Card stacking vs Cherry-picking

What is the difference, when does each apply, and can both be at work at once.

Classic

Card stacking

Presenting only the facts, figures, and quotations that favor one side, and leaving out everything that doesn't, so that a true selection produces a false impression.

Framing and omission

Cherry-picking

Presenting only the data points, studies, or cases that support a conclusion while ignoring the larger body of evidence that would undermine it.

The differences

Card stackingCherry-picking
ScopeThe whole presentation: facts, arguments, quotations, images, all selected for one side.The evidence in particular: data points, studies, or cases chosen to support a conclusion.
OriginOne of the seven devices of 1937, named for a rigged deck.A general term for selective use of evidence, common in science and policy argument.
Typical settingA speech, a brochure, a documentary, a campaign.A statistic, a chart, a literature review, a "studies show" claim.
What is hiddenThe other side's case, in full.The rest of the data, which would change the picture.
The counter-moveAsk what a fair opponent would say and whether you have been given any way to find out.Ask what the full range of evidence shows and who chose the sample.
The tellEverything points one way and nothing is conceded.A specific number or study carries more weight than the body of evidence would.
An example of card stacking

A company's annual report on a product recall mentions the swift response, the customer satisfaction scores, and the industry award, and does not mention the injuries that caused the recall.

An example of cherry-picking

A climate skeptic charts global temperature from 1998, an unusually hot year, to 2012, and shows a flat line. The same chart from 1990 or 1980 slopes steeply upward.

Can both apply?

Cherry-picking is one of the tools of card stacking, so the two often appear together: a stacked presentation will usually include some cherry-picked evidence. They are kept apart because the remedies differ in scale. Against cherry-picking, you need the rest of the data. Against card stacking, you need the other side's whole case, which may have no data in it at all.