Why does Elon Musk always win: The 20-person system and the underlying logic

Not everything you see from Musk is the result of his individual effort. Investor Shaun Maguire—who led the controversial early investment round in SpaceX in 2019—has revealed a little-known truth: Musk does not operate alone but is part of a sophisticated coordinated system of 20 core members.

When Tesla’s market cap was only 4-5 billion USD, when Starlink was still unproven, Shaun—a PhD in mathematical physics, former DARPA employee, and professional eSports player—saw through what others overlooked. He not only helped SeQU IA CAPITAL decide on an initial $20 million investment (then increased it to $600 million), but also revealed a methodology for evaluating talent, organizational governance, and capital allocation at the highest level.

The real secret: 20 people, 10 years, one willpower

When people talk about Musk, they often emphasize the numbers: 112 hours of work per week, running SpaceX and Tesla simultaneously, or continuous technological breakthroughs. But that’s only half the picture.

Behind the scenes is a team. These about 20 individuals are not ordinary members. They have worked with Musk for over 10 years, building trust to the point that— in some cases—they can act according to his thinking without needing permission. They know when to report decisions, when they can handle issues independently. They understand the precise boundary between “being entrusted” and “being authorized to decide.”

“These 20 people can directly realize his will, executing with strength, scale, and precision. This is an operational system that other Silicon Valley entrepreneurs completely lack,” Shaun explains.

This harmony cannot be learned quickly nor recruited from the market. It requires at least 10 years of continuous challenge, pushing each other’s limits, building a shared language, calibrating with common criteria.

The screening mechanism: “Give the rope to hang yourself”

Within this system, there is an extremely strict selection process. Musk does not believe in long-term training programs. Instead, he applies a dual-strategy:

• Rapid promotion if you perform: Your advancement rate will far surpass nearly any other organization in the world
• Almost immediate elimination if you fail: Mess up once, and you are basically removed from the core team

“Continuously letting people advance at a pace far beyond other organizations builds extremely strong loyalty among the best. You give them what others cannot—an opportunity to truly prove their capabilities,” Shaun analyzes.

Combined with generous rewards, this mechanism ultimately retains the top 1% of excellence in the organization—they are the ones truly operating everything. The rest? They will seek opportunities elsewhere.

15 tiers of talent: Why you cannot see the difference

There is a strange phenomenon in the recruitment world: The most talented individuals can see the difference between candidates that ordinary people consider “equivalent.” But the opposite—those three levels below—are completely unable to distinguish.

Shaun offers a highly nuanced framework: In fields of (mathematics, theoretical physics, computer science), there are about 15 clearly distinguishable levels. From “easily earning a PhD at top universities” to “a mathematician of the century, like a Fields Medalist” (such as the recipient of the Fields Medal).

But this is a non-transparent, two-way system.

• High-level individuals can, in 30 minutes of conversation, accurately assess the other’s level (with an error margin of no more than 1-2 levels)
• But those three levels below or more generally cannot distinguish higher levels

For example: A high school math teacher will consider all students with a perfect SAT math score of 800 (absolute perfect score) as “equal”—but among these, some will become Fields Medal winners, while others are just top university graduates.

###Compared to Elo ratings: Why you don’t recognize talent

Shaun uses the Elo rating system in chess to illustrate:

• 2850 points (world champion) vs 2700 points (average grandmaster): Overwhelming dominance, over 99% win rate
• 2700 points vs 2500 points (newly titled grandmaster): Clear advantage, 85-90% win rate
• But to a 1000-point player watching these games: They cannot distinguish the skill difference at all

Conversely, a player rated 2800 can evaluate the level of an opponent’s 10 moves accurately. “This is a core skill for investors: My main job is to evaluate people and talent, especially at the earliest stages. Understanding the level difference of talent is an absolute superpower.”

How Musk perceives engineer thinking from an interview

Shaun shares a detail: Musk interviewed a college student studying economics and immediately concluded, “You won’t develop a business; you should do mechanical engineering,” because he saw the person had an engineer’s mindset.

This person later became a very senior engineer. This ability cannot be faked—it is a true superpower, stemming from a long calibration process.

Shaun’s calibration ability comes from direct contact with extreme talent:

• In 2010, participated in Clay Math Institute’s summer program in Brazil, sharing a residence with 100 top mathematicians, including 2 Fields Medalists
• Interned at DRW with colleagues who are three-time Putnam Competition winners and three-time IMO gold medalists
• Knows many Nobel laureates before they win, such as Kip Thorne

“You witness firsthand how 0.001% of the most outstanding people operate, and your evaluation standards will change forever.”

Investment methodology: Define the required capabilities first

Shaun’s investment framework is very clear and applicable:

Step one: What qualities does this company need to succeed? Some companies don’t require high intelligence (e.g., traditional trash collection), but “robotic trash collection” is extremely important.

Step two: What level is the founder at regarding these key qualities?

It could be sales ability, fundamental mathematical skills (for AI research companies), or simply the ability to withstand pressure. “The key is to identify which qualities are truly critical, then evaluate the person on that dimension.”

###Case: Reading a cold email and instantly knowing technical capability

Shaun invested in Factory, where founder Matan Grinberg impressed just with a cold email. Grinberg mentioned that during college, he co-authored a paper with Juan Maldacena—one of the most renowned theoretical physicists in string theory.

“For me, just this information alone meant—this person has at least a 2600 level of technical skill (compared to chess rating).”

Key insight:

• If collaborating during PhD with Juan, the threshold is around 2300
• But if you can publish with Juan during undergrad, that’s 2600+

“He just said ‘published with Juan,’ not ‘Juan Maldacena’—I really like this high-level way of communicating.” Moreover, this founder combines extremely strong technical skills with excellent sales and empathy—this is the real magic combination.

Most venture investors would completely overlook this signal. Why? Because they lack the calibration ability in this field. This also explains why truly good projects are often missed by outsiders—not because the founders are not good enough, but because evaluators simply don’t understand.

SpaceX 2019: The convincing battle of a prophet

In 2019, when investing in SpaceX with Tesla’s valuation only 4-5 billion USD, Shaun faced the greatest pressure. One partner even scored only 1/10. The discussion was “the most controversial and fierce dialogue” he had ever experienced. But he refused to accept the “no” answer.

The persuasion strategy applied: Start with a $20 million trial investment (at that time, they wanted $600 million). Then, over the next six months, he sent progress updates every three weeks to all decision-makers.

“This approach has two effects: First, it shows your persistence—this is not a spur-of-the-moment idea; second, it shows the data flow—seeing the speed and acceleration of progress. You can often change others’ minds this way.”

This methodology applies to all situations: It’s very hard to change someone’s mind with just one data point, but over time, continuous data streams can recalibrate perceptions.

Today, investing in SpaceX seems obvious, but in 2019, Starlink was still unproven, reusable rockets just started operating, Tesla was not yet a trillion-dollar company. The entire aerospace industry was not understood by mainstream investors. Investing correctly is not about luck but about seeing the level differences others miss.

Why almost everyone underestimates Musk’s companies

The Boring Company: Overlooking technical difficulty

A shocking detail: When Shaun asked Steve Davis (the first SpaceX employee, now head of The Boring Company), about the technical difficulty of the “Zero People In Tunnel, Continuous Mining” boring machine, the answer was: Slightly harder than Falcon 9, easier than reusable Falcon 9.

This device has higher technical difficulty than Falcon 9—but outsiders completely don’t understand. Why? Everyone compares linearly.

They compare one drill to another, but fail to see the generational difference between Falcon 1, Falcon 9, reusable Falcon 9, and Starship.

“Precedents and Limits”: Limits of perception

Musk has a profound observation: People only react to “precedents and limits.”

• Precedents: You have crossed the nonlinear stage, truly achieved the goal
• Limits: Rocket explosion or landing—visual differences that ordinary people can understand

But people do not understand:

• The difference in payload mass
• The difference in achievable orbits
• The complexity behind the technical challenges

“Therefore, The Boring Company now is like SpaceX before 2009—before reaching the milestone ‘Zero People In Tunnel, Continuous Mining,’ outsiders cannot perceive progress. But once achieved, perception will jump tenfold.”

Optimus Robot: Creating “extreme moments”

At Tesla’s event, Shaun witnessed 20 Optimus robots walk out. “They moved from about 9-12 meters, at first I couldn’t tell if they were real actors or robots. I looked at their faces—like looking at humans. Then I looked down at their bodies, and when I saw the hips—very narrow, not human-like—I confirmed they were real robots.”

This experience creates a nonlinear psychological shock, making people truly feel that “the future is arriving.” That’s Musk’s strength: Creating milestones, shocking moments, that visually help everyone understand what’s coming—something charts and data can never do.

Capital allocation: The art of scale betting

Shaun believes Musk’s capital allocation ability is seriously underestimated. The key point is not “what to invest in,” but “when and how much.”

For example, with Starlink: SpaceX started research around 2013, but initial investments were small—they waited until Falcon 9 reusability succeeded. After achieving that in 2016, they spent another two years resolving technical details. By around 2018, Musk confirmed that the components were ready:

• Phased array antenna technology sufficiently mature
• Cost-effective rocket launches (because of reusable rockets)
• The number of launches per year could increase significantly (not building new rockets each time)

“At this point, the bet shifts from small to medium scale. When the unit economics model of the bet is confirmed, then it scales up again.”

Shaun compares this to hedge funds: “If you have a vague idea about a company but don’t understand it well, you buy a 1% stake—this helps you start truly learning about that company. From the outside, it’s hard to go deep, but once you have a position, you develop feelings, more concern. If your thesis is confirmed, you increase from 1% to 5% or 10%.”

From battlefield to investment: The leap in perception

During his PhD, Shaun was recruited by DARPA to serve in Afghanistan. This was completely outside his plan, but he says: “I wanted to serve my country; this was the biggest leap into the fire.”

“In a war zone, the speed of learning is unimaginable—because the feedback and danger are so high.”

In March or April 2012, Shaun drove from one base to another to report. That day, there were no vehicles on the road—normally it takes 45 minutes to an hour, but that day he arrived in just 15 minutes. “I immediately felt something was wrong.”

At the base, he asked about intelligence. People sensed an unusual atmosphere but had no clear information. About 2-3 hours later, a large-scale coordinated attack occurred—6 different bases across the country were attacked simultaneously.

When intuition signals a problem that data cannot show—meaning the data system needs improvement. This experience made Shaun realize: You cannot rely solely on data. Sometimes, a person’s gut feeling about the environment is more accurate than any statistics.

The operators working with Musk: Silent criteria

Figures like Gwynne Shotwell, Steve Davis, and Antonio Gracias—who maintain long-term relationships with Musk—all share certain traits:

First, they are truly willing to work. Antonio Gracias slept at Tesla factories to help increase production—“this level of dedication is hard to fully express respect.”

Second, absolute discretion. “It sounds simple, but when you really want to change the future, sometimes surprises are needed. Most investors will reveal secrets. Keeping secrets is a very low threshold, but most cannot do it.”

Third, support 24/7. Always present in good times and bad—Antonio does better than almost all other investors in this regard.

Fourth, understand ‘AlphaGo decision-making.’ Shaun compares Musk to AlphaGo: “AlphaGo makes moves that Go players can’t understand, but after 17 moves, everyone realizes ‘Wow, that move is crazy.’ Working with Musk is the same—he will do things that are extremely counterintuitive, at first you don’t understand, but after 6 months or a year, you do, and look back as if he were a prophet.”

Those who don’t understand this will reveal their plans, doubt decisions. Those who do will trust the system, trust the direction, be loyal to teammates, and focus on the mission.

Lessons from failure: Selling Nvidia at a $600 billion valuation

Shaun bought Nvidia at IPO in 1999 (when he was 13, because he needed a graphics card for gaming), holding until its market cap reached $600 billion. Back then, data center and gaming revenues were roughly equal, and he felt the valuation was too high, so he sold.

Where was the mistake?

First, underestimating Jensen Huang. “Even holding long-term, I didn’t realize how outstanding Jensen Huang is. He is truly very talented, leveraging his advantages to an incredible extent.”

Second, underestimating how market irrationality can become a self-fulfilling prophecy. Shaun thought the market was somewhat irrational then, but Jensen Huang exploited this so well that Nvidia became extremely valuable, while AMD and Intel became very cheap, forcing them to cut investments. As a result, Nvidia can now ramp up investments wildly, and AMD and Intel’s plans to catch up are essentially crushed.

Third, being discouraged by ‘money flowing into hardware without understanding it.’ Shaun was very annoyed: “I studied Broadcom, TSMC, ASML from a young age, understood semiconductors. But two years ago, money poured in without even knowing what Mellanox was, and they said Nvidia would go to $3 trillion. I thought: Forget it, I’ll withdraw.”

An important addition: Shaun believes Nvidia’s acquisition of Mellanox in 2019 (around $8 billion) “may be one of the best acquisitions in history,” giving Nvidia a huge technological advantage in connectivity—this was a seismic advantage in the data center era.

Lesson: “If I had understood how outstanding Jensen Huang was back then, I would have avoided this mistake. But many funds poured into Nvidia without understanding semiconductors—sometimes, lack of knowledge helps you avoid being bound by ‘valuation.’ Deep expertise can sometimes make you overly cautious.”