the second hundred · metaphor 106

Experience oversamples
the big.

Why does everyone seem busier than average, the gym more crowded than its real usage, your friends more popular than you?

Because experience does not sample life fairly. It oversamples the big, the long, and the crowded — in exact proportion to their size.

Stand at a stop where buses come every ten minutes on average, and you will wait, on average, more than five. Enrol in a university where the mean class holds thirty students, and the class the average student actually sits in is bigger than thirty. Nothing is broken and no timetable is lying; you are simply counting in different units than the schedule does.

Then there is the quiet insult of the friendship paradox: on almost any real social network, most people have fewer friends than their friends do. Buses, classrooms, and popularity are one trick wearing three costumes — length-biased, or more generally size-biased, sampling. Meet the world by the moment instead of by the event, and big things, reaching across more moments, get met more often.

A schedule of gaps between buses — each block is one gap, its width its length

start of the houra random moment lands anywhere along this strip
your measured wait: min
drop yourself in
True average gap
(the schedule)
min
every gap counted once
Average gap you land in (length-biased)
min
Your expected wait to the next bus
min
How many gaps there are — every gap counts once (the schedule’s view) mean min
How much waiting each gap causes — counted by length (your lived view) mean min

A random moment is equally likely anywhere on the strip, so it falls inside a wide block far more often than a narrow one. Widen the spread and watch your expected wait pull away from half the average.

Every number here is computed live from the actual gaps and the actual graph in your browser. Nothing is pre-baked.

experience oversamples the big

Size buys exposure.

The schedule and your body count in different currencies. The schedule counts events — each bus gap is one item on a list, and the average gap is the plain average of those items. You count moments. A ten-minute gap contains twice as many moments as a five-minute gap, so if you show up at a random instant you are twice as likely to be standing inside it. Every interval is weighted by its own length before you ever meet it. That is length-biased sampling: the probability of experiencing a thing is proportional to its size.

Once weighting enters, the average you live inside is not the average that was quoted. The size-biased mean overshoots the plain mean by Var / mean — the whole gap between the brochure and the experience is manufactured by variance. And the wait is worse still: land in a long gap and, on average, you land halfway into it, so your expected wait is not half the average gap but half the length-biased gap. Uneven buses punish you twice — you are more likely to be inside the long wait, and then you still have to wait out its second half.

what to try

Turn the two dials.

A · WAITING

Widen the spread

Push the spread slider up and hold your eye on your expected wait. The true average gap never moves — it is pinned at ten minutes throughout — yet your expected wait balloons well above five. Slide it back to zero and the two numbers snap together: no spread, no penalty.

B · FRIENDSHIP

Sample a random friend

Follow a friendship, again and again, and watch the running average of the friend’s friend-count settle above the average person’s. Then raise connectivity and drop it: the gap tracks the spread of popularity, not its level.

why life feels like more

You are almost always inside someone’s big interval.

This is why the felt world runs richer than the measured one. The crowded hours are, by definition, when most people are present to feel the crowd, so the average person’s experience of the gym is drawn from its fullest moments, not its empty ones. The big class is where most of the enrolled student-hours happen, so the typical student’s day is built from large rooms. The popular friend appears in the most address books, so a randomly recalled acquaintance skews sociable. Sample by the moment, the seat, or the friendship, and you weight every event by how much of the world it occupies.

So the loneliness and the overwhelm are, in part, an artefact of the sampling. Your friends really do have more friends than you — and so do almost everyone’s, simultaneously, with no contradiction, because the popular are counted many times over in everybody’s tally. The average was always a fact about the list of events. Your life is a fact about the moments, and the moments live inside the big intervals.

the honest model

The rule is short: bias in proportion to size.

Give each interval a length L. Sample the world not by picking an interval at random but by picking a random moment — and the chance of finding yourself in interval i is its share of all the moments, Li / ΣL. Everything downstream follows from that one weighting.

P(you land in gap i) = Li / Σj Lj
Size is probability. A gap twice as long is twice as likely to be the one you are standing in.
gap you land in = E[L²] / E[L] = mean + Var/mean
The size-biased average. It exceeds the plain average by exactly the variance over the mean.
your expected wait = E[L²] / (2·E[L])
Half of the gap you land in — never merely half of the average gap, unless every gap is equal.
friend’s friends = E[d²] / E[d] = mean + Var/mean
The same formula with degree d in place of length: a friend is a person size-biased by popularity.

One mechanism, two costumes. Whether the size is a duration, a headcount, or a friend-count, the biased average sits above the plain one by Var / mean, and vanishes precisely when the variance does.

the mapping

Mathematics ↔ life.

MathematicsLife
a long intervalA big class, a long wait, a busy stretch — anything whose size is its reach into your day.
length-biased samplingExperiencing life by the moment, not by the event — so size quietly buys exposure.
expected wait E[L²]/2E[L]Why you always seem to just miss the bus and stand there longer than the timetable promised.
a popular nodeThe friend everyone counts — present at the end of many friendships at once.
the friendship paradox“Everyone has more friends than me” — felt by nearly everyone at once, and not a lie.
the true meanThe fair average you keep hearing quoted and never actually get to live inside.

where the metaphor tears

Three honest failures.

Real crowding is also real.

Size-biasing is not always the story. Sometimes the gym genuinely is oversubscribed and the class genuinely is too large. The paradox explains a gap between the quoted average and the lived one; it does not prove the average is the truth and your experience the illusion. Both can be real at once, and the honest move is to ask how much of the felt excess is sampling and how much is the world.

Equal intervals, no paradox.

The effect is powered entirely by variance. When every gap is the same length — or every person has the same number of friends — the biased mean equals the true mean and your expected wait is exactly half. Slide the spread to zero and watch the whole phenomenon disappear. It is not a law of nature; it is a law of unevenness.

Knowing it doesn’t shorten the wait.

Understanding why you wait longer than five minutes buys you nothing at the bus stop. Size-biasing is a fact about how you sample the world, not a mistake you can correct by noticing it. The only consolation on offer is that you were never quite as unlucky, unpopular, or singularly busy as it felt — you were standing, like everyone, inside the big intervals.