Last week, I wrote about the fun and the pitfalls of viral maps, a feature that included 88 super-simple maps of my own creation. As a follow-up, I’m writing up short items on some of those maps, walking through how I created them and how they succumb (and hopefully overcome) the shortfalls of viral cartography.
As part of my initial story, I included this map about the game show Jeopardy!
Data source: Reddit user trexmatt via a data scrape of J! Archive. Map by Ben Blatt/Slate.
The map relies on data gathered by Reddit user trexmatt that is maintained by the fan website J! Archive. This data set includes more than 200,000 questions and answers, of which 4,790 contain a state name (not necessarily in reference to the actual state) in the answer. The frequency of appearances ranges from 306 for Washington to a mere 30 for North Dakota.
You won’t learn too much from the map above, other than the fact that contestants who answered “West Virginia” probably won more money than those who answered “Florida.” (Please note that clue values have shifted over time, so while $200 was never a top prize, it has not always been the bum amount it represents today.)
Before making the map of clue values, I tried creating a map that showed which category each state was likeliest to appear under. The result is below. But I’ll warn you that the data here is kind of crummy. One issue is the large variety of categories and how infrequently they are repeated. Georgia appears in the answer to 128 different categories—including Early America, Potpourri, and Could It Be … Satan?—of which 114 were unique. It appeared in “G”eography more than any other category, but this only happened three times.
Data source: Reddit user trexmatt via a data scrape of J! Archive. Map by Ben Blatt/Slate.
See more of Slate’s maps.
