who ride during traditonal commuter hours versus noncommuters, we see that although noncommuters are more frequent in both, the spread is larger during the pandemic. Note, that pandemic rider types are down across both categories.

Next comparing riders on weekends versus weekdays, we see that weekday rides are more popular for both groups. However, unlike the commuter hours graph, the spread is not as large in the pandemic time period.

Next comparing the hour in which each ride started we see that nonpandemic has a bimodal spread which roughly matches to commuting hours. Pandemic has a slight peak for the morning commuting hours but has a greater share of rides around noon and early afternoon.

These graphs, displaying how many rides occurred on each day of the week, show that during the nonpandemic time period the days did not have a lot of spread between the number of rides occurring. The pandemic ride numbers show that many more rides are happening on the weekend than previously.

Next, looking at the differences in riders between these two time periods. Comparing these two pie graphs which show the proportional breakdown of members vs casual riders, we see that 2020 has a higher percentage of causal riders.

These graphs show the total amount of rides at the different stopping and starting stations for the two time periods. In the nonpandemic map we see that stations near transport hubs like Penn Station and Grand Central Station are the most frequently used stations. For the pandemic maps, the transportation hubs are no longer hotspots and instead the most frequently used stations are near major medical facilities such as Lenox Hill and Mount Sinai Medical Centers. This may suggest that essential workers at these facilities could partially be turning to Citi Bikeshare to get to and from work.

The animated maps show the amount of rides ending at each of the ending stations for the two periods. Scroll through to see how the patterns change from day to day.

Ride Patterns Compared to Covid

The graphs below show how ridership changed as Covid cases increased in the area.


Despite the regulations imposed, the Corona virus continued to spread. In the combo chart above, we see the relationship between the total rides versus the total number of Covid-19 diagnosed in the area. THe bubble chart shows that although there was an intial drop in rides as the virus cases increased, eventually rides went back up despite increased virus cases

When the city first began shutting down and restricting public transport, we saw a large increase in overall rides. Then as more businesses and services shuttered, the overall rate of riders declined.

Black Lives Matter Protest Related Pattern Analysis

On May 25th, in Minneapolis a man named George Floyd was killed by police officers. Although, Covid-19 related regulations remained in place in NYC and DC, many outraged citizens showed up to demonstrate together. While, residents of NYC did not convene in single large scale protests similar to how DC residents did, multiple diffuse protests occured across the city, including some specifically using bike rides to show solidarity.


This line graph unlike the DC line graphs for the same time period does not show an increase in the overall rides occuring in the days after Floyd's death. However, it is interesting to look at the data in light of the mandatory curfews imposed on the city. These histograms show the distribution of rides each day so a noticeable decrease in rides appears later in the evenings.

This map shows the end stations for May 23, the last Saturday before the death of George Floyd.

It is harder to track specific bike patterns for NYC than it is for DC because NYC also used bike rides themselves as a form of protest rather than as means to get to a specific protest location. Overall, we already know that bike rides have increased by a large amount in this time period, whether this is directly correlated to the protests is unknown. This map showing hot spots for June 7- 25 when a lot of protests were taking place does not show any clear correlations. However, for June 10 & June 19, we can see that locations such as Washington Square Park, associated with protest activities on those days do see a spike.