Capping IT Off

Capping IT Off

Opinions expressed on this blog reflect the writer’s views and not the position of the Capgemini Group

Measuring Predictions - or why astrology is bunk

Predictions are easy to make, I can boldly predict that in 2013 we will discover the secret to cold fusion... if I'm wrong I can make the same prediction next year, and every year until we crack it. Thus on the time I'm right I can exclaim 'I told you so'. Clearly however such predictions are nonsense as its simply a case of 'buying a ticket' and then hoping the prediction comes right this time. Predictions can also be wonderfully vague such as "If there is someone you love, allow yourself to love them. If there is someone you trust, allow yourself to trust them. And if there is someone you are not sure of? Don't allow this to become the reason why you refuse to respect a word they ever say. Many things can explain an odd reaction to a particular person. It may say as much about a failing on our part, as a fault in theirs. I'm not suggesting that you become naive or that you ignore your instincts entirely; but be a little generous with the benefit of the doubt." which I pulled today from an Jonathan Cainer's astrology site. This really isn't even a prediction as its not something on which a real decision could actually be made around an action, its certainly not in anyway a guide to the future, a 'forecast' as it says. So this means we've two clear pieces for a Prediction to be considered a good one

  1. It should be measurable - it should allow actions to be measured against the prediction and give guidance as to what actions should be taken
  2. It should be measured in its broad context - not simply as a 'it was right' but understanding how many times the forecast was right.
  3. It should have a timeframe against the forecast - not simply 'the Sun will explode in the future' type things
Underpinning both of these is a principle, that predictions and forecasts should be measured. Astrology doesn't live up to either of these,here is a fun article summing some of the research on the topic, and so clearly shouldn't be used by a business as a basis to make rational decisions. However on many occasions businesses are doing little better. A forecast for growth is wrong, but the reasons that it was wrong are not investigated another forecast is just thrown up in the hope that it is accurate. The company rejoices when a forecast is accurate in the supply chain, but ignores it as an anomaly when it goes wrong. A marketing campaign is forecast to have a 5% conversion rate and comes in at 8% while another comes in at 2%, one is considered 'good' the other 'bad'... but isn't the real issue that both were forecast incorrectly? If your forecasts and guidance to your business are as vague as the astrology forecast above, or if you don't measure and adapt your forecasts based on what actually happened then you are doing little better than business astrology. The technologies and approaches are out there to massively improve this situation and the business leaders who adopt them will find themselves streaking ahead of the competition. when it goes wrong

About the author

Steve Jones
1 Comment Leave a comment
kpadinja's picture
Nice Post.. One key thing is prediction is the probability of the event happening.. (Confidence level).. if you can predict with 100% Confidence, 100% of the time- then you are GOD.. since we are humans, all predictions should be qualified with a confidence level.. so when the weather man says it will snow with 100% confidence..it should snow.. a prediction with 50% chance is no good.. It is as good as a flip of a coin.. I would say any predictions above 50% confidence are good.. and anything above 80% confidence is brilliant.. and anything below 50% .. is really bad and bogus.. (just like most astrology predictions today). Any prediction should be qualified with a confidence level of the prediction.if you look at weather.com they always qualify weather predictions with a "Chance of.. XX%" today the accuracy of a 24 hour forcast is brilliant.. the accuracy of a 5 day forecast is fairly good, and a beyond 10 days is almost bogus.. I personally don’t believe Astrology is a Bunk.. The prediction part of Astrology is a Bunk because of bad forecasters.. but the actual science behind (Indian) Astrology (i.e. Astronomy and the actual individual horoscope ) is based on solid mathematics and concepts of Big Data.. unfortunately we don’t have good astrologers (forecasters) who can take the mathematical data and make good predictions today..who knows with advanced analaytics .. astrological predictions may become better just like weather predictions..

Leave a comment

Your email address will not be published. Required fields are marked *.