Error and Correction
Admitting error is the hardest thing, but failing to correct after discovering error is the greatest mistake
📌 Concept Explanation
Error and correction = Making errors is inevitable, but persisting after discovering the error is the real mistake; speed of admitting error determines magnitude of loss.
You’re playing golf and hit the ball into water. That stroke is already lost. You have two choices: regret this stroke and affect the next one; or accept it and focus on how to play the next shot. In Duan’s investment philosophy, “error” itself isn’t the problem—“not admitting error” is fatal. The difference between good golfers and bad ones isn’t who never makes mistakes, but whose error rate is lower and who recovers faster from mistakes.
💡 Core Understanding
1. Making errors is inevitable—the difference between masters and ordinary people is low error rate, not never making errors.
This is investment philosophy Duan distilled from golf. Buffett has made many wrong stock picks, but his overall error rate is extremely low—“In stock investing, we expect each investment to succeed. Over 38 years of running Berkshire, the ratio of profitable investments to losing investments was about 100:1.” This is a true moat.
2. Failing to correct after discovering error is the biggest mistake—“sunk cost” is the most dangerous psychological trap.
This is Duan’s description of the most common error pattern—because so much cost has already been invested, unwilling to admit error and stop losses. The core of doing the right thing is: when you discover an error, correct immediately—“no matter the cost, it’s still the minimum cost.”
3. A good corporate culture is an error-correction system that enables early detection and rapid correction.
One important dimension in Duan’s evaluation of companies is their error-correction ability—“Companies like Google appear much stronger than Panasonic in error correction capability—they adjust much faster.” Companies with weak error-correction ability, once they make a directional mistake, often persist in error to the end.
4. In investing, the concrete manifestation of admitting error is: discovering you bought wrong, sell immediately regardless of loss amount.
An important trigger condition for sell logic is: discovering your original buy judgment was wrong (not because the price fell, but because the company itself has problems). Continuing to hold at that point isn’t long-termism—it’s using “long-term” as cover for unwillingness to admit error.
🛠 How to Practice
Distinguish two types of errors:
- Execution-level errors (errors during “doing things right”): These are part of the learning process—tolerable, key is learning from them and reducing probability. Like occasionally hitting into water in golf—it’s normal.
- Directional-level errors (errors in “doing the right thing”): These must be corrected immediately with zero tolerance. Like realizing you’ve taken the wrong road—you must turn back immediately; don’t continue just because “you’ve already come so far.”
Error-correction checklist for investing:
When you discover any of these situations, reassess whether you bought wrong:
- Company’s corporate culture shows fundamental problems (not short-term performance fluctuations)
- Company’s business model has fundamentally changed
- Original buy logic no longer holds
Note: Price decline itself isn’t a correction signal—fundamental company changes are.
📖 Case Study
Buffett’s Energy Future Holdings Investment (2012): Admitting error without delay
Buffett invested approximately $2 billion in utility Energy Future Holdings, later facing risk of total loss. Duan commented: “That means about 1% of market value might be gone…relative to what Old Ba can earn from investing these past few years, this really doesn’t amount to much. It’s somewhat like hitting a bad shot during golf, but overall score remains excellent. As long as you play the game, errors are inevitable—winners are those with low error rates.” This demonstrates positively that “errors are inevitable but what matters is final score.”
Source: Investment Logic Chapter, 2012-02-28
Tesla (before 2019): Made money, but afterward discovered logic was wrong
When asked whether he’d made trades where he made money but later realized facts or logic were wrong, Tesla was Duan’s first example: “My Tesla case is typical.” He made good money on Tesla but eventually sold everything because he “couldn’t understand it,” admitting the original buy logic was flawed. This case illustrates: good outcomes don’t mean correct decisions—validating decisions by results is dangerous.
Source: Business Logic Chapter, 2018-10
⚠️ Common Misconceptions
-
❌ “Lost so much already, let’s wait and see” — This is the sunk cost trap. “Many people make the stupid error of continuing to wait—I’ve already invested tens of millions in this, how can I stop now?” Amounts already lost are history; they cannot become reasons for continuing to hold.
-
❌ “Good companies never make errors” — “When good companies make errors, those are often buying opportunities.” Good companies also make errors; the difference is they have low probability of making major principled errors, plus strong error correction capability. Equating “good company” with “error-free company” causes missing many buying opportunities.
-
❌ “Admitting error means weakness” — Duan repeatedly emphasized that admitting error is the hardest thing but also most important. “The Japanese refusing to admit errors is actually somewhat similar to us.” Not admitting error isn’t strength—it’s the root cause of turning small errors into big ones.
💬 Original Quotes
“The difference between masters and other players lies in low error rate, not how many good shots they can hit.” (Source: Investment Logic Chapter, 2011-09-07)
“As long as you play the game, errors are inevitable—winners are those with low error rates.” (Source: Investment Logic Chapter, 2012-02-28)
“We’ve avoided many errors; fewer mistakes mean greater success probability. My approach to investing is the same. Like golf—you pursue overall score and enjoyment, not hole-in-one every shot.” (Source: Business Logic Chapter)
“Good corporate culture isn’t omnipotent, but it helps companies avoid major principled errors and enables early detection of errors.” (Source: Business Logic Chapter, 2011-02-18)
“There’s no perfect golf, no perfect investment—and life is the same.” (Source: Investment Logic Chapter, 2012-12-20)
🔗 Related Nodes
Upstream Concepts (prerequisites for understanding this concept): Doing the Right Thing · Stop Doing List · Corporate Culture
Downstream Concepts (conclusions derived from this): Sell Logic · Ordinary Mindset · Long-termism
Related Company Cases Alibaba (Wei Zhe incident) · GE (admitting errors post-financial crisis) · Tesla (made money but logic was wrong)
Related Figures Duan Yongping · Buffett
错误与纠错
认错是最难的事，但发现错了不改才是最大的错误
📌 概念解析
错误与纠错 = 犯错不可避免，但发现错了之后继续坚持才是真正的错误；认错的速度决定了损失的大小。
你打高尔夫，把球打进了水里。这一杆已经没了，你有两个选择：懊恼这一杆，影响下一杆；或者接受它，专注于下一杆怎么打。段永平的投资哲学里，“错误”本身不是问题，“不认错”才是致命的。好的球手和差的球手，差别不在于谁从不犯错，而在于谁的错误率低、谁能更快从错误中恢复。
💡 核心理解
1. 犯错不可避免，高手和普通人的差别在于错误率低，而不是从不犯错。
这是段永平从 golf 中提炼出的投资哲学。巴菲特买错过很多股票，但他的整体错误率极低——“在投资股票时，我们预期每一笔投资都会成功，在我们经营伯克希尔的38年中，投资获利的个案和投资亏损的个案比例约为100:1。“这才是真正的护城河。
2. 发现错了不改，才是最大的错误——“沉没成本”是最危险的心理陷阱。
这是段永平描述的最常见的错误模式——因为已经投入了大量成本，就不愿意承认错误、停止损失。做对的事的核心就是：发现错了就立刻改，“不管多大的代价都是最小的代价”。
3. 好的企业文化是一个纠错系统，能让公司及早发现错误、快速改正。
段永平评价一家公司，很重要的一个维度就是它的纠错能力——“看起来 Google 这类公司的纠错能力要比松下这类公司强很多，调整起来快很多。“纠错能力弱的公司，一旦犯了方向性错误，往往会一错到底。
4. 在投资中，认错的具体表现是：发现买错了，立刻卖出，不管亏了多少。
卖出逻辑的一个重要触发条件就是：发现当初的买入判断是错的（不是因为股价跌，而是公司本身有问题）。这时候继续持有，不是长期主义，而是用”长期”掩盖不愿认错的心理。
🛠 如何实践
区分两种错误：
- 执行层面的错误（“把事情做对”过程中的错误）：这类错误是学习过程的一部分，可以容忍，关键是从中学习、降低概率。就像打高尔夫偶尔把球打进水里，是正常的。
- 方向层面的错误（“做对的事”的错误）：这类错误必须立刻纠正，零容忍。就像发现走错路了，必须立刻掉头，不能因为”已经走了这么远”就继续走。
投资中的纠错清单：
发现以下情况时，应该重新评估是否买错了：
注意：股价下跌本身不是纠错的信号，公司基本面变化才是。
📖 案例拆解
巴菲特投资 Energy Future Holdings（2012年）：认错不拖延
巴菲特投资约 20 亿美元的电力公司 Energy Future Holdings，后来面临血本无归的风险。段永平点评：“就是说相当于 1% 的市值可能没有了……相对于这几年老巴投资能赚的钱而言，这个实在是不算什么。有点像一场 golf 里打了个坏球，但总成绩还是很好。只要下场打球，错误是不可避免的，赢家都是那些错误率低的人。“这是对”错误不可避免，但总成绩才是关键”的正面示范。
来源：投资逻辑篇，2012-02-28
特斯拉（2019年前）：赚了钱，但事后发现逻辑搞错了
段永平在被问到”你有没有做过这样的交易：赚了钱，事后却发现事实或逻辑搞错了”时，第一个举的例子就是特斯拉：“我的特斯拉就是典型哈。“他在特斯拉上赚了不少钱，但最终因为”看不懂”而全部放弃，并承认当初买入时逻辑就是错的。这个案例说明：结果好不等于决策对，用结果来验证决策是危险的。
来源：商业逻辑篇，2018-10
⚠️ 常见误区
-
❌ “亏了这么多了，再等等看” — 这是沉没成本陷阱。“多人都犯继续等下去的愚蠢的错误，这个事情我已经投入了几千万呀，怎么停的下来？“已经亏损的金额是历史，不能成为继续持有的理由。
-
❌ “好公司不会犯错” — “好公司犯错误时往往是买入的机会。“好公司也会犯错，区别在于它们犯原则性大错的概率低，且纠错能力强。把”好公司”等同于”不犯错的公司”，会错过很多买入机会。
-
❌ “认错就是软弱” — 段永平多次强调，认错是最难的事，但也是最重要的事。“日本人死不认错这点倒是和我们有点像。“不认错不是强大，而是把小错误变成大错误的根源。
💬 原文金句
“高手和其他选手的差别就在错误率低而不是能打出多少好球来。“（来源：投资逻辑篇，2011-09-07）
“只要下场打球，错误是不可避免的，赢家都是那些错误率低的人。“（来源：投资逻辑篇，2012-02-28）
“我们少犯了许多错误，失误率少成功的机会就大，我做投资也是这样。如打高尔夫，你追求的是整体成绩，是乐趣，而不是一竿进洞。“（来源：商业逻辑篇）
“好的企业文化不是万能的，但能让公司少犯原则性错误，而且可以让公司及早发现错误。“（来源：商业逻辑篇，2011-02-18）
“没有完美的 golf，也没有完美的投资，人生也一样。“（来源：投资逻辑篇，2012-12-20）
🔗 关联节点
上游概念（理解这个概念的前提）： 做对的事 · Stop Doing List · 企业文化
下游概念（由此推导出的结论）： 卖出逻辑 · 平常心 · 长期主义