San Francisco Muni Metro Performance Statistics

Performance has declined during the 2004-2010 time period.
Cf. SF Muni Metro Performance Graphs.

On-time performance
DateJ
Church
K
Ingleside
L
Taraval
M
Oceanview
N
Judah
Average
or Total
Jul-Dec 200480%81%76%70%79%77%
Jan-Jun 200557%72%79%71%69%70%
Jul-Dec 200554%75%87%65%83%73%
Jan-Jun 200670%69%64%61%69%67%
Jul-Dec 200661%68%72%73%73%69%
Jan-Mar 200771%81%71%74%
Apr-Jun 2007*74%72%
2004-2007
average
66%74%75%69%74%72%
standard
deviation
10%6%8%5%6%4%
daily riders
in thousands
1915232331112
scheduled time
for route
in minutes
44'36'35'43'38'40'
*The T line began service in April, 2007

Runs operating within scheduled headways
DateJ
Church
K
Ingleside
L
Taraval
M
Oceanview
N
Judah
Average
or Total
Jul-Dec 200462%100%90%44%50%69%
Jan-Jun 200555%100%100%79%41%75%
Jul-Dec 200537%56%82%57%60%58%
Jan-Jun 200649%57%41%67%58%54%
Jul-Dec 200637%54%52%67%62%54%
Jan-Mar 200742%76%61%65%
Apr-Jun 2007*60%88%
2004-2007
average
47%74%71%63%60%63%
standard
deviation
10%22%23%12%16%8%
daily riders
in thousands
1915232331112
scheduled time
for route
in minutes
44'36'35'43'38'40'

On-time performance
DateJ
Church
K Ingleside
-T Third
L
Taraval
M
Oceanview
N
Judah
Average
Jul-Dec 200767%75%79%59%69%70%
Jan-Jun 2008n.a.74%70%72%65%n.a.
Jul-Dec 200862%61%70%62%73%66%
Jan-Jun 200972%68%73%70%68%70%
Jul-Dec 200972%84%73%68%78%75%
Jan-Jun 201076%70%#77%#68%#80%74%
2007-2010
average
70%72%74%67%72%71%
standard
deviation
5%8%4%5%6%4%
Jul-Dec 201077%54%70%62%72%71%
Jan-Jun 201166%58%70%#67%#78%#68%
scheduled time
for route
in minutes*
35'73'36'40'47'44'
* The changes in time from 2004-7 are due to several changes in routes.
The K and T lines are combined.
# Calculated from FY Q1, Q2, and average figures.

Runs operating within scheduled headways
DateJ
Church
K Ingleside
-T Third
L
Taraval
M
Oceanview
N
Judah
Average
Jul-Dec 200774%75%66%62%50%65%
Jan-Jun 2008n.a.56%53%60%46%n.a.
Jul-Dec 200842%40%50%49%53%47%
Jan-Jun 200953%48%54%44%35%47%
Jul-Dec 200949%66%58%62%47%56%
Jan-Jun 201050%n.a.n.a.n.a.82%n.a.
2007-2010
average
54%57%56%55%52%54%
standard
deviation
12%14%6%8%16%9%
Jul-Dec 2010n.a.44%66%53%n.a.n.a.
Jan-Jun 201148%45%n.a.n.a.n.a.n.a.
scheduled time
for route
in minutes*
35'73'36'40'47'44'
* The J ends at Embarcadero instead of 4th and King Streets.
The N ends at 4th and King Streets rather than the Embarcadero.

  1. San Francisco Muni checks its performance once in every six month period for each line.  Muni data is posted at SFMTA;  however, as of 1/15/13, the last Muni update was on 8/17/11.  Ridership numbers are from sfgate.com.
  2. The Muni goal is to have at least 85% of vehicles run on time according to published schedules, no more than 4 minutes late or 1 minute early.  According to an article in the Bay Citizen, Muni has actually considered 4 minutes 25 seconds late as being "on time" and since 2010, 4 minutes 59 seconds has been considered "on time".
  3. The Muni goal is to have at least 85% of vehicles within the lesser of 30% or 10 minutes of scheduled headway (interval between trains in minutes).
  4. Using linear regression, there has been a gradual decline in the average headway for the Muni Metro system during the 2004-2010 time period;  average on-time performance has declined slightly from 71% to 70%  Muni Metro performance graphs.
  5. Headway, the interval between trains, is more important to most passengers than the on-time performance.  Most passengers do not carry schedules and the schedules are not posted at Muni stops.  If a train is scheduled to arrive every ten minutes, for example, and every train is five minutes late, the interval between trains would still be every ten minutes.
  6. The results for headway between trains are worse than those for on-time performance in the above data.  This may be because the criteria for headway are stricter than the criteria for on-time performance in many cases. A train is not considered late unless it is at least four minutes behind schedule. As the scheduled headway during commute hours ranges from seven to ten minutes for the Metro system, a train that is two to three minutes late may not meet the criteria for headway.

Creative Commons License
Robert Karis
www.sfog.us