Chapter 3 Quiz: Randomness Is Real

15 questions. Answers are hidden — reveal them only after you've committed to your own answer. Be honest with yourself: the goal is to identify gaps in understanding, not to feel good about guessing.

Instructions: For each question, read carefully, write down your answer (or think through it fully), and then reveal the answer. If you got it wrong, write a brief note about where your thinking diverged.

Question 1

Which of the following best describes ontological (true/intrinsic) randomness?

A. A coin flip where the exact physical forces are too complex to calculate B. A quantum event where complete prior knowledge cannot predict the outcome C. A social media algorithm whose logic is hidden from users D. A stock market move driven by millions of uncoordinated human decisions

Reveal Answer **Answer: B** Ontological randomness means the unpredictability is a feature of reality itself, not of observer ignorance. Quantum events (like radioactive decay) are the paradigm case — no additional information, in principle, would resolve the uncertainty. Option A describes deterministic chaos (practically unpredictable, but in principle determined). Options C and D describe systems that are complex and practically unpredictable but are ultimately driven by deterministic physical and human processes.

Question 2

Nadia posts two videos: the polished 11-hour video and the spontaneous squirrel video. She concludes that spontaneous content performs better than polished content on her platform. What is the primary methodological error in her reasoning?

A. She is using qualitative rather than quantitative evidence B. She is drawing a conclusion from an insufficient sample size C. She is ignoring that video quality cannot be measured D. She should have posted both videos on the same day

Reveal Answer **Answer: B** The primary error is insufficient sample size — drawing a general conclusion ("spontaneous content outperforms polished content") from two data points. Two observations cannot reliably distinguish between "spontaneous is genuinely better," "this was a lucky draw for the spontaneous video," "this was an unlucky draw for the polished video," and many other explanations. This is the coin flip illusion applied to content strategy. Options A, C, and D describe real considerations but are secondary to the fundamental sample-size problem.

Question 3

The Law of Large Numbers states that:

A. Large numbers of people always produce better decisions than small groups B. The average of many random trials converges to the true expected value C. Extreme outcomes become impossible with enough repetition D. After a run of bad luck, good luck becomes more likely

Reveal Answer **Answer: B** The Law of Large Numbers is a mathematical theorem stating that the sample mean converges to the population mean (expected value) as sample size increases. Option A describes a different concept (sometimes called "wisdom of crowds," which is not always true). Option C is incorrect — extreme outcomes remain possible in principle with any number of trials; they just become proportionally rarer. Option D describes the gambler's fallacy, which is explicitly *not* what the law of large numbers implies.

Question 4

In the Salganik, Dodds, and Watts music lab experiment, the researchers ran eight parallel social worlds. What was the primary reason for running eight parallel worlds rather than one social influence condition?

A. To increase statistical power by having more participants B. To demonstrate that the same songs could have different fates under different random initial conditions C. To test eight different types of social influence D. To ensure that no single song could dominate all conditions

Reveal Answer **Answer: B** The parallel worlds design was the core of the experiment's logic. If the same songs could end up as hits in some worlds and flops in others — which they did — this demonstrates that success is path-dependent on random early conditions, not solely determined by quality. Without parallel worlds, you couldn't show that the *same song* had *different outcomes* based on initial randomness. Option A is a benefit but not the primary reason. Options C and D do not accurately describe the design.

Question 5

Which of the following is an example of apophenia?

A. A researcher finds a genuine correlation between social media use and anxiety B. A gambler believes that a slot machine is "hot" after watching it pay out twice C. A creator accurately identifies that their videos perform better when posted on Thursday mornings D. A statistician recognizes that a data set is normally distributed

Reveal Answer **Answer: B** Apophenia is the perception of meaningful patterns in data that does not actually contain such patterns. The gambler observing two slot machine payouts and concluding the machine is "hot" is inferring a pattern (the machine's elevated payout probability) from a tiny, insufficient sample. Options A, C, and D describe correctly perceiving genuine patterns — the opposite of apophenia.

Question 6

A friend argues: "If the universe is deterministic — if every event is caused by prior events according to fixed physical laws — then randomness doesn't really exist and luck is just ignorance." Which of the following is the strongest counter-argument from the chapter?

A. Modern physics has proven that the universe is not deterministic at any level B. Even in a fully deterministic universe, some systems are so sensitive to initial conditions that prediction is permanently and practically impossible C. Free will is real, and determinism contradicts the experience of making choices D. Randomness exists because there are too many variables to ever count

Reveal Answer **Answer: B** The chapter presents three counter-arguments to hard determinism: (1) practical unpredictability is permanent even in a deterministic universe due to chaos, (2) quantum mechanics provides empirical reason to doubt hard determinism, and (3) the source of unpredictability doesn't change the mathematical tools we use to analyze it. Option B captures counter-argument #1 — deterministic chaos — which the chapter identifies as a particularly powerful response because it concedes the determinist's physics and still shows that practical prediction is impossible. Option A overstates the physics (quantum mechanics is contested and doesn't straightforwardly refute all determinism). Option C appeals to intuition rather than evidence. Option D is informal and imprecise.

Question 7

Marcus argues that some content creators are consistently successful over many years, which he takes as evidence that skill, not luck, determines long-run success. Dr. Yuki agrees. But what critical concept does Marcus's argument still fail to address?

A. The fact that creators who survive long-term have lower quality than those who don't B. That we only observe creators who persisted, not the equally-skilled creators who quit or never gained traction C. That long-term success proves that the algorithm is not random D. That skill is impossible to develop in creative domains

Reveal Answer **Answer: B** The concept Marcus's argument misses is survivorship bias. We observe consistent long-term successes precisely because they survived — but we don't observe the population of equally or more skilled creators who had worse early luck, didn't get early amplification, and either quit or never gained sufficient traction to be observed. The consistently successful creators we can see may be the skill-*and*-luck winners; there may be a much larger group of skill-only winners who we never observe. This doesn't prove that skill doesn't matter; it complicates the inference from "observed long-term success" to "success was determined by skill rather than luck."

Question 8

You flip a fair coin 99 times and get 99 heads. What is the probability the 100th flip will be heads?

A. Much higher than 50% — you're clearly on a streak B. Exactly 50% — each flip is independent C. Much lower than 50% — tails is "overdue" D. It depends on how hard you flip

Reveal Answer **Answer: B** Each flip of a fair coin is an *independent* event — the coin has no memory of previous flips. The probability of heads on flip 100 is exactly 50%, regardless of how many heads came before. Options A and C describe the hot hand fallacy and the gambler's fallacy, respectively — both are incorrect. Option D (while mechanically interesting in the realm of deterministic chaos) misses the point of the question, which is about the statistical independence of trials. Note: in practice, if you observed 99 consecutive heads, you would have overwhelming evidence that the coin is *not* fair, and you should update your belief about the 100th flip accordingly — but the question specifies a *fair* coin, so the answer remains B.

Question 9

The concept of "cumulative advantage" in the Salganik/Watts music lab means:

A. Talented musicians gradually accumulate advantages over less talented ones B. Early random success generates more success through social proof and algorithm amplification C. Advantages in one domain (music) transfer cumulatively to other domains D. The most-played songs cumulatively become better through more exposure

Reveal Answer **Answer: B** Cumulative advantage (also called "the rich get richer" or the Matthew effect) in this context means that early success — even when random — generates further success. A song that gets a few early downloads shows up higher in listings, gets more exposure, gets more downloads, rises further, and so on. The advantage compounds. This is what created the chaotic cross-world variation in the experiment: whichever song got early (random) advantages kept accumulating them. Option A describes merit-based accumulation, not the dynamic the study found. Options C and D mischaracterize the concept.

Question 10

Which of the following statements best describes the practical implication of the individual vs. ensemble prediction distinction for a content creator?

A. A creator should try harder to predict which individual video will go viral B. A creator should evaluate their strategy based on aggregate performance across many videos, not individual results C. A creator should focus only on consistency, not quality, because quality is irrelevant D. A creator should post randomly with no strategy because outcomes are random anyway

Reveal Answer **Answer: B** The individual vs. ensemble distinction implies that individual video performance is substantially noise, while aggregate performance across many videos is a more reliable signal of the underlying strategy's effectiveness. Option A is what most creators actually do, and what the chapter argues against. Option C is incorrect — quality raises the distribution's floor, it's just that quality alone doesn't determine any individual outcome. Option D is the nihilistic misreading that the chapter explicitly warns against.

Question 11

Regression to the mean (introduced briefly in this chapter, covered fully in Chapter 8) predicts that:

A. After many trials, average outcomes will regress toward zero B. Extreme performances tend to be followed by more average performances, due to the random component reverting C. Good performers eventually regress to the level of their peers as competition increases D. Random processes always return to their starting point

Reveal Answer **Answer: B** Regression to the mean is the statistical phenomenon where extreme performances — high or low — are partly produced by random variation, which is equally likely to be favorable or unfavorable in subsequent trials. So extreme outcomes are followed by less extreme ones not because of any causal mechanism, but simply because the random component is less likely to be extreme again. Option A describes something else. Option C describes competitive dynamics, not statistical regression. Option D is false — random processes have no "memory" of a starting point.

Question 12

Why did the original Gilovich, Vallone, and Tversky hot hand study find no evidence for the hot hand in basketball shooting data?

A. Professional basketball players don't actually have "streaks" in their shooting B. The researchers used flawed statistical methods C. Shooting sequences in the data were consistent with what would be expected from a fixed probability process with random variation D. The basketball courts used in the data had different surfaces that affected shooting

Reveal Answer **Answer: C** The researchers found that the sequences of makes and misses in the data looked statistically similar to what a random process with a fixed success probability would produce. Our brains interpret runs (streaks of makes) as evidence of a changing underlying probability, but statistically, runs appear frequently in random sequences — more often than our intuition expects. The study's finding was not that players are perfectly random, but that the apparent streaks were consistent with randomness. Option B is part of the ongoing methodological debate but not the primary finding. Option A misstates the claim. Option D is irrelevant.

Question 13

Priya's realization — "Randomness isn't an excuse. It's a map." — is best interpreted as meaning:

A. Accepting random luck removes personal responsibility for outcomes B. Understanding the random structure of a system tells you where to direct your effort for maximum impact C. Random processes are beyond human influence and should be accepted as they are D. Luck and strategy are opposites; choosing one means rejecting the other

Reveal Answer **Answer: B** Priya's insight captures the practical value of understanding randomness: knowing *how* randomness operates in a system (job searching, in her case) tells you where human agency has leverage (improving the seed conditions, increasing the number of draws) and where it doesn't (the specific decision any given hiring manager makes in the moment). A "map" guides action; it doesn't eliminate it. Options A and C describe fatalistic misreadings. Option D presents a false dichotomy that the chapter explicitly rejects.

Question 14

"The unit of strategy should be the portfolio, not the post." This advice is most directly an application of which principle?

A. The gambler's fallacy B. Cumulative advantage C. Individual vs. ensemble prediction D. Deterministic chaos

Reveal Answer **Answer: C** The advice follows directly from the individual vs. ensemble prediction distinction. If individual posts are high-variance and substantially noise, evaluating and optimizing any single post is low-leverage. Instead, the strategy should be designed around producing a good *distribution* of outcomes across many posts — the ensemble. Individual draws from that distribution will vary widely, but the aggregate will reflect the underlying strategy's quality. Option B (cumulative advantage) is relevant to understanding *why* individual posts succeed or fail, but it doesn't generate this specific strategic advice. Options A and D are not the relevant principles.

Question 15

A content creator with a consistent, high-quality posting strategy is best understood as:

A. Increasing the probability that any specific video will go viral B. Guaranteeing success through sufficient effort C. Improving the distribution of outcomes across many posts, while accepting that individual results remain substantially unpredictable D. Eliminating the random component of the algorithm

Reveal Answer **Answer: C** A consistent, high-quality strategy raises the expected value — the mean of the outcome distribution — which means better average results across many videos. It does not guarantee any specific outcome (Option B), does not make individual videos predictable (Option A is close but slightly wrong — quality doesn't increase the probability that *a specific* video goes viral so much as it improves the distribution across videos), and cannot eliminate algorithmic randomness (Option D). Option C is the precise, technically accurate statement of what strategic effort accomplishes in a high-variance, luck-influenced domain.

Score Interpretation

13–15 correct: Strong conceptual grasp. Proceed with confidence to Chapter 4.

10–12 correct: Good foundation with some gaps. Review the sections connected to the questions you missed before moving forward.

7–9 correct: Revisit the chapter with the missed questions in mind. Pay particular attention to the individual vs. ensemble distinction and the coin flip illusion.

Below 7: Consider rereading the chapter fully, then retake the quiz. These concepts are foundational — they underpin every subsequent chapter in this book.

Note to self: The questions you got wrong are the most useful information this quiz can give you. Resist the temptation to move on immediately. Sit with what you didn't know.