GraphJam Side A · Music & pop culture in charts
The Plate Archive17 charts

Probability and Paradox, in Charts

The collection where the joke is carried by an actual calculation.

Seventeen charts, thirty links from other sites, and a distinctly higher hit rate than the archive average. The reason is simple and it is the most useful thing this archive demonstrates: a joke with a real calculation inside it gets forwarded, because the calculation can be disputed.

The paradox pair

Two plates here, published six days apart in September 2009, are linked from five sites each, the highest count in this collection, though several charts elsewhere in the archive hold more. They are a complementary pair of Venn diagrams about the dating pool, and Discover Magazine covered both, taking the later one first, which is why several accounts reverse the order.

What they share with the rest of this collection is the mechanism rather than the subject. A three-set Venn diagram sends the reader looking for the region where every condition holds at once, and in both of these plates the whole argument is what is waiting in that region. Each is published individually as a plate rather than as an entry here.

Volume paradox and Monty Python Paradox, the latter filed under films, are the looser members of the same family, and Headphone Paradox, filed under the machines, is the one that names a real physical annoyance rather than a logical one.

Conditionals, drawn properly

Chance of people believing an outrageous statement, published as its own plate, plots credence against the confidence of delivery. Chances The Stranger You Just Spoke to is Jewish is a base-rate chart, and it is drawn as one. Probability Of Girls Crying and Probability of Getting Things both plot conditional likelihood against a stated circumstance.

Chances that the bad guy you just shot a bunch of times is really dead. is the best-constructed of them: a decaying probability against number of shots, applied to a genre convention that ignores probability entirely.

Assertions dressed as findings

The other half of the collection asserts what it cannot know, and does so in statistical costume. Average Accuracy of Adult Male announces a population and a measurement and supplies neither. Difficulty of Task as Perceived by the Average Person invokes an average person, which is the standard move for making an opinion sound sampled. Polydactylism vs Being Killed is linked from three sites, a correlation with no plausible mechanism, which may be the point.

Outcome Analysis for Alternative Options is the most bureaucratically titled chart in the archive and is linked from two sites. Desirability of Splits in selected situations uses “selected situations” the way real papers use it, to mean the ones that worked.

The methodological jokes

The Kniviness of a Small Sampling of Knives is the collection’s quiet masterpiece. Every word in the title is a methodological hedge, a small sampling, of an invented quantity, reported as a finding, and the chart delivers exactly what the title promises. It is a parody of a paper abstract that predates most of the criticism of paper abstracts.

Percentage of this chart that looks like the blue man group is the same joke turned inward: a chart whose only variable is its own appearance. Likelihood of Abduction of Royal Family is the earliest entry, from March 2008, and states a risk assessment for an event with no historical base rate at all, which is what most published risk assessments of rare events actually do.

The distinction the collection keeps returning to, a calculation you can check against one you cannot, is the same line the annual draws between its two halves. Where the outcomes really are countable, the odds of the common games do the arithmetic in full.

Charts here
17
Images recovered
9
Published as plates
3
Referring domains
30
First posted
Mar 2008
Last posted
Mar 2010

The charts

In publication order. Sides run in tens, the way a record does. A chart marked as a plate has its own page and its own retrospective; the rest are published here, at the addresses they held when they were posted.

  1. A1

    Likelihood of Abduction of Royal Family

  2. A2

    The Kniviness of a Small Sampling of Knives

  3. A3

    Percentage of this chart that looks like the blue man group

    · Drawn by CARLWeathers54 · image not recovered

  4. A4

    Probability Of Girls Crying

    · Drawn by Eric C

  5. A5

    Polydactylism vs Being Killed

    · Drawn by kzingirl · 3 domains still link here · image not recovered

  6. A6

    Probability of Getting Things

    · Drawn by Heather

  7. A7

    Outcome Analysis for Alternative Options

    · Drawn by Jason Fahy · 2 domains still link here · image not recovered

  8. A8

    Desirability of Splits in selected situations

    · Drawn by gylbertpenguin · image not recovered

  9. A9

    Chances The Stranger You Just Spoke to is Jewish

    · Drawn by HarleeKwinn

  10. A10

    Chance of people believing an outrageous statement

    · Drawn by OriginalObe · 3 domains still link here · published as a plate

  11. Side B
  12. B1

    Average Accuracy of Adult Male

    · Drawn by jaredcrich · image not recovered

  13. B2

    Difficulty of Task as Perceived by the Average Person

    · Drawn by GrinningColossus · image not recovered

  14. B3

    The Girl Paradox

    · Drawn by gilgatex · 5 domains still link here · published as a plate

  15. B4

    The Boy paradox

    · Drawn by dvhk · 5 domains still link here · published as a plate

  16. B5

    Volume paradox

    · image not recovered

  17. B6

    Chances that the bad guy you just shot a bunch of times is really dead.

    · Drawn by JohnnyReb27

  18. B7

    Chance of Ideas vs. Where They Happen

    · Drawn by Plixxo · image not recovered

Queries answered

What is in this collection?

Seventeen charts that use probability, conditional reasoning or a stated paradox as their mechanism, drawn between March 2008 and March 2010. Three are published as their own plates.

What are the Boy Paradox and the Girl Paradox?

A complementary pair of Venn diagrams about the dating pool, published six days apart in September 2009 and linked from five sites each. Discover Magazine covered both, and covered the later one first, which is why several accounts reverse the order. They are published individually as plates.

Why do these charts travel better than the rest of the archive?

Because a calculation can be checked, and something checkable gets argued about. Across the whole archive the charts that escaped are consistently the ones doing real arithmetic about an unserious subject, and this is the collection where that is the explicit method.

Are the probabilities in these charts correct?

Some are. The paradox charts state a genuine result; the correlation charts assert relationships nobody measured. The collection is a mix, and reading it as a set is the best way to see which is which.

Why do some entries have no image?

Every original chart image was hosted on a service that no longer serves them, so a surviving copy had to be found for each one. Where none survived, the entry keeps its title, date and address rather than a substitute image.