Why do systems fail?

[pika]

Good points Jalfrezi as usual. Point 2 is particularly insightful I think most tend to be blind to that!

S/R can be quite misleading too as it is just winners/total bets. It doesn't say how or when the winners come about. The could all be cluster together. So 10% S/R over 100 bets could have all 10 winners coming in a sequence of about 20 bets or so!

 

[arkle55]

Thanks for the replies everyone.

Arkle

 

[MrDamon]

Reminds of Nicholas Taleb Nasim and his book "Fooled by Randomness". Great book. Very illuminating on the limits of human knowledge and excess of human pride.

The first time I read The Drunkard's Walk: How Randomness Rules Our Lives by Leonard Mlodinow I felt almost liberated and have never looked back. I will say, it is difficult sometimes to recognise when the data you have is being misused, but as Dave said, keeping it as simple and logical as possible does help a lot..

I've definitely seen a major improvement in my work and results since I read it and I cringe now when I see and hear some of the ideas and findings people arrive at through the misuse of data.

 

[pika]

Welcome to the site MrDamon!!

Saw a quote somewhere: If you torture the data hard enough it will confess to anything.

 

[MrDamon]

Hahaha brilliant! We need a Like/Thanks feature

 

[ArkRoyal]

Hahaha brilliant! We need a Like/Thanks feature

The best we can do at the moment is

 

[mlmrob]

Form of racehorses is cyclical and therefore a system based on a set of rules is bound to have counter-cyclical periods.

On the stock markets investors will choose to buy into cyclical businesses if they believe that the economy will do better than the market expects.
Some sectors may follow cycles that lead or lag general economic growth. An important example of this is that consumer and business demand are rarely perfectly synchronised.

It is the same with systems.