Fans Take Metro Home From Soccer Game
Over 7,000 customers took Metro home from the soccer game at Fedex Field on Tuesday night, July 29th, and some used the Silver Line!
While researching the ridership patterns of customers using Silver Line and comparing the usage on Monday, July 28th, to Tuesday, July 29th, we started to see an interesting pattern: a lot more customers were traveling from Morgan Blvd to Wiehle-Reston East. When we drilled down into the data, we found that these trips were occurring in the evening, well past the end of the PM Peak. A few web searches later we discovered that a well attended soccer game had been held at FedEx Field on that night.
We then wanted to know, how many and where did they travel? The chart below compares entries at Morgan Blvd on the 29th and contrasts it against the entries of the previous Tuesday, July 22nd. It is pretty obvious that except for the soccer game, it was a pretty normal day at Morgan Blvd, good for comparison. This data tells us that around 7,500 customers took Metro home from the game. But where did they go?
The chart below shows the stations that our Morgan Blvd customers used to exit the system after the game. There are a lot of unreadable data points on there, true, but a few really stand out, as annotated. Most of the big spikes are end-of-line stations, including our newest, Wiehle-Reston East. This graph also tells us that customers traveling to Wiehle-Reston East made it to their home station faster than those going to Franconia-Springfield. In total, over 500 customers took the Silver Line home to one of the five new stations after the game.
This chart also raises the question, why were so many soccer fans going to Eastern Market, Metro Center, Foggy Bottom and Pentagon City? Those are the earlier spikes shown on the graph. Perhaps those are the neighborhoods where high densities of soccer fans live, or maybe the locations of good sports bars for post-game refreshment.
Did you attend the soccer game and take Metro afterwards? Where did you go and what was your trip purpose?
All the stations highlighted (except Vienna) are non-stop rides on the Blue or Silver lines. It would be interesting to see how many people chanced a late-night transfer to reach a station (end-of-line or anywhere) on the Green, Yellow, or Red lines. All those lines under the peak could add up to a healthy proportion of riders, or it could indicate a need to change service patterns (extra trains, timed transfers) to encourage broader regional Metro ridership to games at FedEx Field.
Also, 7500 customers at Morgan Blvd would stuff 62 Metro cars, or ten full six-car trains, at max capacity. Since the game was (apparently) not anticipated by Metro planners, only two (one Blue, one Silver) westbound trains would have arrived every 12-20 minutes on regular evening/late-night service, which would leave customers waiting 45 minutes to an hour to board. (This can probably be confirmed by analysis of station entry times and how long it took customers to reach their exit stations.) Will Metro consider coordinating with FedEx Field to arrange modified service for heavily-attended events?
The other great piece of information from Michael’s chart of entries is how few WMATA customers use Morgan Blvd. on an average day in July (excluding events at FedEx Field). In 45 minute segments entries never apparently exceed 200 people.
Also agree that I don’t see how all these people were served without additional service. But I don’t see a press release as evidence that there was extra service.
It would also be interesting to aggregate that bottom chart by station and just display a bar chart by station: “Exit points for passengers who entered at Morgan Boulevard between 8:30PM and 11:30PM on 7/29”
Not that surprising to me that some of the highest stations are direct shots on the same line, or the end-of-line stations all the way on the other side of the system (long drive on the beltway/slog through traffic).
I need to point out that while the labels are in 45-minute intervals, the datapoints are actually in quarter-hour intervals. While entries at Morgan Blvd are relatively low, they’re not as low as you are asserting.
I’ll look into additional analysis on this event re train schedules, passenger loads, etc.
(1) interesting that riders reached Wiehle before Franconia since train time should be longer. Was it due to differences in train arrival times at the end of the game or headways on the two lines. This would be verifiable by looking at train schedules that evening.
(2) This also seems like it could be a rich source of data for whether people transferred to Yellow to get to Pentagon City and other Blue/Yellow stops or if they opted for the single seat ride.
About capacity and ridership that night:
1) You’ll note that people started entering Morgan Blvd from the game around 8:45 PM with entries peaking between 10 and 10:15 PM. The ridership is spread out a bit, meaning that not all 7,500 soccer fans entered the Metrorail system at Morgan Blvd at the same time. In fact the max quarter-hour entries was around 2,400.
2) The base service at Morgan Blvd is now double what it was only days before the game. With both Blue and Silver at Morgan Blvd trains now come twice “every 12-20 minutes”.
3) It appears we did add extra service that night. In fact, my records show we provided 10 trains per hour during the 9 PM and 10 PM hours including some eight-car trains. Assuming 100 pax per car, we provided carrying capacity of 6,400 pax per hour.
Hope this helps.
Thanks for the update, Michael. The number of trains per hour could use some clarification. If we count Blue and Silver trains heading eastbound, then at standard 20-minute service intervals, about twelve trains would service Morgan Blvd between 9pm and 10pm. (Three westbound Blue, three eastbound Blue, three westbound Silver, and three eastbound Silver.) However, since the six eastbound trains go only to Largo (which isn’t called out as a major destination in the graph), almost all Morgan Blvd customers would be expected to board the six westbound trains, with a capacity between 3600 (all six-car trains) and 4800 (all eights) for the hour. In that case, “ten trains per hour” would indicate roughly normal service, and not enough service. If, however, “10 trains per hour” means ten in each direction (five, not three, of each color, in each direction, for twenty total), then that would have brought capacity up considerably. In either case, the Silver line helps a lot!
Also, 2400 entries in a quarter-hour works out to eight customers every three seconds (about as fast as you can say “three customers, three customers, two customers”, in the style of “one Mississippi”), constantly, for fifteen solid minutes, with busy periods before and after. Does any data suggest that customers can enter Morgan Blvd station any faster than that? If not, the chart could be interpreted as a crowd of customers arriving at the station starting after 9:45pm, queued and passing through the faregates at the maximum rate until sometime after 10:15pm.
The data Metro provides is wonderful for detecting events like this soccer match, and analyzing the system’s response at a high level. However, it seems some additional datasets (like train arrival rates, as Michael offers above) are needed to truly understand the situation and service level.
Hi, @Kevin
Let me clarify. We provided 10 inbound trains during each of the 9PM and 10PM hours. Sorry I wasn’t clear. I wasn’t even thinking about Largo-bound trains so I neglected to mention which direction I was discussing.
As for the entries, and whether there were queues at the faregates, I can’t say with the data to which I have easy access. The minimum granularity is quarter-hour intervals. Assuming a faregate can process 30 transactions per minute, we would only need 6 in the entry configuration to process a flow of 2,400 entries per quarter-hour interval. I think we have 9 aisles at Largo, and I’d assume most of them would have been in the entry configuration that night.
Awesome. Thanks for the information!
Thanks, Michael!
I’d love to see a comparison of post-game ridership destinations between soccer and football.
I don’t know how much it would help transit planning, but it would sure bring a lot of new readers to this blog.
It would be interesting to see the destination data in map format. I’m guessing that some of those are end of line locations where folks parked, but to some extent it would show the distribution of soccer fans (that take Metro to games).