Return to LIRR Commuters Campaign Home Page
The "On Time Performance" fraud.
      Criticisms:
      The LIRR uses various methods and ‘gimmicks’ to artificially raise the monthly and yearly published OTP number to their advantage. Some of the methods are explained as follows:

        The LIRR “On Time Performance” Fraud
        1. What does it measure?

        1.a Even if the current OTP number is accurate it only measures EMPTY equipment – not passenger experience. It is possible, and happens frequently, that a train can discharge all, or almost all, of its passengers late at their destination stations and still be on time according to the LIRR. A train to Montauk, for example, can discharge the vast majority of its passengers before Speonk and then "make up time" on the rest of the way to Montauk. If the LIRR has some "slack time" in the schedule that can make the statistical difference between being on time or not. If the train then arrives in Montauk within the six minute delay allowance (or at least the time is entered as being within that six minute allowance) of its scheduled time it is marked on time. Yet, all, or almost all, passengers were discharged late at their destinations. The current On Time Performance number measures empty trains - not passenger experience.

        1.b The 6-minute allowance is a very different percentage of running time. Should a train with a 30-minute running time have the same lateness allowance as a train with 180 minute running time? The 30-minute train would be 20% late while the 180-minute train would be 3.3% late.

        1.c The "percentage" measurement also hides the LIRR's worst performance. If the OTP percentage was 90% how late were the 10% of late trains? If they were all one hour late it would be a far worse performance than if they were all 6 minutes late - yet both situations would result in the same 90% OTP.

        1.d In addition the LIRR releases only the overall, AM peak and PM peak OTP numbers. Why not have Monday AM, Friday PM, weekday and weekend numbers? The LIRR also does not list the OTP for each individual destination or individual train it measures. Patchogue, Speonk and Montauk trains are listed together as the Montauk Branch. Likewise, Huntington is not listed separately from Port Jefferson.

        2. Manual (and other) timings.

        2.a Not all LIRR train-timings are "automatic" timings. The LIRR computer train timing system (TIMACS II) allows for the block operators or train directors to 'manually' time a train at a location. All manual block territory timing locations are manually entered by an LIRR employee. This means the operator can enter any time they wish to 'time' a train. There are witnesses who have seen late trains that have been "re-timed" manually to make them on time. Every use of a manual timing should need permission, should be logged and have a reason why it was needed. Manual timings make the accuracy of the entire system suspect at best and corrupt at worst.

        2.b Even the MTA has disagreed with the LIRR OTP numbers and methods saying that instead of 92% it was really about 85% and that the LIRR was timing trains about one mile before their destination in order to get the 92% number. The LIRR still times trains before their destination, again not reflecting passenger experience. The LIRR's parent organization, not an impartial observer, disagreed with LIRR methods and accuracy. It is well known that the LIRR has timed many trains at their destinations (West Hempstead for example) when they passed by the Valley Stream block tower. Apparently the block operator 'knew' the train would be on time and didn't want to wait for an actual timing.

        3. Slack Time.
        This is used to add an extra 'cushion' of time before a trains scheduled destination. For example if the actual running time between a trains second-to-last stop and the last-stop is five minutes the LIRR will deliberately schedule (for example) eight or nine minutes giving them three or four minutes (or more) IN ADDITION to the six (by definition) before a train would be marked late. This is both deceptive and counter-productive. This means a train can arrive ten or twelve minutes past its actual destination time and still be marked on time. It also gives a false impression that the train can arrive "early". This is clearly visible on current printed schedules - such as for Speonk trains. Running time between Speonk and Mastic Shirley can vary from 13 minutes to 21 minutes - giving the LIRR an additional 7 minutes, 12 total, in some cases to be "on time."

        4. Schedule adjustments.
        If a particular train is consistently late the LIRR will 'adjust' the schedule to fit the trains performance – rather than find and correct the cause of the lateness. For example train 2739, 8:06 AM westbound out of Sayville used to be the 8:03 AM. In addition to other gimmicks the LIRR added three minutes to their expected performance. There are MANY other examples of this type of schedule manipulation. This practice also re-enforces the lazy attitudes and does not promote any kind of responsibility or determination of WHY trains are late. Adjust to the lateness and always slacking off seems to be the method of operation choice for the LIRR.

        This example is for only ONE train at ONE station. Multiply this by the all the LIRR stations and for the LIRR's approx. 740 trains and it is a pervasive, deceptive and degrading practice.

        5. Cancelled trains are not late. When the LIRR has severe equipment or weather problems it frequently cancels trains. It has a strong motive to do this - canceling trains improves the OTP number. Cancelled trains are not counted in the Monthly OTP statistics. More people are inconvenienced and then stuffed onto another already overcrowded train (safety anyone?) and the OTP number is better? To say this is counterproductive and a disincentive is an understatement. Yet 'work' trains with no passengers onboard are given very liberal schedules, arrive empty and on time and are counted in the OTP.

        Deceptions:
        The "59 minute" train and other ploys.
        When the Ronkonkoma electrification was finished the LIRR publicly announced the "59 minute ride from Ronkonkoma to Penn Station." Where is that train now? According to the schedule there is currently no such train. Yet there were TV and Newspaper reports about "the 59 ride to Manhattan." I guess the media only cover the LIRR when they are lying or creating a PR stunt. Even when the "59 minute train" was supposedly in the schedule it never ran on time. Cheap meaningless public relations stunts like that typify the LIRR's disregard for its passengers and the public at large.

        The “Delay Analysis” Process.

        The delay analysis system used by the LIRR is really a misnomer. There is no analysis. The system should be called a delay reporting system. This system was originally written by the LIRR using only WHOLE minutes – it was not possible to enter or track seconds. When the LIRR re-wrote this system they changed the former 'within 5 minutes' delay definition to the current 'within 6 minutes' late -- 5 minutes and 59 seconds -- definition. This computer system is where the LIRR tallies their delay numbers to come up with the perpetual 92%, or better, number. Maybe they will discover another 'bug' and change the OTP allowance to 7 or 8 minutes. Meanwhile Metro North, according to their own public relations people, still uses the official 4 minutes and 59 seconds for its lateness measure - yet it almost always has much better numbers, and actual performance, than the LIRR.


        Suggestions:

        The LIRR SHOULD analyze and cross-reference delays by season, date, time of day, day of week, weather, crew, train, equipment, branch, location, passenger load and perhaps some other factors. Currently they hide, lie and deceive about this; therefore, they do not even have the correct information to determine the causes of delays. They are condemned to repeat their mistakes. The OTP should be a 'how are we doing' indicator instead of a phony public relations ploy.

        Many companies do failure analysis to determine why something fails and therefore what they can do to prevent future failures. The LIRR, by being deceptive and deceitful keep themselves from even having the data and methods to do this. The LIRR ensures the same mistakes will be made again and again. Regular commuters on the LIRR experience this first hand.

        A "Passenger Hours Delayed" index should be compiled and used as a performance indicator of passenger experience.

        The LIRR should track and publish a "Passenger Hours Delayed" on time performance index. This should be published on the Internet and available in real-time. If the LIRR does not want to stop keeping their beloved percentage OTP number they could use both measurements - to give passengers a more complete and meaningful indication of their performance. Since the LIRR does not do this one can conclude that giving passengers a meaningful measurement of LIRR performance is not something they want to do. State Departments of Transportation, the U.S. DOT, The LITP 2000 and many Universities use a “passenger hours late” standard as a measure of performance. Apparently, it would be more difficult to manipulate so the LIRR continues to use the percentage method.



        Free Counters
        Search Engine Optimization Services
LIRR, LIRR Commuters, LIRRCommuters, LIRRCC, LIRR Commuters Campaign, LIRR Commuters Council, Commuters, Long Island Railroad, Long Island Railroad
Return to LIRR Commuters Campaign Home Page