The Challenge of Predicting Individual Events

Why is it so awfully difficult to predict single events accurately, when the synergy of multiple events aggregated together seems far easier to guess?

Many people are able to predict how things will develop in their personal lives over a period of, say, the next 12 months. For example, job prospects or personal events.

If you take the time to write down all the expected events planned or envisaged in your life over the next 12 months, you will probably realise in a year’s time that most of them have actually happened and more or less as expected.

Perhaps you started with things that are set in stone, the fact that you already have two holidays planned abroad, and more importantly (or not, as the case may be!) that Aunt Emma will be celebrating her 90th Birthday with a small party at your house.

You are kind of expecting a mediocre school report from your son, and you have the gut feeling that next year is finally the one where you are promoted to ‘Head of Department’ at work…

A year on and time for your personal review.

Surprisingly, each of your predictions happened, more or less, but the details turned out to be very, very different from your expectations!

The vacation plans went a bit awry and instead of the usual lazy, beach holidays you went to the mountains.

Aunt Emma made it to her birthday but the party was cancelled, replaced by an even larger and unexpected family gathering at a local hotel.

Your son’s school results were as expected but instead of the usual low maths mark, he made the top half of the class and slipped down the order in sports instead.

You are indeed now Head of Department, exactly as expected, but in a different company.

What we’re trying to say is that although overall trends can be predicted with a fine degree of accuracy, it is almost impossible to make an accurate prediction about a single event which is just one small part of the whole chain.

Sales Predictability

Another example, from the world of sales.

Once a year, professional companies plan budgets by reviewing their existing client portfolios estimating renewals and/or future sales.

Any professional can predict the increase in the number of customers by ‘X’%. However, it is impossible to say which customers will remain faithful, or how many new clients will be converted.

The only certainty is that after a year there will be ‘X’% more sales. 40% of existing customers will increase their purchases, but which ones will is anyone’s guess.

Of all the budget predictions for the forthcoming year, most will be correct but some will probably have disappeared into thin air to be replaced by positive events that were totally unexpected.

Otherwise, the final overall balance turns out pretty much as predicted.

Predictions in Football

Before the 2011-12 season our statistics suggested Bayern Munich would have just one or two 0-0 results in their home fixtures in the Bundesliga, but obviously nobody could tell beforehand which games would end goal-less.

Their campaign concluded with just one 0-0 result at home, on 14.04.2012 against Mainz.

The same prediction also applied for one goal matches in Bayern home matches.

There were two such games, 07.08.2011 against Mönchengladbach (0-1) and 19.11.2011 against Dortmund (0-1).

The statistics predicted Bayern would have between two and four home matches with less than 1.5 goals.

During 2011-12 they played three such matches.