Bus Turnaround performance metrics are available and documented as part of a public API
, and related analysis code is open for anyone to review, use, and adapt on Github
. This API and the analysis code were produced by Nathan Johnson, Kuan Butts, JD Godchaux, Lela Prashad, Martin Lehner, and Neil Freeman. The Bus Turnaround Coalition owes special thanks to Nathan and Neil, without whom these metrics would not exist.
Each route is assigned speed and reliability grades based on thresholds in the table below. Those grades are then assigned to a traditional 4-point scale and averaged to create the route’s final grade, per the table below. Additionally, a bus route that receives an “F” on either performance metric automatically fails, as do buses with a “truancy” rate of greater than five percent (according to data from NYC Transit).
||> 4 buses/hour
||< 4 buses/hour
We have strived to fairly grade NYC buses by holding them accountable to a high but achievable standard that acknowledges the constraints on bus operations in a dense, metropolitan city. In other words, we are not grading on a curve. The Turnaround Coalition’s approach to defining the above grade thresholds is as follows:
We set a goal of increasing speed to a median of 10mph, which is approximately the average speed in Chicago and Boston. The current median speed among NYC bus routes is 6.6mph.
We set a goal of reducing bunching on high-frequency routes to 5%, which would represent 1 in 20 buses bunched, or for a bus rider that rides twice per day on weekdays, one bus every two weeks. Current median bunching among NYC bus routes is 9%.
We set a goal of running at least 75% of low-frequency buses on-time—the same goal that Boston’s MBTA has set for itself. The current median for on-time performance among NYC bus routes is 59%.
We have defined grade thresholds in relation to these goals and to the current statistical distribution of performance (recognizing that, no matter what, some routes will perform better than others). Bus routes achieving the above goals for performance are assigned a minimum of a “B” for that performance metric (for example, a bus with an average speed of 10mph earns a B for speed), and other grade thresholds are based on percentiles within an “ideal” performance distribution, in which the current distribution remains but has the median shifted to achieve our performance goals. Email us if you want to get all-the-way into the weeds: firstname.lastname@example.org