Truth, Accuracy and the Myth of the Tested System
When I write an article, I always have some idea of what I want to say and how I want to say it. On the odd occasion, the article doesn’t always turn out the way that I had planned and the end-point turns out not to be the one that I envisaged. As I begin to put pen to paper, as it were, and write this article, I develop a sneaking suspicion that this is going to be one of those ‘odd’ occasions where the end-point won’t be the envisioned one. This is an interesting approach for me because it spurs me on to complete the article in order to find out how it ends. Hey, ho, such is life.
Let us ponder this question: Does accuracy imply truth and vice versa?
No, and here’s a good example as to why.
‘My height lies somewhere between 1 inch and 1 mile’, is a true statement, but hardly an accurate one.
‘My height is 11.125789247 inches’, is certainly an accurate statement since it includes nine decimal places. However, the statement is untrue because I am, in fact, five foot seven inches tall.
So, what’s all this got to do with the testing of a system?
Systems are tested in order to create ‘facts’ relating to the system’s performance. These ‘facts’ can then be presented to those who would either purchase the system, or its selections. The scrupulous purveyors of such systems and selections will test/proof their systems for at least six months, prior to sale, (and will be able to prove this). System statistics, which are accurate to two decimal places, will be available by the yard in order to convince the would-be buyers that the system and its selections are ‘good and wholesome’.
The fact that the statistics are accurate to two decimal places most certainly does imply accuracy, but does it imply truth – that the system will produce a future profit? Well, yes, it does IMPLY truth but it may not necessarily BE the truth.
To illustrate this point, I have a (true) story to tell.
Last year, I was contacted by a company that placed lay bets on behalf of its members using the funds that they had provided. They suggested that I invest a lump sum of £5,000, but would give due consideration to me investing more. Their profitability claims were quite staggering. I investigated their claims and found them to be totally accurate.
The company claimed that their system had a strike rate of over 99.99%. I investigated this also and found that, within the context in which the system was being operated, this figure was also accurate.
The company claimed that, only once in the two-year history of the system, had they had two consecutive losing bets. They claimed that 3 losing bets in a row was an impossibility.
All employees of the company were invited on occasional long week-end breaks in the sun, at the company’s expense, such were the profits being generated by the system.
I hasten to add at this point that I did not accept their offer, in spite of the fact that their claims appeared to be true.
Then, one day, in mid-December 2007, the ‘impossible’ happened. The system generated 3 lay selections. All 3 selections won. Within the space of two hours, not only had they lost all of the profit accumulated by the system over the previous two years, they lost a lot more besides. In fact, they lost the whole of their betting bank.
The Directors of the company re-mortgaged their homes, sold their cars and took out loans in order to create a new bank to recover their losses and generate further profits. The reason that they did this was because lightening had struck once where it should never have struck at all. Now, they considered that lightening would never strike twice in the same place and that the system was completely bullet-proof.
Within a couple of weeks, the system was running again and generating profits.
Then, one day in late January 2008, the system identified 5 lay selections. The ‘impossible’ happened again when three of them won in succession. Within the space of two hours, they lost the whole of the profits generated over the previous month and a lot more besides. The whole of their betting bank had been lost. In fact, with the space of six weeks, they had lost two complete betting banks.
Unable to raise yet more funds, the company went into liquidation owing more that £330,000. All of the company’s employees lost their jobs and those who invested lump sums in the system lost all of their money.
So, there we have it. The system was fully tested for six months before it went live and behaved impeccably for almost two years thereafter. Then, without warning, it failed in a spectacular way, not once, but twice within the space of six weeks. After it failed the second time, I continued to monitor the system. I found that it’s performance was nothing short of disgraceful.
What this example shows is that the past performance of a system is not a reliable indicator of its future performance.
There is another fact that most would-be buyers of systems and tips fail to appreciate. When a system is tested, theoretical, rather than actual, bets are placed on the selections generated by the system. The system is therefore tested within the context that no bets are physically placed on any of the selections. Profits from the system are therefore calculated within this context.
When the system goes live, bets will be placed by customers on the selections generated by the system. This will adversely affect the odds of the selections. The more customers there are, the more adverse the affect will be. As a result, the value in the selections will either be much reduced or will totally disappear. Profits, in turn, will be adversely affected. In fact, such may be the reduction in the value of the selections that losses may be incurred. No amount of testing can calculate what effect this will have.
There is one last thing that I would like to draw your attention to. I receive a lot of emails about one subject in particular. The emails generally take the form: ‘I followed system X for several months, during which time it made a substantial profit. The minute I joined, it entered a losing phase and I lost a considerable amount of money.’ I received so many emails on the subject that I decided to do some research to determine why this was. Not only did I investigate my own systems, I also investigated over 100 systems created by other people. My research indicated that all systems are essentially cyclic in their nature in that they go through profitable phases and then unprofitable ones. I also found, as a general rule, that the more profitable a system had been in the past, the more unprofitable the system will be in the future. I also found that, generally, the longer the profitable period lasted, the longer the unprofitable period lasted. Now, when is a new system, or the tips generated by it, put up for sale? During one of the system’s less successful phases or during one of its successful ones? Obviously during one of its more successful phases. What follows a successful phase? An unsuccessful one. Therefore, customers generally join a tipping service just in time for it to go into one of its losing phases. They then wonder why they lose money when they first join a tipping service!
The conclusion that I would therefore draw, from the above, is that the only thing that system testing can tell us is that the system was tested. Nothing more and nothing less. The past performance of a system is not a reliable indicator of its future performance. To draw any other conclusions from system testing is sheer folly.
Psycho
Related Posts:
- Betfair Risk Free Betting – Myth Or Reality? It may not exactly be alchemy but the idea of...
- Betting Systems – Can You Really Win? I’ve been a full-time gambler for well over 5 years,...
- Lay to Lose – Gold at the End of the Betfair Rainbow? Truth and Fiction, Part 1 What is Lay to Lose: Ultimately trading horse races on...
- Horse Racing – How I Transformed Myself From Loser to Winner Being a horse racing punter is easy, you just go...
- Lay to Lose on Betfair – Gold at the End of the Rainbow? Truth and Fiction – Part 2 As an example let us assume that our chosen favourite...