Coding Challenge

[Dua Yong Rew]

Let's play a game.

Let's select a strategy to code and we see how our code earn different amount of money.

Rules

1. Select a trading system (How to enter or exit should be stated.)

2. Select a time frame and a testing period

3. Select a symbol

 

Anyone game for it? 

[Simon Gniadkowski]

doshur:

Let's play a game.

Let's select a strategy to code and we see how our code earn different amount of money.

Rules

1. Select a trading system (How to enter or exit should be stated.)

2. Select a time frame and a testing period

3. Select a symbol

 

Anyone game for it? 

If all the solutions are correct they will all have coded the Strategy correctly so will all generate the same result for the same symbol, timeframe and date range.  You need to evaluate in a different way . . .  if the subject is coding evaluate the code.

[Dua Yong Rew]

RaptorUK:
If all the solutions are correct they will all have coded the Strategy correctly so will all generate the same result for the same symbol, timeframe and date range.  You need to evaluate in a different way . . .  if the subject is coding evaluate the code.

Maybe point one should change. Example if we trade certain candle stick pattern. Not all solutions will generate equal profit.

[Ubzen]

doshur:

Let's play a game.

Let's select a strategy to code and we see how our code earn different amount of money.

Rules

1. Select a trading system (How to enter or exit should be stated.)

2. Select a time frame and a testing period

3. Select a symbol

 

Anyone game for it? 

Wooo, games ... me like games. Ok I'm In. You choose the trading strategy. Do we have to post the codes ?

[Simon Gniadkowski]

doshur:
Maybe point one should change. Example if we trade certain candle stick pattern. Not all solutions will generate equal profit.

So then it's no longer a coding test but a test of the Strategy ?

[Alain Verleyen]

doshur:
Maybe point one should change. Example if we trade certain candle stick pattern. Not all solutions will generate equal profit.

In my opinion you have to define the rules precisely, and then ask who wants to participate in the challenge. If you try to define the challenge itself on the forum, this will lead to eternal discussion.

But the idea is good.

[Ubzen]

angevoyageur: In my opinion you have to define the rules precisely, and then ask who wants to participate in the challenge. If you try to define the challenge itself on the forum, this will lead to eternal discussion.

But the idea is good.

Yep. I'm attempting to do something similar for mql4_forum. Hope it doesn't become eternal.

[Dua Yong Rew]

Ubzen:
Wooo, games ... me like games. Ok I'm In. You choose the trading strategy. Do we have to post the codes ?

Nope. We will compare trade result

[Ubzen]

doshur: Nope. We will compare trade result

You have more explaining to do.

Is this going to be a Live test for earning || or a back-test for earning?

Does everyone choose their own strategy ||or does everyone code a selected strategy?

If the codes are hidden and its based on back-test than it not much of a challenge Imo.

[Dua Yong Rew]

oops, maybe I'm not too clear about this challenge.

First, we will select a strategy. I will go to FF and check out which is a interesting price action strategy. we will all code this.

Second, all our expert will run in back test mode for a pre defined period, let's say the whole of 2013, in a fixed timeframe like H1 on the same broker.

we are free to code our own filters, trail stop, break even...

we will upload back test results and compare the results. 

[Ubzen]

doshur:

oops, maybe I'm not too clear about this challenge.

First, we will select a strategy. I will go to FF and check out which is a interesting price action strategy. we will all code this.

Second, all our expert will run in back test mode for a pre defined period, let's say the whole of 2013, in a fixed timeframe like H1 on the same broker.

we are free to code our own filters, trail stop, break even...

we will upload back test results and compare the results. 

Wouldn't that lead to curve fitting with maximum lot sizes?