When Good Intentions Meet a Sinkhole: MTA’s July OTP

Published in

cylussec

4 min readAug 6, 2019

Spoiler: The sinkhole did not make On Time Performance go up (photo: Brian Seel)

Note: For information about the methodology behind this citizen driven data project, see this post. To help with server costs, chip in on Patreon!

In case you missed the coverage, there was a massive sinkhole in Downtown Baltimore. This put a damper on the rather impressive improvements the system saw after its summer schedule change.

Overall, the system’s On Time Performance dropped from 54.1% to 49.97%, but the biggest hits came on some of the lines that were most effected by the sinkhole.

Left: OTP for 1 minute early to 5 minutes late. Green is on time, blue is early and purple is late. Right: OTP for 2 minutes early to 7 minutes late. Pink is on time, red is early, and orange is late.

While every high frequency route took a hit, four of them dropped by double digits: Brown, Navy, Yellow, and Purple.

Note that all comparisons are with the new summer schedule, so the old percentages are from June 23–30, and the new percentages are from July 1–31.

Citylink Brown

Citylink Brown dropped from 46.77% to 35.78%, a drop of 11%.

We can dive deeper and look at how late the buses were over the course of the month. The following graph shows the average number of seconds late each Citylink Brown was per day. The sinkhole opened on July 10.

Number of seconds off schedule averaged by day, were negative seconds means early. The x-axis is the date.

This shows that the delays got much worse after construction started, and slowly got a bit more under control about 11 days afterwards.

Citylink Navy

Citylink Navy dropped from 49.56% to 38.51%, a drop of 11%.

We see a similar trend on the delays per day on the Navy, although with a slightly less stark inflection point on the 10th.

Citylink Yellow

Citylink Yellow dropped from 47.6% to 36.73%, a drop of nearly 11%.

The Yellow sees a delay graph that closely mirrors the Navy.

Citylink Purple

Citylink Purple dropped from 52.47% to 40.46%, a drop of 12%.

The Purple is more curious, as it does not actually pass the sinkhole. Could it be that more buses were pulled from this route to cover the bus bridge?

We can look at the speed of the Purple buses as they go along their route to see if we can figure out what happened. Lets focus on the Eastbound, weekday buses. First lets look at June 1-July 10 to get a baseline.

Citylink Purple bus speeds as it goes through its route for weekday, eastbound buses from June 1-July 10. X axis is the stop order, and the Y axis is the speed. The three lines are the 25/50/75 percentile speeds.

Now, lets look at the post sinkhole numbers.

Citylink Purple bus speeds as it goes through its route for weekday, eastbound buses. X axis is the stop order, and the Y axis is the speed. The three lines are the 25/50/75 percentile speeds.

It really appears that the reduction in the OTP for this month was simply because the buses were slower along a large part of their route, which could be due to more traffic diverting from Pratt St to Baltimore St. Although, that shouldn’t matter since Baltimore has a bus lane, that should negate much of that delay if its enforced properly.