RR6
Christopher McSweeny
Chaos Theory was a fascinating point in this chapter because it stood out in a different light
than what Malcolm had explained form Jurassic Park all those years ago. As silver explains it, it has
two key distinctive features, it is dynamic or ‘ever-changing’, and the systems are non-linear. This
can be seen in numerous systems within the world, but weather forecasting is the best example of
this, because the atmosphere is always changing and moving so a continuous update on a forecasting
model is necessary to keep up with its dynamic change. If a forecaster did not update the model, it
would rapidly change very quickly because one wrong forecast would turn into an even worse
forecast. This also is important to note with data inaccuracy as well. As silver points out with
Loren’s team forecasting program, slight errors in data can create a big problem and this only grows
over time with forecasting if it is wrong in the first place, it grows at a more rapid rate as you
continue to try and model with that data. I was thinking to myself, what we are learning in class that
pertains to this topic and I think its perfectly relevant to forecasting error. The importance of
calibrating your forecasting error is ever more important because if you don’t it is more likely that
your errors may become the same or worse over time.
I also found it interesting that the Weather Channel instinctively held a bias based on
consumer response to their forecasting models. They intentionally forecasted higher chances of rain
when rain had a chance to happen intentionally to ensure that if it did rain, the consumer (the
viewer), was informed and made an informed decision. This makes sense because of the boy who
cried wolf is very relevant to where if the weatherman didn’t forecast rain, and then it rained, people
are more apt to remember it than if they had done the opposite. If the weatherman forecasted rain,
and it hadn’t rained, the consumer was more likely to forget about it. This is definitely an eye opener
to understand perceptual biases moving forward.
