失败一直是市场经济最基本的组成部分。市场有效，是因为人 们在不断尝试新想法。这些想法大多都会失败。那些成功的想法又会让早先的想法失败。在美国，每年约有10%的企业消失。这幅景象有点让人难堪——但试错法 (trial and error)可能正开始获得作为一种商业技巧的公正地位，而不再是资本主义肮脏的小秘密。

有些迹象让人心怀希望。哈 佛商学院(HBS)的斯特凡•托姆克(Stefan Thomke)认为，计算机运算技术的进步使得新产品实验可能成为家常便饭。例如，现在很容易实验网站布局改变的效果，向不同访客显示不同的配置，并实时 跟踪结果。与此同时，谷歌(Google)定期发布新产品的试用版(beta)。而像《魔鬼经济学》(Freakonomics)作者之一史蒂文•莱维特 (Steven Levitt)这样的学术超级明星一直在教授商业实验方面的管理课程。

关于从错误中学习的心理学我们也开始了解的越来越 多。《轻推》(Nudge)作者之一、行为经济学家理查德•塞勒(Richard Thaler)发明了“享乐主义的加工”(hedonic editing)这个短语，用来形容我们将小损失与较大的收益放在一起以掩饰损失之痛苦的习惯行为。粉饰太平乃人之本性，但也可能导致你无法从失败中吸取 教训。塞勒和同事甚至研究了游戏节目《一掷千金》(Deal Or No Deal)中参赛选手的行为。他发现，选择失误的人接下来会开始不顾后果地冒险，这往往会让他们错上加错。

如果失败暂时夺去了我们的判断 力，我们就很难从失败中吸取教训。当我们开始理解为什么试错法是一个非常痛苦的过程时，或许就能够更加有建设性地使用这种方法。这场金融危机让我们意识 到，无法容忍一点失败的体系是危险的。一家机构“太大以致不能倒闭”的理念过去听起来很让人信服。现在不再如此。

Gain from the pain of failure

By Tim Harford 2010-03-11

Undercover Economist

Failure has always been a fundamental part of a market economy. When markets work, they do so because new ideas are constantly being tried out. Most fail. Those that succeed cause older ideas to fail instead. In the US, about 10 per cent of businesses disappears each year. This is an awkward insight – but trial and error could be starting to take its rightful place as a business technique, rather than the dirty little secret of capitalism.

There are some hopeful signs. Stefan Thomke of Harvard Business School has argued that advances in computation have made it possible to experiment on new products as a matter of course, trying many things and expecting many failures. It is now easy, for example, to experiment with changes in the layout of a website, showing different configurations to different visitors and tracking results in real time. Google, meanwhile, routinely launches new products with a “beta” label on them. And academic superstars such as Steven Levitt, co-author of Freakonomics, have been teaching executive courses in business experimentation.

We are also starting to learn more about the psychology of learning from mistakes. Richard Thaler, the behavioural economist behind Nudge, coined the phrase “hedonic editing” to describe our habit of lumping small losses together with larger gains in order to mask the pain of the loss. Sugar-coating is human, but it's also a recipe for failing to learn from failure. Thaler, with colleagues, even studied the behaviour of contestants on Deal or No Deal. He discovered that people who had made unlucky choices then started to take reckless risks, which often compounded the error.

It's hard to learn from failure if it briefly robs us of our judgment. As we start to understand why trial and error is so painful a process, we may be able to use it more constructively. The financial crisis has made us aware that a system that cannot tolerate a bit of failure is a dangerous one. The idea that an institution was “too big to fail” used to sound reassuring. Not any more.