Phlogiston/Risks of Forecasting and Metrics

Jump to navigation Jump to search

Things that can go wrong with forecasting[edit]

Some top risks[edit]

  • Forecasting can provide a motive to consciously cheat
  • Forecasting can trigger unconscious bias that ruins the validity of the metric
  • The cost of measurement is not zero.
  • Forecasting and measuringin general provides a bias toward doing things that are more easily measured (and against people doing things with less tangible benefits)

Additional Reading[edit]

  • Examples of Systemic Bias in Forecasts[edit]

    These articles include examples of charts where a number of forecasts are made over a series of time, such as years and decades, and every single forecast is completely wrong and visibly in conflict with the observable trends.

    The International Energy Agency consistently underestimates wind and solar power. Why?

    Obama’s Plans For Withdrawal From Afghanistan Have Changed Again

    Transpo Agencies Are Terrible at Predicting Traffic Levels

    Why crude oil prices keep taking us by surprise

    Itanium Sales Forecasts