MRT performance improved or “meaningless” statistics?

MRT network performance up 30% as trains travelled longer distances before delays?

The Straits Times article on 12 January, “MRT network performance up 30 per cent in 2016 from 2015” states that “MRT trains generally travelled longer distances before encountering delays last year, compared to the year before, going by latest data from the Government showing improved train network performance.

Last year, trains clocked 174,000 train-km travelled between delays of more than five minutes, a 30 per cent increase from 133,000 train-km in 2015.”

Did Downtown Line stage 2 contribute to the improvement?

Since “The best performing MRT train line was the Downtown Line, which achieved 260,000 train-km between delays of more than five minutes last year, up from 45,000 train-km the previous year” – to what extent did the additional track, stations and trains of the new downtown line stage 2 extension which started on 27 December 2015 – contribute to the improvement?

Who cares how many km travelled between delays?

Arguably, what commuters may be more irked by – may not be how many km were travelled between (before the next delay) delays of more than five minutes, but rather how many delays of more than five minutes there were and whether these have increased.

How many longer delays?

What is perhaps even more important from the perspective of commuters may be how many delays there were of more than 10, 15, 30 minutes, one hour, etc – and whether these have increased?

As an analogy – it may be akin to telling you that you walked longer before you got hit by the next accident, but without telling you how serious the accidents were, or whether the number of accidents have increased?

Less meaningful statistics?

This may in a sense be somewhat like telling you the statistics that has less meaning or arguably, no meaning at all to you.

Learn about statistics?

Maybe the people who run tains may need to have a refresher course in statistics – can use SkillsFuture Credit or not?