Parking Probability

firebus's picture

I was parking at a metered spot today, and didn't have enough money to pay for the time I needed. But I figured it was probably worth paying for part of the time, to cut the chances of getting a ticket.

It would be great to create a parking calculator that would tell you when it was a good bet to pay the meter and when it was a good bet to risk a ticket.

You'd need to know

  • How long you will be parked
  • Hourly rate for the meter
  • Amount of the fine if you get a ticket
  • Probability of getting a ticket per hour parked

The last item above is pretty hard to calculate. You could do a reasonable guess if you knew

  • How many parking enforcement officers are working at a given time
  • How many blocks per hour an enforcement officer can cover
  • How many blocks there are in the city that have restricted parking

The above is sufficient if officers are equally likely to patrol a given block. In reality, cities probably put more enforcement on areas with higher fines and/or higher hourly rates (both from a financial and a justice standpoint, higher cost locations should have more enforcement). So it would help to have a probability weighting based on either the rate or the fine, relative to where that rate or fine is in the overall pricing structure.

Coming up with the above data is beyond the capacity of the n of 1 research methodology. However, it's maybe interesting to turn the question around: Given a parking rate, a length of time parked, and a fine, we can calculate how often the city needs to patrol a given block in order to make it worthwhile to pay for parking at the meter.

For example, at my space this morning, the meter was $2/hour. I wanted to park for an hour. The fine would be $55. So this block needs to be patrolled at least once every 27.5 hours for it to be worth paying the meter. It seems possible, but maybe not probable, that this spot gets patrolled daily (it's a kind of remote corner of the city).

However, complicating the calculation is the fact that there are different numbers of officers on the street at different times of day, and most meters are only in effect for certain parts of the day.

So that's not 27.5 hours of real time - the meter is only in effect from 8AM to 6PM, or 10 hours per day. So 27.5 hours is almost 3 days of real time. For some reason it now seems more probable that it's worth paying the meter, but that might be an illusion...


Comments

firebus's picture

Errata?

If I made any sort of egregious logic error in the above, please let me know :(


the panopticon

Just like in a panopticon, you never know when you are being watched and the penalty if you are caught is nearly as steep.

Though currently too chicken to do this myself, I admire the modern day cool hand luke who spray painted many of the windows on the meters parking meters in highland park gray. Thus making it impossible to tell if the time is expired.

I suppose a can of gray spray paint can also come in handy when parking is scarce. A red zone can quickly be modified to accommodate the weary traveler.


firebus's picture

the revolution will be graphed

you might be right that spray paint is a better weapon than statistics in this war we both fight


Powered by Drupal - Design by Artinet - Amazon Affiliate