Bill Gates wisely pointed out that "most people overestimate what they can do in one year and underestimate what they can do in 10 years".

This observation sheds light on planning for one's retirement. Most people rely on the assumptions of half-formed plans without really taking an in-depth look at their plan for their "golden years".

The process of retirement planning often starts with putting dreams and ambitions into tangible and defined goals.

Wouldn't it be nice for our retirement planning purposes if shares consistently gave us 8 to 10 percent returns each year? After all, that's what share markets have delivered on average over the very long term.

Indeed, between 1935 and 2016, US shares returned 11.4 percent annually, Canadian shares returned 9.6 percent, and international shares averaged 8.3 percent.

The trouble is that share returns are anything but predictable and so while they may average 8 to 10 percent over a 25 or 50-year period, every single year could deliver panic-inducing losses, euphoric gains, or something in-between.

Since 1988, the S&P 500 had single-year returns as low as negative 37 percent (2008) and gained as much as 37.58 percent (1995) in a single year.

Only in three of those 29 years did the S&P 500 deliver annual returns between 8 and 11 percent. The rest of the years were all over the place.

Why does this matter to your retirement planning? Because it's not enough to just plug "8 percent" into your retirement projections and call it a day.

Probability modelling

As far as retirement planning is concerned, probability modelling is somewhat related to gambling. It is a technique used to reduce the gamble that many people take when they decide to retire and live off their savings.

Probability modelling answers whether you will outlive your retirement assets or have enough money saved to last you well into your late 80s or 90s.

To understand probability modelling better, it is helpful to compare it to other types of retirement planning tools.

Many simple retirement calculations produce projections of required retirement savings using fixed average annual rates of return (eg, 7 per cent or 8 per cent) on investments.

These fixed figures are based on an expected investment strategy (eg, how much shares are held in an investor's portfolio) and historical rates of return.

While better than doing no calculation at all, fixed-rate planning tools have one major flaw. History tells us that there is not a diversified portfolio that will reliably produce the expected average return annually or even after decades have passed.

Instead, investment returns, especially for shares, are usually all over the map and can go through prolonged slumps such as in 2000-2002 and 2007-2009.

For example, if you count on an average annual return of 10 per cent over 30 years and spend according to this expectation, but instead average a return of 7 per cent, you might be living a severely reduced lifestyle by the time you are 80.

Of course, to be fair, there is also the probability that you will earn a higher return than planned. But this "upside risk" is not the one most people are worried about.

Instead, they are concerned about the chance of outliving their money.

So how does probability modelling help? By inserting additional criteria into a retirement planning calculation. Many financial planners use 30-year standard deviations to test the expected rate of return on retirement projections.

Standard deviation is a measure of market volatility (eg, highs and lows) of investment returns.

Financial advisers often use specialised software to randomly change the rate of return to cover a wide range of possible outcomes. With each change, the software records how much money a person is left with at the end of his or her life.

After a probability modelling is complete, a financial planner can show what per cent of the time a client still had money left over after a long period.

The client then seeks to craft a plan that provides both an acceptable spending level and an acceptable probability that assets won't be depleted.

Often, key variables in the simulation (eg, age at retirement and amount of money needed) are adjusted to find an outcome that works.

Of course, it is up to investors and their financial advisers to make necessary portfolio adjustments to match their desired simulated outcome.

They must also monitor, and revise, retirement plans as required to ensure no unpleasant surprises occur at a time down the road when a retiree can do little about it.

• This article represents general information only. Before making any financial or investment decisions, we recommend you consult a financial planner to consider your investment objectives, financial situation, and individual needs. A disclosure statement can be obtained free by calling 0800 878 961.