The trick with misinformation isn't that people are lying to you. The trick is that they either tell accurate facts while withholding additional necessary information or they tell accurate facts while greatly implying things that aren't entirely accurate.

Take this chart from CNBC and Citi Research. It first appeared in Feb-2013 as part of a Patti Domm's overly optimistic piece: "US Is on Fast-Track to Energy Independence: Study." Now, fresh off CERAWeek, the energy optimism Super Bowl, Domm is back with the first in a four-part NBC news series examining the "energy boom dawning in America." Right on cue, a version of this same chart appears in the NBC news piece.

So what's so bad about this chart? First, let's be clear - everything in this chart is 100% true. The problem isn't the data, the problem is the chart itself.

There's absolutely no reason for two Y-axis: If you handed this oil production/consumption/import data in spreadsheet form to a room full of science, math, and engineering undergrads - then asked them to chart the data - none of them would come up with the chart CNBC produced. None of them.
Both Y-axis use the same unit of measure m b/d or million barrels of oil per day. Because they're the same unit, the proper way to chart this is to use one axis. The problem? Using the same axis for all three lines would show the REALITY that consumption is much much larger than production.

At some point, someone in the CNBC decision-chain made the conscious decision to go the extra step of using a second axis.
The 'scale' of the two axis are different: The right axis has a much smaller m b/d range from one end to the other. The result is to make the consumption line look pretty flat, while exaggerating the big swings of the production line.
They never bother to tell the reader which lines correspond to which axis: If you're familiar with the energy industry, you probably know which line goes with which axis, but most normal people will have zero idea - that's the point. For most people - those that will spend half a second at most glancing at the chart, without context for the values, the numbers might as well not even be there.

In the end, to the untrained eye, you get a carefully crafted chart that suggests a few things to the reader: It suggests that US oil production has somehow caught and passed US oil consumption - a ridiculous notion. Second, because neither axis begins at zero, it also suggests to the reader that US oil imports are not only low, but nearing zero.

Can one prove that this is the intent of the authors? No, of course you can't. That's the ultimate trick, they can always fall back on the argument that "Everything in the chart is factually true."

But if they wanted to be free of suspicion all they needed to do was simply create the chart the normal way - the same way any undergrad would do it.