Chapter 11 Quiz: The Birthday Problem and Other Probability Surprises
Chapter: 11 — The Birthday Problem and Other Probability Surprises Format: 15 multiple-choice questions with hidden answers Instructions: Commit to an answer before revealing. Write your reasoning first.
Question 1
In a group of 23 people, what is the approximate probability that at least two people share a birthday?
A) 6.3% B) 23% C) 50.7% D) 99%
Show Answer
**Correct Answer: C) 50.7%** **Explanation:** This is the core birthday problem result. With 23 people, there are 23×22/2 = 253 possible pairs, each with a 1/365 chance of sharing a birthday. The exact calculation via complementary counting yields approximately 50.7%. Most people guess 6% (the chance of sharing YOUR birthday specifically), which is the single-target error — comparing against one reference point rather than all pairs. Choice D (99%) is the probability for approximately 57 people. This problem is the foundational result of the chapter.Question 2
Which of the following correctly explains WHY the birthday problem result is so much larger than most people expect?
A) Our intuition overestimates how many possible birthdays there are B) We naturally think about whether someone shares OUR birthday, but the problem asks about any two people sharing a birthday — and the number of pairs grows much faster than group size C) People tend to cluster their birthdays in certain months, which increases collision probability D) We instinctively use the wrong formula (multiplication instead of addition)
Show Answer
**Correct Answer: B) We naturally think about whether someone shares OUR birthday, but the problem asks about any two people sharing a birthday — and the number of pairs grows much faster than group size** **Explanation:** The core intuition failure is a reference class error: we compare against one target (our own birthday: ~22/365 ≈ 6% for 23 people) when the correct calculation considers all possible pairs (253 pairs for 23 people). Number of pairs grows as n(n-1)/2, which is quadratic in group size — much faster than the linear growth of the group itself. Choice C is slightly true (actual birthday distributions have some clustering) but it's a minor effect that doesn't explain the main intuition failure. Choice D describes an arithmetic error, not the conceptual error.Question 3
In the Monty Hall problem, you pick Door 1, Monty opens Door 3 (revealing a goat), and offers you the chance to switch to Door 2. What are the correct probabilities?
A) Door 1: 50%, Door 2: 50% B) Door 1: 1/3, Door 2: 2/3 C) Door 1: 2/3, Door 2: 1/3 D) Door 1: 1/3, Door 2: 1/3, Door 3: 1/3 (unchanged)
Show Answer
**Correct Answer: B) Door 1: 1/3, Door 2: 2/3** **Explanation:** Your initial choice (Door 1) had a 1/3 probability of being correct. Monty's action — crucially, he ALWAYS opens a goat door, never the car door — carries no information that could change the probability of Door 1. But it eliminates Door 3, forcing all probability that was "not Door 1" onto Door 2. Since P(not Door 1) = 2/3, Door 2 now has P = 2/3. Choice A is the common wrong answer, arising from the (false) intuition that two remaining doors must each have 50% probability. This would only be true if Monty's door selection were random (he might have opened the car door), but it isn't — he always opens a goat.Question 4
Why does the Monty Hall answer change if Monty randomly opens a door (and just happens to reveal a goat by coincidence)?
A) It doesn't change — the math is the same whether Monty is deliberate or random B) In the random version, seeing a goat is additional information that updates all probabilities, and the two remaining doors become equally likely (50/50) C) In the random version, Door 1 becomes more likely because Monty avoided it D) The random version makes Door 2 more likely than 2/3
Show Answer
**Correct Answer: B) In the random version, seeing a goat is additional information that updates all probabilities, and the two remaining doors become equally likely (50/50)** **Explanation:** This is one of the subtler points in conditional probability. If Monty randomly opens a door and reveals a goat, that event has probability 2/3 (since 2 of 3 doors are goats). The fact that we're conditioning on this event (random goat revealed) changes the calculation. Using Bayes' theorem: P(car in Door 1 | random goat in Door 3) = P(car in Door 1) / P(goat in Door 3) = (1/3)/(2/3) = 1/2. Similarly for Door 2. So yes — in the random version, it really is 50/50! The key is that Monty's *knowledge and deliberate action* is what creates the 2/3 advantage for switching. Without that deliberate action, the information content changes.Question 5
The inspection paradox explains which of the following observations?
A) You tend to arrive at the bus stop just as the bus is leaving B) When you arrive at a random time, you're more likely to land in a longer gap between buses, making your average wait longer than the average gap would suggest C) The bus schedule is designed to create uneven intervals D) Buses that run late cause longer waits than buses that run early
Show Answer
**Correct Answer: B) When you arrive at a random time, you're more likely to land in a longer gap between buses, making your average wait longer than the average gap would suggest** **Explanation:** The inspection paradox is a sampling bias: longer intervals contain more "arrival moments" for a random observer, so you are systematically more likely to arrive during a long gap than a short one. This means the average gap length you *experience* is larger than the average gap length in the schedule — even if the schedule is completely regular by design. Choice A describes Murphy's Law-style bad timing (a psychological perception, not a mathematical result). The inspection paradox is mathematical, not about perception.Question 6
The friendship paradox states that:
A) People with many friends are less likely to consider others as true friends B) For almost any social network, most people have fewer friends than their friends have on average C) Social media creates friendships that are less satisfying than in-person friendships D) The number of friends you have correlates negatively with how lucky you are
Show Answer
**Correct Answer: B) For almost any social network, most people have fewer friends than their friends have on average** **Explanation:** The friendship paradox is a direct consequence of the inspection paradox applied to social networks. Highly connected people (popular nodes) appear in many people's friend lists. So when you average the connections of all your friends, that average is pulled up by the highly connected nodes who appear in your list precisely because they're highly connected. The paradox was proven mathematically by Scott Feld (1991) and has been empirically confirmed on Facebook, Twitter, and other platforms. It's why everyone on social media feels like everyone else has more connections — they're sampling a biased group.Question 7
The St. Petersburg Paradox demonstrates which of the following?
A) That people are irrational because they won't pay enough to play a positive-EV game B) That expected value calculated in dollars can be infinite while the rational amount a person would pay is finite — showing that utility-adjusted EV is more appropriate than raw dollar EV C) That probability calculations always break down at extreme values D) That casinos can always be beaten with the right strategy
Show Answer
**Correct Answer: B) That expected value calculated in dollars can be infinite while the rational amount a person would pay is finite — showing that utility-adjusted EV is more appropriate than raw dollar EV** **Explanation:** The St. Petersburg game has infinite dollar EV (each coin-flip sequence contributes $1 to the infinite sum). Yet most people rationally won't pay more than $20–$30 to play. The resolution is that utility is not linear in dollars: winning $2 billion instead of $1 billion adds essentially zero utility to any normal person's life. Once you apply a concave (logarithmic) utility function, the utility-EV of the game is finite and modest. This is why rational people refuse to pay large amounts — they are correctly calculating utility EV, not dollar EV. Choice A misidentifies the refusal to pay as irrationality; it is actually a correct application of utility theory.Question 8
A family has two children. You're told "at least one is a boy." What is the probability both are boys?
A) 1/2 (50%), because each child independently has a 50% chance of being a boy B) 1/4 (25%), because both being boys is just one of four equally likely outcomes C) 1/3 (33%), because knowing at least one is a boy eliminates the girl-girl case, leaving three equally likely scenarios, only one of which is boy-boy D) 2/3 (67%), because we know one is a boy so only the second child is uncertain
Show Answer
**Correct Answer: C) 1/3 (33%), because knowing at least one is a boy eliminates the girl-girl case, leaving three equally likely scenarios, only one of which is boy-boy** **Explanation:** The four equally likely two-child arrangements are: BB, BG, GB, GG. "At least one is a boy" eliminates GG. Of the remaining three equally likely arrangements (BB, BG, GB), only BB has both boys. Probability = 1/3. Choice A (1/2) is the most common error — applying independence of individual events to a conditional question. Choice D confusingly suggests 2/3 by implying one child is "known" and only the second is uncertain — but this applies if you know the FIRST child specifically is a boy (which gives P = 1/2), not if you only know "at least one."Question 9
If you know that in a family of two children, "the OLDER child is a boy," what is the probability that both are boys?
A) 1/3 B) 1/4 C) 1/2 D) 2/3
Show Answer
**Correct Answer: C) 1/2** **Explanation:** Knowing the OLDER child is a boy specifies a particular child. The possible arrangements become: B(older)B(younger), B(older)G(younger). These are the only arrangements consistent with "the older child is a boy." Of these two, exactly one has both boys. Probability = 1/2. Compare to the previous question (1/3): "at least one is a boy" doesn't specify which child, leaving three possibilities (BB, BG, GB). "The older child is a boy" specifies the older child, leaving only two possibilities (BB, BG). This is the boy-or-girl paradox: HOW you received the information (which child is specified vs. just knowing one is a boy) changes the conditional probability dramatically, even though the content sounds similar.Question 10
In a city with 1 million professionals, you estimate each person knows approximately 200 others in their field. At a professional event of 100 people, a "coincidence" (shared connection) is:
A) Extremely unlikely — professional networks are too large to overlap randomly B) Expected to be rare but occasionally happening C) Essentially mathematically inevitable — the number of pairs and connection density makes it almost certain D) Only likely if attendees are from the same company
Show Answer
**Correct Answer: C) Essentially mathematically inevitable — the number of pairs and connection density makes it almost certain** **Explanation:** With 100 people, there are 100×99/2 = 4,950 pairs. Each pair has roughly a 200×200 / 1,000,000 = 4% chance of sharing a connection (extremely rough estimate). Even with this rough approximation, the probability that at least ONE pair shares a connection out of 4,950 pairs, each with ~4% individual probability, is effectively 1 − (0.96)^4950 ≈ 1.0. The exact calculation confirms: shared connections at professional networking events are not coincidences — they are mathematical certainties. This is why Priya found three shared connections in one evening: it would have been extraordinary if she hadn't.Question 11
Why does the six degrees of separation phenomenon exist? Which is the best explanation?
A) Human psychology drives people to maintain exactly six connections between themselves and any acquaintance B) Social networks have a small-world structure where local clustering and a few long-range connections combine to create short average path lengths between any two nodes C) The number six is a mathematical constant determined by the average number of human relationships D) People deliberately seek out connections that bridge otherwise-separate communities
Show Answer
**Correct Answer: B) Social networks have a small-world structure where local clustering and a few long-range connections combine to create short average path lengths between any two nodes** **Explanation:** Watts and Strogatz's 1998 small-world model showed that networks with high local clustering (your friends tend to know each other) combined with even a small fraction of "long-range" connections (you know someone in a completely different social cluster) produce very short average path lengths between any two nodes. The six-degree result is not a psychological constant (Choice A) or a mathematical universal (Choice C) — it's an emergent property of network structure. The exponential growth of networks (100, 10,000, 1,000,000 at degrees 1, 2, 3...) means that even a handful of degrees spans the entire population, producing the small-world result.Question 12
Nadia's most-viewed video has 2 million views. She makes a follow-up video expecting similar results. Which probability concept from this chapter is most relevant to calibrating her expectations?
A) The birthday problem (she should look for overlap with other creators' audiences) B) The St. Petersburg Paradox (extreme outcomes have high variance; she should not expect the extreme outcome to repeat) C) The inspection paradox (her large-gap success means a small-gap is coming next) D) Six degrees of separation (her network will automatically amplify the next video)
Show Answer
**Correct Answer: B) The St. Petersburg Paradox (extreme outcomes have high variance; she should not expect the extreme outcome to repeat)** **Explanation:** The 2-million-view video is an extreme outcome in a high-variance distribution. The St. Petersburg Paradox illustrates that when outcomes have very high variance and fat tails, any individual extreme outcome (like a massive viral hit) carries very little information about the expected value of future outcomes. The expected value of Nadia's next video should be calculated from her overall distribution, not from her outlier video — just as you should not pay unlimited amounts to play the St. Petersburg game based on the possibility of an extreme outcome. Regression to the mean (Chapter 8) is also relevant here. The inspection paradox (Choice C) doesn't work quite this way — it affects wait times, not content performance.Question 13
Priya is deciding whether to email a well-known professional in her field who she doesn't know personally. She knows: 72% chance of no reply, 23% chance of a brief helpful reply, 5% chance of a meaningful connection. Which chapter principle most directly supports emailing anyway?
A) The inspection paradox — connected people get more emails, making reply more likely B) The birthday problem — the more people she emails, the more "collisions" (shared connections) she'll find C) The Monty Hall problem — she should switch her initial instinct to not email D) The St. Petersburg Paradox — even rare huge outcomes justify the action
Show Answer
**Correct Answer: B) The birthday problem — the more people she emails, the more "collisions" (shared connections) she'll find** **Explanation:** The birthday problem structure applies directly to the strategy of broad outreach: each individual outreach has a modest probability of success, but with many outreach attempts, the probability of at least one meaningful connection (a "birthday collision") becomes high. If Priya emails 20 professionals, the probability that at least one replies with value is much higher than the probability that any one specific person replies, just as 23 people are likely to have at least one birthday match even though each individual pair has only a 1/365 chance. Choice D (St. Petersburg) is partially relevant (asymmetric upside), but the birthday problem is the more direct and accurate framing for a strategy of broad networking. Choice A is a real effect (the inspection paradox/friendship paradox) but describes who she'll encounter, not why she should email.Question 14
The "boy-or-girl paradox" demonstrates which general principle about conditional probability?
A) Gender is never relevant in conditional probability calculations B) The mechanism by which you receive information — which child you learn about, or just that "one is" — changes the reference class and therefore the probability calculation C) Conditional probability only works when all outcomes are equally likely D) You need to know more than two facts to calculate conditional probability correctly
Show Answer
**Correct Answer: B) The mechanism by which you receive information — which child you learn about, or just that "one is" — changes the reference class and therefore the probability calculation** **Explanation:** "The older child is a boy" → P(both boys) = 1/2. "At least one child is a boy" → P(both boys) = 1/3. The information content feels similar but the reference class is different. In version 1, you've identified a specific child (the older one), leaving two possibilities. In version 2, you only know a property of the pair without identifying which child, leaving three possibilities. This is a general lesson for all conditional probability: you must always ask "what is my sample space given exactly this information obtained in exactly this way?" The mechanism of observation matters. This has real implications for how scientists design experiments and how detectives gather evidence.Question 15
Which of the following is the best summary of why probability surprises (birthday problem, Monty Hall, etc.) are relevant to the science of luck?
A) They prove that true luck is rare and most apparently lucky events have mathematical explanations, meaning luck does not really exist B) They show that probability intuition is wrong in specific, predictable ways — and calibrating these errors helps you see that things you call "lucky coincidences" are often mathematically expected, while things you dismiss as "too unlikely to pursue" may actually be positive EV C) They demonstrate that mathematics is always more reliable than intuition, so we should always calculate rather than act instinctively D) They explain why some people are luckier than others due to their network position and nothing else
Show Answer
**Correct Answer: B) They show that probability intuition is wrong in specific, predictable ways — and calibrating these errors helps you see that things you call "lucky coincidences" are often mathematically expected, while things you dismiss as "too unlikely to pursue" may actually be positive EV** **Explanation:** This is the meta-lesson of the chapter. The birthday problem teaches us that coincidences we experience as miraculous are often mathematically expected. Monty Hall teaches us that we resist updating on new information. The inspection paradox teaches us that our sampled experience of the world is systematically biased. The St. Petersburg paradox teaches us to evaluate extreme-outcome bets in utility rather than dollars. Together, these calibrations change what we see as "lucky" and what we see as "worth pursuing." Choice A goes too far — saying luck doesn't exist. Choice C overcorrects toward pure calculation at the expense of judgment. The correct stance (B) is calibrated intuition: use math to correct specific known biases while retaining the value of learned pattern recognition.Chapter 11 Quiz complete. The next section (Part 3) shifts from mathematics to psychology — how our minds shape the luck we experience.