Deception isn’t always malice. Sometimes it’s just a structural trick. Or a phonetic illusion.
Today we are playing with three types. Statistics. Surveys. Sound.
Can you solve them?
Super Syllabus
Start with a school. Two classes. End of year 1 arrives. Every pupil gets graded. We line them up. Best to worst. Middle to the median. It sits on a C.
Solid. Unremarkable.
Then comes Year 2. The school swaps out the curriculum. New syllabus. Fresh start. At year’s end, they grade them again. We sort the list. Middle to the median. Now it is a D.
Worse. Obvious right? The new syllabus failed. It dragged everyone down.
Wait.
What if the new syllabus actually improved every single pupil’s grade?
Think. The median dropped. The middle fell. Yet every individual got better. How do you pull a D into a C or a B when the median slips downward? It is possible. You just have to manipulate the shape of the data. Not the performance.
“The median is the middle value.” Not the average. The middle.
Try it. Devise the scenario. Make every student smarter but make the school look worse.
Peculiar Poll
Two companies. Smith Surveys and Jones Polls. They both ask the same question. Do you like the new government policy?
Each polls 125 people.
Here is the data from Smith :
– Men support it: 21 out of 25. That is 84%.
– Women support it: 80 out of 100. That is 80%.
Men win here.
Here is Jones :
– Men support it: 22 out of 100. A dismal 22%.
– Women support it: 5 out of 25. Even lower at 20%.
Men win here too.
So… men love the policy. Both companies say so.
Or do they?
Let’s aggregate. Smash them together.
Total men polled: 25 + 100 = 125. Total yes: 21 + 22 = 43.
43 / 125 = 34.4%.
Total women polled: 100 + 25 = 125. Total no: 80 + 5 = 85.
85 / 125 = 68%.
Women actually support the policy by nearly double the rate of men.
How? Both polls showed higher support for men individually. But combined? The result flips completely. Men drop behind.
It is not a contradiction. It is a weighting trap. One group is small in one poll. Huge in another. When you lump them together the big numbers drown the small ones out. The aggregate lies about the components. Or vice versa. Which one is real? The group? Or the whole?
Anguish Languish
Forget numbers for a moment. Let’s try words.
There is a thing called Anguish Languish. Invented by Howard L. Chase. An American linguist with too much time and a good sense of irony. He takes English sentences and rewrites them with English words that sound like the original but mean garbage.
Read this aloud. Do it quietly so your neighbor doesn’t call the police.
“Ones her punnet I’m inner smell vial itch their lift a misty verse boy culled Pitter.”
Say it fast. Smooth. Connected.
If someone hears that without seeing it… they think they heard: “Once upon a time, in a small village, there lived a mischievous boy named Peter.”
Your brain fills the gap. The audio cortex ignores the semantic nonsense because the phonetic path is clear. It hears meaning where there is only noise.
Kit Yates wrote that example. It appears in his book You Don’t Know What You Miss. He explores the hidden gaps in data, surveys, and language. Things we assume we understand.
There is a prize. Send me a sentence. Write it in Anguish Languish rules. Use only common English words. Make it funny. If I like it… I’ll give you a copy of Kit’s book.
Deadline is 4 PM today. UK time. I will pick a winner and post some favorites at 5. I’ll also post the solutions to the syllabus and poll problems then.
Until then? Just try listening. Not reading.
Did you catch the trick in the median drop?
