Tuesday, March 6, 2018

Transport safety: Statistics can be misleading


I saw a "scary" headline the other day concerning New York City's Metropolitan Transport Authority (MTA).

MTA buses were in more than 21K collisions in just 3 years

MTA buses racked up at least 21,823 crashes, collisions and other mishaps over 31 months beginning in 2015 — an average of 23 per day, The Post has found.

At least 2,520 people — or 2.7 per day — were injured during the time period, according to the MTA. At least 14 people died, including a 25-year-old skateboarder, a 62-year-old pedestrian, a 60-year-old motorist and a 70-year-old with a walker who was mutilated by a hit-and-run bus.

There's more at the link.

That sounds pretty bad, until you look at a few more statistics.  The MTA reports that it has no less than 5,725 buses in service, covering 2,952 route miles.  If we take the "scary" statistics above, that means in three years, each bus, on average, would have suffered less than 4 accidents, or just over 1 per year, ranging from the tiniest of scrapes to a major collision with casualties.  Given the state of New York City traffic, I'm astonished that it's that low!  Have you seen the number of cars on the city's roads, and how everyone drives there?  The video clip below is pretty typical of rush hour.





The casualty figures, while very sad, even tragic, are also not unreasonable when viewed over that size of fleet.  They translate to about 2¼ casualties per bus in three years, or less than one per year.  Again, given the traffic volume in New York City, and the number of people living there, that really isn't as bad as the headline made it sound - particularly when one realizes that a single accident might produce several casualties, whereas dozens of other accidents might produce none at all.  The average doesn't reflect that reality.

It would be wonderful if an organization like the MTA could operate without any accidents, damage or casualties . . . but that would be humanly impossible.

Peter

8 comments:

P. Lester said...

I have driven in Manhattan and would never do it in anything but a fully insured rental car. People will bang into you and never give it a thought. They will gridlock intersections so others can't proceed and apparently the horn is necessary for reasons that are not readily apparent to those of us from "Fly Over" country. Maneuvers that would put one at risk of being a homicide statistic in Texas are considered of no consequence there. In their defense, I have to say I have found the people of NYC to be as courteous and pleasant as any others. However in traffic they seem to have developed a more gladiatorial culture than I am used to.

Ben Yalow said...

As a native New York City resident, I've been driving in Manhattan for four decades, and a pedestrian there for six. And the difficulties, while very real, are exaggerated by that video, which shows a rush hour traffic in a busy area.

And driving isn't gladitorial. But it does require a recognition that traffic density is really high, and you need to be very aware of how much space your car occupies, so you can fit into tight spaces.

The other thing that this video really doesn't show is the effect of pedestrians. New York pedestrians walk far more than those in other major cities -- it's fast and easy to get around. And since parking is so expensive, it's not unreasonable to park a half mile or so from where you want to be and walk there, if you're heading into a major traffic area.

And the right of way rules for pedestrians are simple -- "If it's safe, go." Traffic lights are there to give you some idea of what cars might be doing, and crossing at corners is pretty optional. As my mother (also a native New Yorker) taught me when I was first leaning to cross streets, "Look at the cars, not the lights; nobody ever got run over by a street light."

Which means that what you can't see in the video (but I'm sure is there) is a stream of pedestrians crossing the street through that traffic, weaving their way around the mostly-stopped cars.

New York also has lots of different routes to get between places, and natives all drive with the car radio set to one of the news/traffic stations (or maybe more than one on the buttons, since one gives reports every ten minutes on the ones, one on the fives, and one on the eights), and you replan the route depending on what's clogged. Yes -- there are traffic reports on the local news stations 24/7/365. It means you use local knowledge, not a GPS, for navigating around the city.

So, if you're driving in Manhattan, you need to take all those behaviors into account. But, if you do, then it's still a very practical way to go.

Steve Sky said...
This comment has been removed by the author.
Dave said...

Pedestrian behavior like Ben describes would get you run over in Texas - and the driver would be in the right. Texas transportation code specifies that pedestrians must cross at cross walks. Cross somewhere else and, well, enjoy the interaction with that brushguard-equipped F250.

And while Peter noted the number of busses and number of route miles, the truly important stat is the number of vehicle miles driven. The average NYC bus probably drives more miles in a month than the average car does in a year.

I'd be willing to bet that if you compared the rate of accidents per 100,000 vehicle miles for the MTA buses to the same stat for normal cars, the buses would come out ahead. Probably even moreso if you had some way to determine 'at fault' accidents

lpdbw said...

Be careful with that whole "at fault" concept, Dave.

We have some toy trains here in Houston, er, I mean light rail, and the poor design means the trains and cars co-exist on the same roadway.

Curiously, the city that runs the trains, pays the police, and runs the courts seems to find that in every single accident, it is the private driver who is at fault.

Weird coincidence, that.

Ben Yalow said...

And Dave is right -- it's illegal in New York to cross at other than a crosswalk, or against the light, etc.

Every few years, a mayor tries to enforce that part of the code. We've had lots of different methods tried.

They fail. People still cross when/where it's safe, not where the crosswalk is, or when the light says to. And learning how to cross streets requires learning judging distances, location, and driver behavior, not just the vehicle code.

And the police are all aware of that behavior, and, except when a mayor is in anti-jaywalking mode, all ignore it. I've crossed in the middle of a block in front of a parked, occupied police car, and neither I nor they thought this was unusual behavior.

When I'm showing friends around the city, I cross very differently from the way I do when I'm walking around by myself. I don't expect them to flow the way a native would.

If you look at the first few minutes of this video of a walk along Broadway, you'll notice people crossing in the middle of a block, against lights, etc. It's simply normal pedestrian behavior, and New York drivers are used to it.

https://www.youtube.com/watch?v=JiM5A6bTfKs

TheOtherSean said...

"Pedestrian behavior like Ben describes would get you run over in Texas" - Heck, it can get you run over in NYC, if you're unlucky. My great-grandmother met her demise when crossing the street while Christmas shopping, decades ago.I never got a chance to know her because a truck turned her into roadkill.

 Ashley said...

Which proves the old saying, "lies, damned lies, and statistics." Except I would add it's not the statistics at fault, but the lack of understanding about what the statistics imply. Your take down is a perfect example of understanding the implications, rather than just throwing the raw data out there.