forecasting
Update in small steps
The rule
Change your mind slowly as new facts come in. Don't cling to your first view, and don't flip on every little noise.
Where it flips
But changing too eagerly turns every headline into a trade. The fix is to ask whether the new fact really changes your reason to hold, or is just noise.
Good forecasters shift their odds a little at a time as facts land, instead of holding stubbornly and then panicking. A weak quarter nudges the odds down a bit. It rarely means sell everything, and it rarely means nothing.
A worked example
A composite holding reports a soft quarter. The steady move is to trim your '70% this idea works' to about 60%, not to dump the stock or to ignore the miss. [illustrative]
How to spot it
- ·belief unchanged even after real new evidence
- ·or a full U-turn on a single data point
- ·reacting to the price move, not the facts
Philip Tetlock · Superforecasting