If you want a prediction, go to the lady in the tent at the local fair.
If you want a forecast, talk to those who have an intimate knowledge of the drivers of the outcomes you are seeking to forecast.
Good forecasting is an iterative process, the more you do, the better you get, so long as you understand why the forecast is (almost) never right on each occasion it is done. Continuous improvement techniques are the core functions of good forecasting.
Forecasts are also improved when you leave aside some of the algorithms that manipulate the past into a forecast, and look instead at the drivers of demand, sometimes a qualitative input, to get a better picture of the sales that may come along. If you are selling ice-blocks, it is useful to look out the window to see how hot it may be, and factor that into forecasts, not just rely on sales over the last few weeks.
Agree, but as we both know, a project plan is as much a tool to manage the deviations that inevitably occur, as it is to co-ordinate and predict the flow of activities that need to occur to complete the project. Without a mechanism to provide transparency and enable managment of deviations, a project plan is as useless as a candle in a tornado.
I like the first sentence…
Forecasting when it comes Project Management is actually a science, using some highly developed techniques such as EVM. Prediction, on the other hand, is what usually make projects fail.