Much ado about nothing
There's been a lot of press in Toronto recently because there has been a huge spike in "car versus pedestrian" fatalities (14 so far this year). So, of course, the city is spending money to study the problem.
Guess what? It's probably not a problem.
The more likely explanation is a fluke of statistics called a "Poisson Burst" (named for the mathematician, of course). The basic idea is that when you have an event that occurs randomly, sometimes it will occur more often, and sometimes it will occur less often.
Consider this - are there any differences between, say, last January and this one, that would account for the extra fatalities? Well, it's less snowy and slippery, but that should drive the numbers down, not up, as people are less likely to slip under a car.
So, in the absence of any new factor (i.e. any change in the rules of the system), we need to take a step back and consider if there's any cause to look for.
Consider this statement: on average, 24 pedestrians are killed every year by cars in Toronto. (I am too lazy to look up the real number, just bear with me here.) Does that mean that, every year, 24 people die? On December 31, do we need to do a tally, and if we're a bit short, throw a couple of people under a bus? Obviously, no. What that means is, if we take the number of deaths every year for a number of years, then the average is 24. No more, no less. Nothing in that precludes a year where nobody dies, followed by a year where 48 die. The average of those two years would still be 24.
Of course, politicians need to be seen do be doing something, and the press needs something to write editorials about, and so lots of money gets spent and lots of ink gets spilled. None of it will do any good, but at least we're "doing something", right?
Guess what? It's probably not a problem.
The more likely explanation is a fluke of statistics called a "Poisson Burst" (named for the mathematician, of course). The basic idea is that when you have an event that occurs randomly, sometimes it will occur more often, and sometimes it will occur less often.
Consider this - are there any differences between, say, last January and this one, that would account for the extra fatalities? Well, it's less snowy and slippery, but that should drive the numbers down, not up, as people are less likely to slip under a car.
So, in the absence of any new factor (i.e. any change in the rules of the system), we need to take a step back and consider if there's any cause to look for.
Consider this statement: on average, 24 pedestrians are killed every year by cars in Toronto. (I am too lazy to look up the real number, just bear with me here.) Does that mean that, every year, 24 people die? On December 31, do we need to do a tally, and if we're a bit short, throw a couple of people under a bus? Obviously, no. What that means is, if we take the number of deaths every year for a number of years, then the average is 24. No more, no less. Nothing in that precludes a year where nobody dies, followed by a year where 48 die. The average of those two years would still be 24.
Of course, politicians need to be seen do be doing something, and the press needs something to write editorials about, and so lots of money gets spent and lots of ink gets spilled. None of it will do any good, but at least we're "doing something", right?