Geographic Information Systems
Pick Six and Instant Win Ticket
Distribution
New Jersey Lottery
Initial Concept
I don't normally play Pick Six. I don't like waiting until late at night to see if I won anything. Like most modern Americans, I want entertainment with instant feedback, so I usually play Instant Win scratch-off tickets.
I received a buy one, get one free coupon for Pick Six in the sunday newspaper, so I decided to buy a ticket. With my two quick picks in hand, I checked the state lottery's website to see if I had won.
I didn't. However, I noticed that the site allows you to look up past winners by ZIP code. I thought, "what if one area has a better 'track record' of winning? I could go buy tickets there!" Honestly, I know that winners are chosen at random and that the distribution should also reflect the population (more people, more ticket purchases, thus more winners), so I really didn't consider moving.
Data Collection
First, I needed to collect the data. I wrote a Perl script to make requests to the website and collect the number of matches returned for each ZIP code.
download Perl script "Lotto Grabber"
The software returns a tab-delimited file. I loaded the file into Microsoft Excel, and saved the information as a DBase IV file. I then added the database to ArcView. ESRI has a media kit with state information such as ZIP code boundaries as polygons. I can use the polygon layer, but I am unable to distribute it. I will get census-based ZIP code data and attempt to recreate the data with a distributable shapefile.
Download DBase Tables for ArcView (right-click and 'Save as...')
- pick_six_zips.dbf (8 kb)
- instant_win_zips.dbf (8 kb)
Constructed Data
 |
Lottery Winner Distribution Dot density map of Pick Six and Instant Win winners. Distribution correlates to New Jersey population. |
 |
Location Quotient Chloropleth Location quotient shows the ratio of per-ZIP code winners to total winners divided by the quotient of the population of that ZIP and the total population of the state. High values (darker colors) show areas where there are an above-average number of winners. This doesn't mean you'll win if you buy a ticket there. |
Conclusion
This is really just a fun exercise in what you can do with a simple idea and fun tools.
New Jersey Housing Trends
US Census
Initial Concept
The Census has a lot of information. Literally, tons of information. I was poking around the Census website and I found that you can download housing data for "county subdivisions" (municipalites, in NJ's case). There were some interesting attributes recorded, like number of bedrooms and such.
Data Collection
Getting data from the Census is a pain. Not because they make it difficult to obtain, it's just that there is so much information it becomes unmanageable.
The process is this: you fill out a web-based form in the American Fact Finder and the information will be presented to you in an HTML table. ArcView needs this information in either a Microsoft Access database (the new "Geodatabase" format) or a DBase database (Shapefile format). The Census allows you to download the information as a Microsoft Excel spreadsheet. From Excel, you can manipulate and reformat the data so that you can export it as a DBase file.
I have a shapefile and seperate databases for joining included in this ZIP file (320 kb).
Constructed Data
 |
Recently Constructed Houses Chloropleth Shows the percentage of total households in a municipality that were built sometime between 1990 and 2000. |
 |
Percentage of Households with Five or More Bedrooms
Chloropleth Shows the percentage of total households in a municipality that have five or more bedrooms. |
Conclusion
This is barely scratching the surface of what the Census has to offer.
Percentage of Adult Smokers by State
MSN Health
added 20 January 2004
Initial Concept
I was browsing the web and I came across a writeup about how smoking varies by state. The page included a table, the data in which would be better represented through the use of a map.
Constructed Data
I retyped the HTML table into a DBF and joined it to the Census's state boundary file to produce the map.
 |
Percentage of Adult Smokers by State Chloropleth Shows the percentage of adult smokers in each state. |
Conclusion
Look at the table, then view the map. It would take you a while to see that there's a concentration of adult smokers around Kentucky (the highest at 32.6%) and West Virginia using only the table. The map shows that instantly.
John J Reiser
newrisedesigns dot com