How To Spot A Made Up Chart
And here I thought I was the king of the made up chart. After my master class in fake charting in the blockbuster post; Thinking About Constraints On Partially Simulated Universe Scenarios, I just assumed the other pretenders to the fake chart throne would pack it in and go home. Apparently we have a young upstart out there who things he can go toe to toe with diamond d in the world of fake chartery, and guess what people, I gotta hand it to the kid, he’s the real deal. Clearly he knows his way around the absolute basics of Microsoft excel, including the use of error bars. Damn! It took me years to figure out that adding error bars to any chart consisting of totally made up data based on wild ass guesses would lend it a certain level of gravitas that you can’t get from just using a square for the data points instead of the default diamond. Have a look at this beauty.
What exactly is the source of the “data” for this “chart” you ask? No doubt rigorous double blind case control studies were conducted in leading research laboratories around the world. Then the data was compiled and analyzed by a different set of researchers to prevent bias, finally it was reviewed by yet a third group before finally getting published. Or the guy from the website sent out an email and asked a bunch of people what they thought. He then put the numbers in an excel spreadsheet, averaged them, and charted it.
And this is the MIT Technology review, I mean wtf MIT. Yours truly had his first brush with fame when some work he was doing in graduate school was featured in a short blurb in this very magazine. I still keep the copy I have. And wtf computer science, computer scientists, am I going to have to strip you of the right to the word science after computer?. You do not want to end up in the same category as data “science” do you? Not even a science, just a tool of science?