Annotations (8)
“At first I was puzzled. How could things be fine at 60 startups and broken at 80? It was only a third more. Then I realized what had happened. We were using an O(n²) algorithm. So of course it blew up. The solution we adopted was the classic one in these situations. We sharded the batch into smaller groups of startups, each overseen by a dedicated group of partners.”
Operations & Execution · Leadership & Management · Technology & Engineering
DUR_ENDURING
Scaling breaks at complexity thresholds, not size
“If you correct course at a high enough frequency, you can be simultaneously decisive at a micro scale and tentative at a macro scale. The result is a somewhat winding path, but executed very rapidly, like the path a running back takes downfield. And in practice there's less backtracking than you might expect. Founders usually guess right about which direction to run in, especially if they have someone experienced like a YC partner to bounce their hypotheses off.”
Strategy & Decision Making · Operations & Execution · Leadership & Management
DUR_ENDURING
Micro decisiveness plus macro tentativeness
“Most startups have the same problems. No two have exactly the same problems, but it's surprising how much the problems remain the same, regardless of what they're making. Once you've advised 100 startups all doing different things, you rarely encounter problems you haven't seen before. That's one advantage of funding large numbers of early stage companies rather than smaller numbers of later-stage ones. You get a lot of data.”
Business & Entrepreneurship · Strategy & Decision Making · Operations & Execution
DUR_ENDURING
Volume creates pattern recognition expertise
“Focus is doubly important for early stage startups, because not only do they have a hundred different problems, they don't have anyone to work on them except the founders. If the founders focus on things that don't matter, there's no one focusing on the things that do. So the essence of what happens at YC is to figure out which problems matter most, then cook up ideas for solving them, ideally at a resolution of a week or less, and then try those ideas and measure how well they worked.”
Operations & Execution · Strategy & Decision Making · Leadership & Management
DUR_ENDURING
Priority, hypothesis, test, measure, weekly
“Founders will sometimes come in to talk about some problem, and we'll discover another much bigger one in the course of the conversation. Founders will come in to talk about the difficulties they're having raising money, and after digging into their situation, it turns out the reason is that the company is doing badly, and investors can tell. Often founders know what their problems are, but not their relative importance. They'll come in to talk about three problems they're worrying about.”
Strategy & Decision Making · Psychology & Behavior · Leadership & Management
DUR_ENDURING
People misrank problems by urgency vs impact
“That was another big surprise: how often founders don't listen to us. A couple weeks ago I talked to a partner who had been working for YC for a couple batches and was starting to see the pattern. 'They come back a year later,' she said, 'and say We wish we'd listened to you.' It took me a long time to figure out why founders don't listen. At first I thought it was mere stubbornness.”
Psychology & Behavior · Leadership & Management · Philosophy & Reasoning
DUR_ENDURING
Counterintuitive advice requires experiential proof
“The knowledge of the YC partners and the founders have different shapes: It wouldn't be worthwhile for a founder to acquire the encyclopedic knowledge of startup problems that a YC partner has, just as it wouldn't be worthwhile for a YC partner to acquire the depth of domain knowledge that a founder has. That's why it can still be valuable for an experienced founder to do YC, just as it can still be valuable for an experienced athlete to have a coach.”
Leadership & Management · Business & Entrepreneurship · Strategy & Decision Making
DUR_ENDURING
Breadth and depth are complementary, not substitutes
“If you look at history, great work clusters around certain places and institutions: Florence in the late 15th century, the University of Göttingen in the late 19th, The New Yorker under Ross, Bell Labs, Xerox PARC. However good you are, good colleagues make you better. Indeed, very ambitious people probably need colleagues more than anyone else, because they're so starved for them in everyday life.”
Culture & Society · History & Geopolitics · Leadership & Management
DUR_CONTEXTUAL
Greatness clusters geographically and institutionally
Frameworks (1)
Problem Hierarchy Diagnosis
Separating presenting problems from root causes
A systematic approach to diagnosing which problems truly matter when faced with multiple competing issues. Used when individuals or organizations present surface-level problems that mask deeper structural issues. The framework distinguishes between presented problems (what people say is wrong), underlying causes (what actually drives the issue), and hierarchy of impact (which problems will be fatal if unaddressed).
Components
- Listen to the presented problem
- Investigate the mechanism
- Identify the root cause
- Rank problems by lethality
Prerequisites
- Access to concrete business metrics
- Ability to ask direct, uncomfortable questions
- Understanding of what lethal vs. non-lethal problems look like in the domain
Success Indicators
- The person experiences an 'aha' moment recognizing the true problem
- Resource allocation shifts dramatically toward the diagnosed problem
- The presented problem resolves as a side effect of solving the root cause
Failure Modes
- Accepting the presented diagnosis without investigation
- Identifying a root cause but lacking authority to address it
- Correct diagnosis but insufficient will to face the implications
Mental Models (9)
Economies of Scope in Data Collection
EconomicsVolume of similar activities enables pattern recognition that individual instances cannot provide.
In Practice: After advising 100 startups, YC partners rarely encounter new problem types
Demonstrated by Leg-jdr-001
System Complexity Thresholds
Systems ThinkingSystems that work at one scale break at another not because of gradual degradati
In Practice: YC's organizational structure breaking at 80 startups despite working at 60
Demonstrated by Leg-hf-001
Root Cause vs Presenting Problem
Decision MakingPeople often present symptoms rather than causes.
In Practice: Founders coming in with fundraising problems
Demonstrated by Leg-jdr-001
Counterintuitive Truth Rejection
PsychologyWhen someone hears advice that contradicts their intuition, it sounds wrong, not insightful.
In Practice: Founders not listening to YC advice, returning a year later saying we wish we had listened
Demonstrated by Leg-jdr-001
Feedback Loop Frequency as Competitive Advantage
Systems ThinkingThe speed of your feedback loop determines how fast you can learn and adapt. If
In Practice: YC's focus on weekly iteration cycles: prioritize, hypothesize, test, measure
Demonstrated by Leg-hf-001
Micro-Decisiveness, Macro-Tentativeness
Decision MakingYou can be highly decisive at small time scales while remaining flexible at large ones.
In Practice: Explaining how rapid course correction enables simultaneous decisiveness and tentativeness
Demonstrated by Leg-jdr-001
Navigation Quality Compounds with Speed
TimeImproving navigational ability (knowing which direction to move) has a double ef
In Practice: Explaining how YC's value comes from improving founders' navigational confidence
Demonstrated by Leg-hf-001
Complementary Knowledge Shapes
Strategic ThinkingDifferent roles require different knowledge distributions: depth in one area vs. breadth across many
In Practice: Explaining why founders (domain specialists) benefit from YC partners (startup generalists)
Demonstrated by Leg-ac-001
Network Effects in Peer Quality
MathematicsThe value of being part of a high-quality peer group grows non-linearly with group size and quality.
In Practice: Discussing how great work clusters geographically and institutionally
Demonstrated by Leg-ac-001
Connective Tissue (4)
O(n²) algorithm complexity in computer science
In computer science, an O(n²) algorithm is one whose runtime grows quadratically with input size. When you double the input, runtime quadruples. YC discovered their organizational structure was O(n²): each of N partners needed to know each of M startups, creating N×M relationships. At 60 startups with 6 partners, that's 360 relationships. At 80 startups with 8 partners, that's 640 relationships, a 78% increase in complexity from a 33% increase in batch size. The solution, sharding the batch into smaller groups, reduced complexity to O(n) by having each partner know only their assigned startups. This is identical to how database sharding works: split one large database into several smaller ones to reduce query complexity. The organizational scaling problem maps directly to computational complexity theory.
Analyzing why YC's organizational structure broke at 80 startups when it worked at 60, Graham explicitly used algorithmic complexity as the analytical framework
Horror movie trope: focusing on boyfriend while door is ajar
In horror movies, the heroine often focuses emotional energy on a trivial interpersonal problem (boyfriend cheating) while remaining only mildly curious about an obvious lethal threat (mysterious door ajar, strange noise). The audience sees the misallocation of attention clearly and wants to shout at the screen. Graham uses this as a metaphor for how founders misrank problems: they obsess over moderate issues while dismissing or minimizing the problem that will kill the company. The horror movie structure works because the audience has information the protagonist lacks (genre awareness tells us the door matters more than the relationship). In business, the advisor has pattern recognition the founder lacks (experience tells us product-market fit matters more than fundraising tactics). The tension between what the protagonist prioritizes and what will actually matter creates dramatic irony in film and diagnostic opportunity in business.
Illustrating how founders misjudge relative importance of problems, Graham directly invokes the horror movie trope as an explanatory parallel
Football running back's downfield path
A football running back does not run in a straight line to the end zone. The path is winding, with frequent micro-adjustments based on where defenders appear. But the execution is rapid: the back is moving at full speed while making these adjustments. The result is a somewhat circuitous route executed at maximum velocity. The key insight: high-frequency correction enables aggressive speed. If the back tried to plot the perfect path before running, defenders would close in. If the back ran straight without adjustment, defenders would stop them. The winding path at high speed is optimal. Graham uses this to explain how startups can be simultaneously decisive (at weekly resolution) and tentative (at strategic resolution). Each weekly decision is made with conviction and speed; the macro strategy remains flexible. Like the running back, they correct course fast enough that wrong turns don't matter much, and the speed of execution creates advantage.
Explaining how rapid iteration enables both speed and flexibility through frequent course correction
Athlete and coach relationship
An experienced athlete still benefits from a coach despite having deep domain expertise. The athlete has experiential knowledge of their sport that the coach cannot match. The coach has breadth: they've worked with dozens of athletes, seen hundreds of competitions, and pattern-match across more situations than any single athlete experiences. The knowledge shapes are complementary: depth in one domain (athlete's sport) and breadth across situations (coach's pattern library). Neither can efficiently acquire the other's knowledge shape. The athlete would waste time trying to coach other athletes when they should be training; the coach would waste time trying to match the athlete's execution skills. Graham uses this to explain why even experienced founders benefit from YC: founders have depth in their domain, YC partners have breadth across startup problems. The relationship is valuable precisely because the knowledge is complementary, not overlapping.
Explaining why experienced founders still benefit from YC despite their own expertise
Glossary (1)
counterintuitive
VOCABULARYContrary to common sense or intuition; surprising because it opposes expected understanding
“And when you tell someone something counterintuitive, what it sounds to them is wrong.”
Key People (1)
Harold Ross
(1892–1951)Founder and editor of The New Yorker magazine from 1925 to 1951
Concepts (1)
O(n²) algorithm
CL_TECHNICALAlgorithmic notation meaning runtime grows as the square of input size; doubling input quadruples time
Synthesis
Synthesis
Migrated from Scholia