(Any excuse to make a spreadsheet.)
Since the election is over (as predicted, Obama landslide–yes, I am a happy snake) and I’m now lacking my daily dose of poll data, I’ve turned to crunching other data to support my statistical addiction. Obviously, the first data I looked at was the final election results so I could update my “blue shift” map. Naturally, this led to an examination of the differences between my previous map (from the last blog post), which used predictions based on polls, and the final results. The map this data produced basically only shows where polls were over- and understating support for a candidate. While I was at it (and because it was trivial to drop new data into my SVG generator program), I generated maps for the 2004 and 2008 final presidential election results. All four maps follow–click for the full image, obviously.
The aspect that jumps out the most in the first image (Blue Shift) is how Arkansas is the only state shifting decidedly more Republican; Louisiana and some other adjacent states are also slightly more red or neutral. As I alluded to last time, it appears that the demographics in these states are such that: A. there was profound white prejudice/misinformation against Obama and B. there wasn’t enough black/minority support to overcome this.
The poll performance map is even more interesting. Immediately, North Dakota and Wyoming stand out as performing way better for McCain than expected–I’m going to attribute these two to the limited number of polls. It’s also likely that Rhode Island, Vermont, and Hawaii fall into that same category except in the Democratic direction. Arkansas, as previously discussed, also under-performed prominently for Obama. But I think the biggest conclusion we can draw from this map can be seen if we look at the bigger picture. The Southeast and Central states are all mostly under-performers for Obama, while the western states over-performed (in general). Perhaps, when they get in the polling booth, rural westerners are generally more open-minded and less bigoted (and thus more accepting of a black president) than rural southerners? It could also say something about the turnout regionally. More information is required before we can make a decisive conclusion, but it’s interesting nonetheless.
It occurs to me that the maps aren’t particularly useful for seeing specific results. I considered rectifying that at one point but decided that the purpose of the maps was mainly to compare states to each other, which the maps do quite well. However, if you want to see the data I used, I uploaded the spreadsheet here (Excel 2002 format).
More recently, I’ve been studying the visitor demographics for my site–pondering my precarious IE support and wishing IE6 would just go away already. Though sadly, the data suggests that IE6 usage will persist. It’s like the herpes of web development. Anyways, I collected the data using this site’s raw logs, counting only distinct IPs for a particular browser (i.e. multiple hits are filtered out). Then, I made some pretty charts in Excel to show off the results. Nothing is really surprising here, but… it’s interesting nonetheless. 😉 Charts below.
As a status update, I’ve been working on my Deus Ex guide again and… yes… it’s almost complete 😮 . And I’ve only had to play DX what must be 12 times over 7 years to write it! I’m not complaining, though. Really, it feels like the definitive guide to Deus Ex on the internet and I’m quite proud. Look for an updated GameFAQs version soon.
Also, I’m working on a massive overhaul of the SnakeByte content system which will feature tabviews (emulated with clever HTML, DOM, JavaScript, and CSS, of course), more images, changelog support, and even some AJAX. It’s about 75% complete as of this post, with expected completion before the end of the year.
