May 25, 2022

MBTA Station Prints Quiz 2

Here on the Data Blog, we previously posted a station prints quiz where readers were given daily ridership graphs for anonymous stations and tasked with identifying which station the graph belonged to (we also provided the answers). With new travel patterns emerging because of the pandemic, we would like to put our readers’ subway knowledge to the test again with another quiz. This time, we will be looking at ridership before and during the pandemic, showing average daily ridership prints for both 2019 and 2021. Interesting trends can be found station-by-station, and comparing ridership from before and during the pandemic adds another level of analysis. This kind of study can help us understand how ridership has changed during the pandemic both at different stations and times of day. Edit: the answers to this quiz are now available here!

In our previous post, we reviewed common ridership trends throughout the day, particularly when it comes to peak time periods. Pre-pandemic, we would often see a clear peak in the morning between 7:00 – 9:00 AM, a clear evening peak between 4:00 – 6:30 PM, or both, with a specific spike at “peak of peak” that was generally in one or two fifteen-minute periods in both the AM and PM. The presence of these peaks depended on where the station was located. Stations in the downtown area, such as State, Government Center, Aquarium, and Park Street, typically only had an evening peak since they are near many workplaces. Stations near residential areas, such as Malden Center, Braintree, Davis, and Wonderland, tended to only have a morning peak. Some stations like Central, Maverick, Back Bay, and South Station had both morning and evening peaks. This could have been because there are both people that live in the area and people that work in the area, or because riders enter the station via transfers from bus or Commuter Rail. For example, many Commuter Rail riders transfer to the T at Back Bay and South Station.

Ridership has seen significant changes since the start of the pandemic. There is, of course, lower ridership across the board at all stations and almost all times, but there is a trend of peak periods being less extreme and more even levels of ridership throughout the day at many stations. Furthermore, there is a “spreading” of the peaks at some stations, where the peak is less concentrated around 8:00 – 8:15 AM for the morning period and 5:00 – 5:15 PM for the evening period and is more spread throughout the whole peak range. We can see an example of this for Andrew station. The data shown in the first graph represents the time period of 1/1/2019 – 12/31/2019, and the second graph represents 1/1/2021 – 12/31/2021. The graphs show the distribution of validations by 15-minute periods averaged across all days in the year, and a validation is defined as a card or ticket tapped at a faregate.

Andrew Station, 2019

Line graph showing ridership as validations per 15 minutes. Line first increases at 5 AM and reaches peak at 8 AM with 160 validations per 15 minutes. Then decreases until 10 AM and stays at about 55 validations until 2 PM, then increases to evening peak with 90 validations per 15 minutes from about 4 to 5:30 pm. Line then decreases to 12 AM.

Andrew Station, 2021

Line graph showing ridership as validations per 15 minutes. Line increases at 5 AM and reaches peak at 50 validations per 15 minutes from 7 to 8 AM. Then decreases to about 33 validations from 10 AM to 2 PM. Line then rises to PM peak at approximately 50 validations per 15 minutes from 4 to 6 PM, then decreases to 12 AM.

It can be hard to be sure exactly what’s changing due to the difference in y-axis ranges along with the changing trends in the graph. A simpler way to visualize the changes is to look at the percent of retained ridership in 2021 compared to 2019. The graph below shows validation retention by 30-minute windows throughout the time the T is running, around 5:00 AM to 1:00 AM.

Line graph of validation retention from 2021 compared to 2019 as a percentage. Line starts at 60%, decreases to 30% at 8:30 AM. Then increases to 60% at 10 AM and stays around there until 10 PM. Increases to 95% at 11:30 PM, decreases to 70% at 12 AM, then increases to 90% at 1 AM.

This graph makes it easier to see that the morning rush hour has seen the most significant decrease in ridership, especially the 8:00 – 8:30 AM and 8:30 – 9:00 AM periods. These periods retained just over 30% of their ridership from 2019 to 2021. The adjacent periods of 7:30 – 8:00 AM and 9:00 – 9:30 AM retained slightly more ridership, at 37% and 38%, respectively. The 7:00 – 7:30 AM period retained 45% and the 9:30 – 10:00 AM period retained 48%. This trend creates the U-shape seen on the validation retention graph in the morning rush hour period, and results in a flattening or spreading out of the morning peak period. We can also see a slighter and bumpier U-shape around the evening rush hour. The periods from 4:30 PM through 6:30 PM saw a slightly more severe decrease than the adjacent hours, ranging from 51 – 54%. The rest of the service periods hover around 55 – 70% retention, with high retention in the late evening between 70 – 100%.

This trend of more significant decreases at peak times and moderate retention otherwise was seen across many stations – both outlying and downtown stations for morning and evening peaks. This is likely due to many people commuting to work less often or not at all, especially in white-collar 9-5 roles.

A second common difference between the 2019 and 2021 validation patterns that we saw was, well, a lack of a difference! Although all stations saw overall decreases in validations between the two years, some showed nearly indistinguishable patterns in 2019 and 2021. The best example of this is Chinatown.

Chinatown is a downtown station with a major evening peak and a sustained level of validations throughout the later evening. In the below charts showing 2019 and 2021 validations, we can see that there are some minor differences in the detailed small hills and valleys, but the major evening peak and late evening activity are both still very apparent and don’t seem diminished much compared to the rest of the day.

Chinatown, 2019

Line graph showing ridershipas validations per 15 minutes. Line goes up starting at about 6 AM until 6 PM peak with about 180 validations per 15 minutes, then goes down until 12 AM.

Chinatown, 2021

Line graph showing ridership as validations per 15 minutes. Line increases starting at 6 AM until the peak at 6 PM with about 58 validations per hour, then decreasing.

Line graph of validation retention from 2021 compared to 2019 as a percentage. Line is relatively stable going slightly above and below 40% retention from 7 AM to 12 AM. Lowest retention is at 5 PM with 30% and highest is at 1 AM with 80%.

The validation retention graph for Chinatown is relatively flat and constant throughout the day except during the very late evening and early morning hours. There are dips in the retention at morning and evening rush hour periods, but they are much subtler than the dip in the morning peak at Andrew that we looked at previously.

The last station we’ll be discussing is one that we used as a mystery station for our last station prints quiz due to its unique pattern, Community College. Pre-pandemic, this station had tight validation peaks about every hour and a half that seemed to align perfectly with a class schedule. The 2021 station print for Community College does not tell the same story.

Community College, 2019

Line graph showing ridership as validations per 15 minutes. Increases starting at 5 AM then multiple peaks at 8 AM (60 validations),10 AM (50 validations), 12 PM (60 validations),1:30 PM (80 validations), 2:30 PM (95 validations), 4 PM (90 validations), 5:30 PM (110 validations). Line then decreases until 12 AM.

Community College, 2021

Line graph showing ridership as validations per 15 minutes. 2021 ridership line graph is much smoother than 2019. Small AM peak at about 8 AM with 16 validations per 15 minutes. Line decreases to about 11 validations per 15 minutes from 9 AM to 12 PM, then increases to peak from 2:30 to 3:30 PM with 28 validations per 15 minutes. From here the graph mostly decreases until 12 AM.

The clear validation peaks have disappeared, making Community College station almost unrecognizable from its 2021 print. The loss of the class peaks is likely due to many students attending classes online or in a hybrid format.

The station now has its highest validation period from 2:00 – 3:30 PM. It does still show evidence of a slight morning and evening peak, as well as a small peak around 11:00 PM as it did in 2019. Otherwise, you might never guess that this is the same station! The validation retention graph doesn’t show any sort of trend. It is definitely the “spikiest” graph we’ve seen yet.

Line graph showing ridership retention for 2021 compared to 2019 as a percentage. It stays around 30% from 7 AM to 9 PM, with many little peaks up and down throughout. After 9 PM line increases to over 100% retention at 1 AM.

QUIZ TIME

Now that you are familiar with some common validation patterns and what kinds of changes we’ve seen during the pandemic, let’s see if you can figure out some mystery stations. As a reminder, these graphs represent stations with faregates, so ungated stations on the Green Line branches and the Mattapan Trolley are not possible choices.

Mystery Station 1, 2019

Line graph showing ridership as validations per 15 minutes. Increases starting at 5:30 AM to small peak at 8 AM with about 140 validations per 15 minutes. Decreases to about 70 validations from 10 AM to 2 PM, then increases to peak at 5:30 PM with 340 validations per 15 minutes. It then decreases until the end of the day at 12 AM.

Mystery Station 1, 2021

Line graph showing ridership as as validations per 15 minutes. Line increases starting at 5:30 AM, then very small peak at 7 AM at about 32 validations per hour. Then decreases to about 25 validations at 8 AM and gradually increases to 38 validations at 2 PM. Then increases to large peak from 4 to 5:30 with 178 validations per 15 minutes. Line then decreases to 12 AM with a small peak at about 7:30 PM with 90 validations per 15 minutes.

Line graph showing validation retention for 2021 compared to 2019 as a percentage. Line starts at 66% at 5 AM then decreases to 20% at 8 AM. It increases from here for a small peak at 3 PM with 60% retention. Decreases and stalls at about 50% until about 10 PM, then increases to 85% retention at 1 AM.

In Mystery Station 1, the morning peak period saw a significant decrease in validations as we have seen with other stations, with about 20% retention from 8:00 – 9:00 AM. However, this peak was minor to begin with – the major peak at this station was in the evening, and the validations during the evening peak were actually retained quite well. The validation retention ranges from 49 – 53% between 4:00 PM and 6:30 PM, which is on par with most of the non-peak hours. The other intriguing thing is that the shape of the validations graph is markedly similar throughout the afternoon and evening between 2019 and 2021. The peaks occur at the same exact time periods, and the particular shape of those peaks and their size relative to other areas of the graph is consistent between the two years. For example, there are peaks at 3:30 PM, 4:30 PM, and 5:00 PM, as well as a few smaller peaks at 7:15 PM and 11:00 PM.

Mystery Station 2, 2019

Line graph showing ridership as validations per 15 minutes. Starts increasing at 5:30 AM to large peak at 8 AM with 110 validations per 15 minutes. Then steep drop off, levels off from 10 AM to 2 PM at 40 validations, then a sharp increase to 2 PM with about 135 validations per 15 minutes. From 2:15 PM to 6 PM ridership stays roughly at 80 validations per 15 minutes, then decreases until 12 AM.

Mystery Station 2, 2021

Line graph showing ridership as validations per 15 minutes. Line starts increasing at 5:30 AM to peak at 8 AM with about 28 validations per 15 minutes. Decreases and stays around 23 validations per 15 minutes until 2 PM, then a large increase to about 63 validations at 2 PM. Line stays at about 85 validations until 6 PM, then decreases until 12 AM.

Validation retention graph for 2021 compared to 2019 as a percentage. Line starts at 55% then decreases to 8 AM with about 25% retention. It increases and stays around 50% from 11 AM to 9 PM, the increases until 1 AM with 100% retention.

This station had a large morning peak pre-pandemic that noticeably shrunk in 2021. It also had a set of mid-afternoon peaks that have been retained well at about 50% in 2021 compared to 2019. Hint: This station is on the Orange Line.

Mystery Station 3, 2019

Line graph showing ridership as validations per 15 minutes. Gradual increase from 6 AM to 2 PM, then large spike reaching its peak at about 105 validations per 15 minutes at about 5:30 PM. Then steep drop off, leveling off at about 15 validations per 15 minutes from 8PM to 12 AM.

Mystery Station 3, 2021

Line graph showing ridership as validations per 15 minutes. Line stays below 10 validations per 15 minutes until 2 PM, then increases to about 43 validations per 15 minutes at about 5:30 PM. It then decreases to under 10 validations per 15 minutes from 8 PM to 11:30 PM, then a small spike to about 17 validations at 11:30 PM.

Line graph showing validation retention for 2021 compared to 2019 as a percentage. Starts at 80% at 5 AM, then decreases to lowest point at 35% at about 8:30 AM. Increases to about 50% at 9 AM, then gradually decreases to about 25% at 9 PM. Then increases to 95% at 1 AM.

This station has an evening peak and fairly low ridership. The evening peak had a moderate level of validation retention that was comparable to that of the non-peak service hours.

Mystery Station 4, 2019

Line graph showing ridership as validations per 15 minutes. First increase is at 6 AM and peaks at 8 AM with about 410 validations per 15 minutes. Then decreases until 10 AM at about 60 validations. Increases gradually until 4 PM, then large spike at 6 PM with about 630 validations. Decreases until 12 AM.

Mystery Station 4, 2021

Line graph showing ridership as validations per 15 minutes. Increases starting at 5 AM. AM peak is at 8 AM with 75 validations per 15 minutes, then decreases to about 30 validations at 10 AM. Then increases gradually until 6 PM peak with 110 validations per 15 minutes and decreases until 12 AM.

Validation retention graph for 2021 compared to 2019 as a percentage. Over 100% retention at 5 AM, then decreases sharply to about 15% at 8 AM. Increases to 40% at 10 AM, stays about there until drop off at 3 PM with low at 5 PM with about 18% retention. Increases gradually to 60% at 1 AM.

This station has morning and evening peaks with the evening peak being much larger. It has very high ridership in 2019 but retained very little of it in 2021.

Mystery Station 5, 2019

Line graph showing ridership as validations per 15 minutes. First increase at 5 Am, reaching peak from 6 to 8 AM with about 20 validations per 15 minutes. It decreases until 10 AM, then stays at about 6 validations until 2 PM. Increases to about 8 validations from 2 to 6 PM, then decreases until 12 AM.

Mystery Station 5, 2021

Line graph showing ridership as validations per 15 minutes. Line first goes up at 5 AM, stays at about 6 validations per 15 minutes from 5 to 6:30 AM, then peak of 8.5 validations per 15 minutes at about 7 AM. Then decreases gradually until 12 PM with 3 validations. Increases until 3 PM at 5 validations then decreases gradually until 12 AM.

Line graph showing validation retention for 2021 compared to 2019 as a percentage. Line starts at 70% at 5 AM, then decreases to low at about 31% at 6 AM. Then increases gradually until 1 PM at 64%, then decreases to 42% at about 3:30 PM. Then increases gradually yo 90% at 1 AM.

This station has a morning peak region and very low ridership overall.

Bonus Question: In 2019, the highest number of average validations at a single station in a 15-minute period was 1,026. In 2021, this station was still the highest, but it decreased to 257. What station and time period was it?

Thanks for reading! We hope you enjoyed learning about validation patterns with us. You can comment your answers below, and we will post the answers next week so you can see how you did (comments may take a little while to post). Good luck on the quiz!