In god we trust, all others bring data
Yesterday I read a chapter called ‘Learning to drop your familiar tools’ in a book written by David Epstein ‘Range: How Generalists Triumph in a Specialized World‘.
‘For too long, we’ve believed in a single path to excellence. Start early, specialize soon, narrow your focus, aim for efficiency. But in this groundbreaking book, David Epstein shows that in most domains, the way to excel is something altogether different. Sample widely, gain a breadth of experiences, take detours, and experiment relentlessly. Range is an urgent and important book, an essential read for bosses, parents, coaches, and anyone who cares about improving performance’ – Daniel H. Pink, bestselling author of Drive and To Sell is Human
The chapter opens up talking about the Carter race study, something that I was unfamiliar with until yesterday. It goes like this:
“Carter Racing has to make a decision under pressure about whether or not to race given they have experienced engine failure at a rate of 29%. Based on the opinion of the engine mechanic, temperature is a key driver in the cause of engine failure and Carter Racing has never raced in such a cold temperature.
The data gathered by a more senior mechanic, who designed the turbo-charging system, does not demonstrate a correlation between temperature and gasket failures.
There are three possible scenarios to consider:
First, if they race and win, Carter Racing stands to gain a $1 million sponsorship from Goodstone and the team will feel an immense sense of pride. However, if luck allows them to win, Carter Racing also stands the chance of dismissing the engine failure issues in preparation for future races.
Second, if they race and experience engine failure, Carter Racing will face their biggest blow up on national television at a major event; not to mention the loss of substantial new sponsorships from Goodstone and their existing $500k oil contract.
Third, if Carter Racing chooses to withdraw, the team morale will be negatively impacted and their reputation with key sponsors may be in limbo. Withdrawal will also leave them $50k in the hole for the season.”
The team made their decision using their familiar tools, their trusted process, which was ultimately the wrong decision. By simply asking for more data, which obviously existed, the right decision was obvious.
Its easy to say in hindsight though, the chapter then moves on to a real life example, the 1986 NASA Challenger shuttle disaster.
It followed a similar pattern, the go, no go choice was made based on incomplete data, again, if they had requested all the data the issue was pretty obvious.
Process and protocol are good, but flexible thinking should be encouraged to challenge each and every instance, especially where there is any doubt.
The title of this post, apparently now hangs on the walls of the NASA space center.