Why Horse Racing Selection Systems Have a Shelf Life

Firstly, what do I mean by a shelf-life?

This is the period in which a selection system is profitable.

Horse racing markets, be they betting exchanges or traditional bookmakers, are, for all intents and purposes, approximately perfect.

What do I mean by the term ‘a perfect market’?

In a perfect market, the strike rate of a horse is directly proportional to its odds. For example, if a horse is trading on a betting exchange at odds of, say, 4.0, on average, the horse will win one in four of its races. Likewise, if a horse is trading on a betting exchange at 25.0, on average, it will win one in twenty five of its races. A similar argument may be applied to the odds of horses that are offered by bookmakers.

Why is this important?

Well, let’s take this example:

Suppose that our system generates one selection in each of four races and the odds of each selection are 4.0 on a betting exchange.

Because the betting market is approximately perfect, on average, we would expect one of our selections to win and the other three to lose.

If we placed $1 on each of our four selections to win, we would lose $1 on each of the three losing horses making a total loss of $3. On the winning horse, we would win 1 x (4.0 – 1.0) = 1 x 3.0 = $3. However, betting exchanges levy a commission of 5% on all winning bets. Therefore, our winnings would be $3 (winnings) – $0.15 (commission) = $2.85. Overall, we would lose $3 and win only $2.85. Therefore, over the four bets, we would lose $3 – $2.85 = $0.85.

The above argument applies regardless of the odds of the selections and regardless of whether the four horses are backed to win or layed to lose.

Now, horse racing systems generate selections (horses). The selections will have odds quoted for them. If the odds of the selections generated by a system are summed over a long period of time and then averaged, the average odds will determine the long term strike rate of the system. If we then apply the above argument, we can deduce that no selection system can win, long term, if the market that the bets are placed in is approximately perfect.

This is a bold statement to make. So, what proof is there for this?

Firstly, I know of a web site that proofs the selections of over 300 horse racing selection systems. Although over the short term, some of the systems win, over the long term, they all lose.

Secondly, there are web sites that gather and analyse horse racing results. If the results are interrogated, they show that, indeed, there is a direct relationship between the odds of a horse and its strike rate. This implies that betting markets are approximately perfect.

Thirdly, there are web sites that cater for the testing of new selection systems. Over the short term, some of the systems do show a profit. Over the long term, none that I have ever witnessed have made a profit..

Fourthly, given the sizeable profits that the horse racing betting industry make, it is obvious that there must be far more losers who bet on horses than there are winners.

So, if, over the long term, all horse racing selection systems lose, why is it that some systems win over the short term?

For the most part, the short and long term strike rates of a system will remain approximately equal. At such times, a system will neither be profitable as a backing nor as a laying system i.e. it will break even if the betting exchange commission is ignored. When betting exchange commission is taken into account, the system will lose.

When the short term strike rate of a selection system rises above its long term strike rate, it will become profitable as a backing system since the selections will win more often than is required to achieve a break-even situation. When the short term strike rate of a selection system falls below its long term strike rate, it will become profitable as a laying system since the selections will lose more often than is required to achieve a break-even situation.

What causes the short term strike rate of a selection system to deviate from its long term strike rate?

Random chance, no more, no less.

Now let us consider the theory of Mean Reversion. This theory states that the short term mean will always gravitate towards the long term mean. If we substitute the strike rate of a selection system for the mean, the theory now states that the short term strike rate of a system will always gravitate towards its long term strike rate. Therefore, if random chance causes the short term strike rate of a system to deviate from its long term strike rate, the short term strike rate will eventually return to the long term strike rate.

In practice, when random chance causes a disturbance to the short term strike rate of a system such that it deviates from the long term strike rate, the short term strike rate will begin to oscillate. The oscillation will be centred around the long term strike rate. Over time, the oscillation will deteriorate. Eventually, it will deteriorate to the point where the oscillation totally ceases. At this point, the long and short term strike rates will become approximately equal and the system will, again, neither be profitable as a backing nor as a laying system. The system will remain in this state until such time as random chance disturbs the short term strike rate.

The shelf life of a profitable selection system is therefore that time between a system’s strike rate being disturbed by random chance and the strike rate returning back to its normal state.

There is a second reason why a system has a shelf-life.

When a backing system becomes profitable, it attracts subscribers. When it attracts subscribers, it becomes popular. When it becomes popular, so do the selections that the system generates. When this happens, the selections become ‘over bet’.

What is meant by ‘over-bet’?

A horse becomes over-bet when its odds fall below its true chances of winning.

What causes the odds to fall below the horse’s true chances of winning?

Too much money being placed on the horse to win.

When people start to bet money on a horse, its odds begin to fall. The more money that is placed on a horse, the more its odds fall.

Is this a problem for backers of the horse?

Yes, because backers of the horse will win less when the horse wins.

Why is this a problem?

Because the winnings will not be sufficient to off-set the losses incurred on those horses selected by the system which lose. As a result, the system will become unprofitable and its shelf-life will have expired.

So, even though the strike rate of the system will be unaffected by the increased interest in the selections, the system will become unprofitable because the odds on the selections will be lower than they ought to be.

Similar arguments can also be applied to systems which select horses to be layed.  In this case, the odds increase, rather than decrease.  This leads to increased losses.

So, there we have it. Two reasons why a system has a shelf-life.