Craig Sauvé, who is the director of mobility in the Plante administration, as well as the vice-president of the STM, was the subject of scrutiny on Twitter yesterday afternoon for seemingly misleading claims about the reliability of metro service.
TL;DR Sauvé's claim that, in 2018, the metro experienced the least delays than it has in the last five years has been disputed.
While making a presenation, Mr. Sauvé claimed that 2018 was the STM's best year yet in terms of metro interruptions. He mentioned that there was an increase of 5% in metro users, and that interruptions were down by 8%. This makes it the best year, in terms of interruptions, in five years.
The second part of that statement was disputed by HuffPost Quebec journalist Olivier Robichaud.
Le conseiller @CraigSauve affirme que le #métro a connu sa meilleure année en 5 ans. C’est faux, archifaux. Sauf 2017, c’est la pire année pour laquelle on a des données. #polmtlhttps://t.co/KVrkOFJDT4
In fact, 2018 was the second-worst year for metro interruptions in recorded history. The worst recorded year was 2017.
Sauvé replied, stating categorically that he had been telling the truth: there had been a 5% increase in passengers and an 8% decrease in delays.
This may be true, but only because 2017 was an exceptionally bad year for metro delays. And it certainly doesn't mean that this had been the best year in five years.
Robichaud gives us the statistics for metro interruptions lasting more than five minutes in this tweet: we can clearly see that the last two years have had above-average numbers of interruptions:
Pannes de plus de 5 minutes:
Moyenne sur 5 ans: 1005
Moyenne sur 10 ans: 968
Sauvé attempted to qualify his statements, saying that, given that metro trains had in total travelled longer distances, it was the best year in five years.
Oui, @ORobichaud il y a eu plus d'événements bruts, MAIS il y a eu ben plus de kilomètres parcourus par le métro en 2018. Et même avec plus du monde dans le métro, il y a MOINS d'intérruptions de service 5min et + par kilomètre parcouru. 1/2 #polmtl
However, it is hard to see what this has to do with the number of delays.
That «point» is also false. The 8% decrease he cites has nothing to do with delays per km. It is the 1170 total interruptions of 2017 (a record year) minus the 1073 total interruptions of 2018. No one has yet given a stat concerning km travelled. #stm#polmtl
The official STM account jumped on the thread this morning, repeating Sauvé's first statement, and stating that there will soon be a new indicator to measure metro delays.
La STM a effectivement connu une baisse de 8 % des interruptions en 2018 et ce, alors que l'achalandage est en hausse (~ 5 %) et que le métro roule davantage (+ 3,8 %). Un nouvel indicateur basé sur l'offre de service sera d'ailleurs présenté dans notre prochain rapport annuel.
However, it is very easy to manipulate statistics into saying whatever one wants them to. We always have to be critical of the information presented to us, especially when this information seems dubious.
Are you wary of this new indicator?