GIGO

Forbes has helpfully ranked the most dangerous cities in the country so you will know where you’re most likely to get a cap busted in you:

  1. Memphis, Tenn.
  2. St. Louis, Mo.
  3. Kansas City, Mo.
  4. Detroit, Mi.
  5. Miami, Fla.
  6. Tulsa, Okla.
  7. Nashville-Davidson, Tenn.
  8. Indianapolis, Ind.
  9. Oklahoma City, Okla.
  10. Stockton, Calif.

I would have really expected Chicago, New York, New Jersey, Las Vegas and Los Angeles to make that list. I guess they really must be liberal heavens. It isn’t like crimes in some cities might be under reported to make officials look better. Of course illegal immigrants report every crime they see. And no one has ever suggested that crimes in some high-crime areas are ignored or go unreported by the population. Oh wait… really? Apparently all of those things happen. Much like a poll on a website that asks any visitors to choose between “Justin Beiber” and “Justin Timberlake”, the data isn’t entirely accurate.

The FBI says not to use their stats to create rankings of cities:

The Federal Bureau of Investigation provides detailed crime statistics every year, but warns against creating rankings solely based on their data, since they are individually reported by agencies and reflect vastly different socioeconomic circumstances.

The FBI was probably just wanted to be polite and not mention that other stuff, but all of their stats are voluntarily submitted by the cities police departments. And they specifically say not to create rankings based solely on their data. So what does Forbes do to “fix” their data. Apparently they are in contact with the top climate scientists at the IPCC and decided the best way to fix their bad data was to combine it with some other data:

By combining the data with traffic fatalities we feel our list offers a proxy for safety.

A wrong and a right cancel each other out is apparently how this works. In computer programming we have a saying, Garbage In – Garbage Out. You can’t combine up bad statistics with good statistics and hope the good ones will reform the bad ones. The bad ones go on a bender with the good ones and they all turn up passed out on your computer screen the next day.

There are parts of Oklahoma City and Tulsa that I wouldn’t go to without an actual can of whoopass. However, Chicago doesn’t even make the top 10? Come on. Maybe a stats person can correct me here, but I don’t see any reason to trust such a ranking.

Advertisements
Both comments and trackbacks are currently closed.

Comments

  • Tam  On October 16, 2010 at 9:17 pm

    As it turns out, Chicago et al were deliberately excluded for reasons only muzzily explained…

    • trackerk  On October 16, 2010 at 10:51 pm

      Was the disclaimer added after the fact? I’m not the brights bulb in the pack, but can’t imagine I missed the big fat “we don’t have data on probably the most dangerous city in the country” explanation.

%d bloggers like this: