Strategy Idea!

Faros · Jan 12, 2011

Hi, I have been reading forum this morning and got interesting idea in my mind. There are so many strategies and EA, some of them are good, some off them are not very good.

The idea is simple and I can bet that somebody already had it but...

There are four strategies with 50% win rate. We need strategies that work on the same time period and use different instruments to make a forecast. Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%

Now what will we have if we take ten, fifteen systems. Yes we will have only couple trader per month but chance to win will be very high.

wackypete2 · Jan 12, 2011

You have a 50% win rate...so what happens if the 50% losses are greater than the 50% wins?

Peter

Hoggums · Jan 12, 2011

Faros said:
Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%

That is the most bonkers maths and reasoning I've ever seen!

So if three fail and one succeeds you still make a profit? That must mean you have a win lose ratio of minimum 3:1 - and the strategy wins 50% of the time???

If this is the case then you have the best trading system EVER created and you should be a multi-millionaire within months.

When you come back to reality try re-working your maths...

NVP · Jan 12, 2011

Faros said:
Hi, I have been reading forum this morning and got interesting idea in my mind. There are so many strategies and EA, some of them are good, some off them are not very good.

The idea is simple and I can bet that somebody already had it but...

There are four strategies with 50% win rate. We need strategies that work on the same time period and use different instruments to make a forecast. Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%

Now what will we have if we take ten, fifteen systems. Yes we will have only couple trader per month but chance to win will be very high.

hey mate

brush up on your basic probability and then also revise the values/effect of extended consecutive losing runs based on your set percentage win/lose ratios

then contact LTCM and ask them how their year went in 1998

NVP

Shakone · Jan 12, 2011

I think what the opening poster is trying to suggest, is that you take 4 systems each with around a 50% chance of winning. You don'ttrade all 4, you just wait until all 4 systems are lining up in the same direction on the same timeframe, and take that trade, hoping that the chance of all 4 being wrong is 6.25%.

The maths doesn't work out though faros, since there is dependency. The concept of using different factors to make your trade less likeyl to fail is a good conept though, and lots of people use filters for their trades (trend filter for example), or take trades with confluence of different types of support and so on.

alexander · Jan 12, 2011

Faros said:
There are four strategies with 50% win rate. We need strategies that work on the same time period and use different instruments to make a forecast. Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%.

That is only true if the strategies have 0 correlation, which is very unlikely given that most indicators are derivatives of one another and most currencies are correlated.

Hotch · Jan 12, 2011

This is hilarious.

Strategy Idea!

Faros

Member

wackypete2

Legendary member

Hoggums

Senior member

NVP

Guest Author

Shakone

Senior member

alexander

Well-known member

Hotch

Experienced member

Similar threads