Before going out yesterday I checked the weather: sixty per cent chance of precipitation. I stood in the doorway for three seconds and put the umbrella down.
Halfway there, the rain came. Not heavy, but enough to soak my shoulders through. That evening I mentioned it to someone, and they said: the forecast was way off. I said, it told you sixty per cent. He paused, then said: who can read that?
I have had this conversation many times. It plays out again almost every time it rains.
The way people read a forecast is interesting: sixty per cent becomes “probably will not”, thirty per cent becomes “will not”, ninety per cent becomes “will”. The percentage gets squashed in the mind into a switch — on or off — and the whole wide middle is quietly deleted. Then, when the outcome does not suit us, we turn round and blame the forecast. It did not lie. We translated it wrong.
Why do we always mistranslate? Because people cannot stand suspension. A probability asks you to hold two answers at once: it may rain, it may not, so I take an umbrella but I do not cancel the trip. Most people, at the moment of walking out the door, will only hold one. Holding two means admitting you might be wrong — one of the things people least want to admit, especially about something small.
One layer further down, complaining that the forecast was “wrong” is a very cheap disclaimer. If it had said “rain, guaranteed” and you carried an umbrella for nothing, you would feel cheated; if you left the umbrella and got wet, at least there is someone to blame. Complaining rewrites “I judged badly” as “it deceived me”. The cost shifts away at once, and you come out clean.
This kind of mistranslation does not stay in the weather. A doctor says the operation has a ninety per cent success rate; the family hears “no problem”. A fund manager says six per cent expected annualised return; the client hears “six per cent a year”. A friend says “I will try to come”; the host hears “he is coming”. Any word carrying uncertainty gets sanded flat into certainty as it passes from mouth to ear. It is not that the listener is stupid. Certainty is comforting, and comfort is more addictive than accuracy.
I do not want to make this a moral matter, and I do not think anyone should be scolded for it. Aversion to probability is temperament, not something cultivation fixes. But one thing can be fixed: if you know you translate this way, keep something in reserve. Carrying an umbrella is cheap. Spending ten seconds on the small print that says expected returns are not a promise is cheap. Silently converting a friend’s “I will try” into “probably not coming” and leaving a space at the table is cheap. These moves are tiny, almost superfluous — and what they buy is precisely the “what if”.
I have paid for this myself. One year I put together a project schedule, reporting every stage at its most optimistic estimate, each one linked neatly to the next; it looked handsome. What happened next is not hard to guess: the third stage stuck, everything downstream collapsed, delivery slipped a month. Going back over it, no single stage was a surprise; every delay was within the range of ordinary sense. I had simply added them all up at the good end. After that I set myself a clumsy rule: anything said with “probably”, “should be”, “roughly” gets planned at the worst end and hoped for at the best. The first half is cost, the second half is mood.
As for the forecast itself, it has in fact been getting more accurate for years — far more reliable than when I was a child. Most of the scolding it takes is not about accuracy; it is that we demand a sentence it cannot honestly say. Ask it for a guarantee and it can only be blamed when it errs. Ask it for a probability and it hands the choice back to you: umbrella or no umbrella is your business.
I got half soaked that day, went home and changed, and did not feel wronged. Sixty per cent chance of rain, and I bet on the forty. I lost, that is all. Losing a bet and paying up is more decent than turning round to blame the forecast.
Written by Bu Shi
说些什么吧！