Brad Feld’s review of Jeffrey Ma’s book House Advantage has us geeking out about leveraging statistics to do everything from improving ad relevance to selecting the optimal combination of caffeine drinks to keep us effectively hopped up throughout the day. It also inspired this week’s TWIR as we looked far and wide for people doing cool stuff around finding patterns out of data.
Some of you may remember Jeff Ma from the MIT Blackjack team covered in the best selling book Bringing Down the House or the less than stellar movie version, 21 (you know, the one where Kevin Spacey jumped the shark). Jeff’s book is a must read for anyone involved in online marketing. He covers the fundamentals of statistics and how you can leverage them to drive better business decisions. At first glance that may sound like a snoozer of read but Jeff brings the topics to life through blackjack, sports bookies and stories that make you want to hop on the next flight to Vegas, baby, Vegas.
Finding patterns out of data is a major theme in House Advantage and what better way to do that then by making pretty pictures. David McCandless’ recent TED talk in July is a great intro to how properly designed visualizations can help people make sense of huge data sets, and discern trends that might at first glance be obscured. He uses examples like Global Military spending and Facebook status updates. If you like what you see take the time to checkout his blog for more examples.
Another data visualizer (is this a new career path?) worth mentioning is Peter Warden. He’s got some great posts around data visualization including data about start-ups via Crunchbase (triggered by fellow Boulder resident Brad Feld) and relationships via 210 million public Facebook users, done geographically on a map of the United States. The Facebook data presents some very interesting qualitative conclusions on how the country is connected/divided, socially. Apparently we’re from Socalistan … that’s good right? Btw, if you want to draw your own pretty pictures with data checkout Peter’s cool OpenHeatMap tool.
Of course data visualization isn’t a recent phenomenon. If you want to really wow your friends (ok your nerdy friends) with a little data visualization history Shawn Allen offers up somehistorical context starting with William Playfair, whose Commercial and Political Atlas of 1786 was the first widespread distribution of line and bar graphs.
Because we do have a few art historians in the TWIR community we couldn’t help but end TWIR this week with a story of data as art. Japanese artist Isao Hashimoto has created avisualization and sonification of nuclear weapons testing in the 20th century. By rendering the different magnitudes of bombs as different tones, he creates a powerful representation of the Cold War’s horrific potential for destruction.
And lastly, we couldn’t help but cover the Printing Impressions review of Ad Tech London’s digital vs traditional advertising surveys. The delta of user perceived relevance between traditional and digital advertising is stark but not surprising given the self-diagnosed low levels of digital expertise by advertisers. We expect a natural maturation and improvement in advertising quality as marketers become more comfortable the with digital medium over time. That said, a shift in focus may be neccessary to truly grasp the vast potential of online advertising and capture user attention. As our own Dave Schwartz points out in the comments, to-date, many of the digital advertising industry leaders have spent their time with math models figuring our how to buy traffic more efficiently, as opposed to how to merchandise their products and services to their target audience. Digital advertising offers unparalleled controls to reach specific audience segments. Its time for advertisers to take advantage and focus on bringing the right, relevant, message to those audience segments. Visualizing this article re-written in 2 years with inverse statistics is what keeps us up at night. Not healthy, we know.