金庸武侠的武功层级是怎么划出来的？
解释武侠读者用来把人物分成“一流”“绝顶”“宗师”的判据，并演示它们在十四部作品间的可迁移性。
16 min 8 scenes
Explore by topic
Mind & Society
Psychology, education, sociology, politics, and economic behaviour.
How Unique Is a Random Shuffle?
A random shuffle of 52 cards produces an order that has almost certainly never existed before, because the 52 possible arrangements so vastly outnumber all shuffles humans have ever performed that repetition is the rare event.
8 min 4 scenes
The Worst Day That Decides the Week
A contestant's worst day is not just a minor dip — in elimination formats it acts as a make-or-break threshold that can override an otherwise strong week.
18 min 9 scenes
Step Leaderboards and the Real Walking Champion
A step leaderboard ranks people by their cumulative weekly total, so the leader is whoever accumulates the most steps over the period, not whoever records the single highest daily walk.
8 min 4 scenes
Running in the Rain: Wet or Dry?
Running in the rain reduces wetness, but the benefit is smaller than intuition suggests because the time saved in the rain is roughly canceled by rain hitting the front of the body faster.
16 min 8 scenes
When Hours Pile Up: Sunk Costs and the Stories You Tell Yourself
Sunk time does not raise the value of a goal; it raises the cost of admitting the goal was wrong, so the brain reframes the goal as more valuable to keep self-narrative consistent.
14 min 7 scenes
排队两小时，东西为何更值钱？
人们会通过抬高物品价值来为已付出的时间正名，这就是'努力合理化'效应——非金钱成本同样会重塑感知价值。
8 min 4 scenes
When Order Matters: Non-Commuting Everyday Motions
Non-commutativity shows up whenever a motion changes your orientation or reference frame, so sequence-sensitive everyday tasks can be predicted by checking whether a step rotates you or shifts a fixed frame like a car's interior.
16 min 8 scenes
Why Do Concert Tickets Sell Out in Seconds?
Tickets disappear fast and resell at huge markups because real fan demand far outstrips the fixed number of seats, and resale markets surface that unmet demand at a clearing price.
16 min 8 scenes
Why a Riffle Shuffle Fails
A human riffle shuffle can look mixed yet retain structured order because its release pattern is not independent, individual cards do not have equal selection chances, and rigid spacing preserves positional relationships.
18 min 9 scenes
Where Do Pollsters Find 1,000 People?
The real-world sampling frames pollsters use to recruit about 1,000 respondents, and why weighting makes that sample represent millions.
14 min 7 scenes
True Randomness in Shuffles
A truly random shuffle gives every exact ordering of a deck the same chance—and 'looking unpredictable' is not the test.
14 min 7 scenes
三种餐厅，三种选址逻辑
看懂三类餐厅选址背后的顾客动机：顺路、专程与周末打卡，分别对应流量型、目的型和周期型选址。
14 min 7 scenes
Why Do Food Spots Open Side by Side? A Location Detective Investigation
Why food spots that seem like rivals open side by side—and the conditions that make that choice smart.
14 min 7 scenes
How Can 1,000 People Represent Millions?
How a random sample of about 1,000 people can estimate the opinions of millions, and why the same method fails when sampling is not random.
14 min 7 scenes
租金与时间的拔河：空地餐饮定位之谜
理解选址时租金预算与顾客时间成本如何互相权衡，并预测不同场景下什么餐饮形态能活下来。
14 min 7 scenes
Why Lying Subway Maps Win
Why a deliberately distorted map can be more useful for navigating a train network than a true-to-life one.
14 min 7 scenes
开店选址：新店该开在谁旁边？
弄清楚扎堆效应什么时候会带来客流、什么时候会分流顾客，并据此判断新店该靠近哪类邻居。
16 min 8 scenes
高峰流量下，绕行收益为何经常失灵？
探索高峰流量下绕行收益被变道和合流‘吃掉’的机制，并找到让绕行真正省时的临界条件。
16 min 8 scenes
Can 1,000 People Speak for a Country?
Why a random poll of about 1,000 people can reliably reflect a country of millions—and when it can't.
14 min 7 scenes
从奶茶街到夜市：餐饮聚集的现实形态
用决策时间和租金两个关键变量，看懂奶茶街、夜市乃至美食广场等餐饮聚集形态的成因。
14 min 7 scenes
餐饮为什么爱扎堆：同一条街上的邻居们
餐饮品牌扎堆不是偶然，而是一套可预测的聚集效应选址逻辑；看懂它就能解释从奶茶街到夜市的种种现象。
14 min 7 scenes
怎么判断一个「研究发现」靠不靠谱？
区分「靠谱」与「不靠谱」研究发现的三条可操作标准：研究设计是否严谨、结论是否被夸大、以及来源是否可追溯。
16 min 8 scenes
Designing AI-Allowed Assignments That Invite Curiosity
A reusable design pattern for AI-allowed assignments that separates automatable production from non-delegable inquiry, defended by a worked example and a boundary check.
16 min 8 scenes