- By rewarding or promoting these individuals, the business risks having lucky managers rather than competent ones - fine, until their luck runs out.
- Also, although spontaneous decisions may turn out to bring some business benefit, they don't teach us anything. We can't use them to improve the way we take decisions, or to instruct others.
Wednesday, 25 November 2009
When things go wrong: Towards better decision making - quality of decision and the role of chance
Business results are the outcome of the interaction between our decisions, our actions and chance. Even if we make no error, there is always the cahnce that a bad outcome will result from a 'good' decision. For example, we might play dice game version A (http://spreadsheets.google.com/pub?key=te9MzyHoIN6EyuoHmfDxMaw&output=html) ten times and lose every time, despite having established that hte risk had a positive expected value. But how would such a decision be regarded in business?
If we were rewarded solely on results, with no attention paid to the way we took our decisions, our $10 loss would look pretty bad. Our performance report might read as follows: 'Despite your poor results, you played this game again and again, throwing good money after bad on the off-chance of things somehow coming right. You recklessly gambled company money on an uncertain future. Your poor results are evidence of your bad judgement. What were you thinking?'
But if we were rewarded on the quality of our decision-making process, our actions would appear in a very different light, resulting in a different review: 'Although results have been poor, due to circumstances beyond your control, the quality of your decision making was excellent. You obtained all the information that you could on possible outcomes, and the probabilities of each, and took a decision on that basis. The negative results, though disappointing, have not bankrupted the company. You will be rewarded on the basis of decision-making quality.'
The flip side of this is that people might make decisions on impulse, or randomly, and still get good results by chance.