Let's Try an Experiment

[Mysticforex 13]

Let's try some random trading. For 10 trading days, starting Monday 7/22 thru Friday 8/2 we will open 1 trade on my Oanda demo account.

It will be 1 trade a day opened at the NY Open ( 09:30 am EST ) The size will be 1 dollar a pip. The trade will have a 20 pip TP and a 20 pip S/L If for some reason the trade has not closed by 5 pm of the same day it will be closed ( no holding trades over night ).

The pair will be EUR/USD.

The method we will use to determine the direction of the trade will be a coin toss. Heads is buy, tails is sell.

I will flip and call the first trade on Mon 7/22. I need 9 more members to register here in the thread to flip and call the remaining trades ( one each ).. First come first served. Please try to have the trade call in 12 hours before market open.

 

After the experiment we will decide if it was bullshit and have a good laugh, or maybe realize that with a little fine tuning ,,,, well who knows.

[GlassOnion 12]

I will take Weds 7/24. Heads/buy

[Mysticforex 13]

I will take Weds 7/24. Heads/buy

 

 

Noted G. O. you're in.

[SIUYA 11]

coin flipped.....tails.

I also randomly picked the Friday 8/2

[optiontimer 10]

Flipped a nickel and it came up tails for 7/23.

 

Mystic, you might want to edit your initial post to include a list of the dates taken, and whether heads or tails.

[zupcon 10]

random trading systems, can I possibly resist ?

 

Yes, I think I will, for the time being at least.

[tradingwizzard 10]

interesting..........will sit on the sidelines and watch this one......if you will need one more participant at the end, I will jump in.....watching

[Mysticforex 13]

Flipped a nickel and it came up tails for 7/23.

 

Mystic, you might want to edit your initial post to include a list of the dates taken, and whether heads or tails.

 

Optiontimer, was planning to do that at the end of each day till filled. Then post a final list.

 

Here it is so far:

Me 7/22 Mon. Heads - Buy

 

Glassonion 7/24 weds. Heads/Buy

 

Siuya 8/2 Fri. Tails - sell

 

Optiontimer 7/23 Tues. Tails/Sell

 

 

So far I am putting them in the order posted. When all the spots are filled I will post them in chronological order.

[jpennybags 11]

Well, this is just stupid... flippin da coin:

 

Tails / Sell 7/25

[Mysticforex 13]

I should point out that thist is not about winning or losing. It does not matter who flipped a Tails on a day the pair goes down. It is just about a group of traders who got together to amuse themselves.

[BlueHorseshoe 10]

I'd love to get involved, but I'm a little short on change . . . Could anyone possibly lend me a coin?

 

BlueHorseshoe

[zupcon 10]

I'd love to get involved, but I'm a little short on change . . . Could anyone possibly lend me a coin?

 

BlueHorseshoe

 

RANDOM.ORG - Coin Flipper

[SIUYA 11]

I'd love to get involved, but I'm a little short on change . . . Could anyone possibly lend me a coin?

 

BlueHorseshoe

 

I did one better for you Blue....

the first serial number on the £20 note I pulled out of my pocket started with a D - hence you can have a sell day (D for down)

The last number ended in 1 which would give you Thursday 1st....

Random enough??

 

Blue - sell 1st August.

[Mysticforex 13]

To date:

 

Here it is so far:

Me 7/22 Mon. Heads - Buy

 

Glassonion 7/24 weds. Heads/Buy

 

Siuya 8/2 Fri. Tails - sell

 

Optiontimer 7/23 Tues. Tails/Sell

 

 

So far I am putting them in the order posted. When all the spots are filled I will post them in chronological order.

 

and Drum Roll Please..............................

 

Jpennybags, 7/25 SELL.

 

( why am I starting to feel like a bookie ? )

[BlueHorseshoe 10]

 

( why am I starting to feel like a bookie ? )

 

I've never trusted that Siuya, so I flipped a coin for Thursday 25th . . .

 

Heads - Buy.

 

BlueHorseshoe

[jperl 11]

Unfortunately guys/gals 10 coin flips is not going to tell you much. You need something

like 1000 flips to get any realistic statistical data.

[BlueHorseshoe 10]

Unfortunately guys/gals 10 coin flips is not going to tell you much. You need something

like 1000 flips to get any realistic statistical data.

 

Hi Jerry,

 

I'm not sure how logically consistent your statement is . . .

 

Do you think that 1000 coin flips will suddenly become something other than random? If every coin flip from the ten produces a winning trade, will we suddenly have a great trading system in our hands? If 1000 coin flips were to produce 1000 winning trades (can happen, but I can't do the maths!), what then?

 

Larger sample sizes usually mean greater statistical significance, but this is not always the case.

 

Anyway, this is all just a bit of entertainment

 

BlueHorseshoe

[SIUYA 11]

Unfortunately guys/gals 10 coin flips is not going to tell you much. You need something

like 1000 flips to get any realistic statistical data.

 

are you kiddin!

You think we are after statistical data - leave that alone - we are after a system.....something that will take all the hard work out it and be a money making machine.

if it works then we have a system that we can collectively market, sell and make plenty and show to us that you can make money being a complete idiot with totally random systems.....at least thats enough to fool some.....

[Mysticforex 13]

I've never trusted that Siuya, so I flipped a coin for Thursday 25th . . .

 

Heads - Buy.

 

BlueHorseshoe

 

Bluehorseshoe added to the mix.

Thus. 7/25 Heads/Buy

[Mysticforex 13]

Unfortunately guys/gals 10 coin flips is not going to tell you much. You need something

like 1000 flips to get any realistic statistical data.

 

 

Not to worry Jperl... just having some fun.

[jperl 11]

 

 

 

Do you think that 1000 coin flips will suddenly become something other than random? If every coin flip from the ten produces a winning trade, will we suddenly have a great trading system in our hands? If 1000 coin flips were to produce 1000 winning trades (can happen, but I can't do the maths!), what then?

 

Larger sample sizes usually mean greater statistical significance, but this is not always the case.

 

BlueHorseshoe

 

For what it's worth, fair coin tosses should follow a binomial distribution. Stock prices do not. So any system based on some fixed random distribution, binomial or otherwise will ultimately fail.

As far as guessing 1000 winning trades in a row, it's 1/2^1000. Even getting 10 winning trades in a row is pretty small, approx 1 in 1000.

Enjoy the entertainment.

[jpennybags 11]

Bluehorseshoe added to the mix.

Thus. 7/25 Heads/Buy

 

Wait... I had 7/25... tails / sell. You can move one of us to 7/26... like it will really make a difference.

[BlueHorseshoe 10]

For what it's worth, fair coin tosses should follow a binomial distribution. Stock prices do not. So any system based on some fixed random distribution, binomial or otherwise will ultimately fail.

As far as guessing 1000 winning trades in a row, it's 1/2^1000. Even getting 10 winning trades in a row is pretty small, approx 1 in 1000.

Enjoy the entertainment.

 

Binomial might be a bit of a red herring - a market always closes up or down from it's prior close - that's a binomial outcome, isn't it?

 

Stock prices don't behave like a coin flip, I agree . . . but nor do they consistently behave unlike a coin flip. Sometimes they are auto-correlated; sometimes they are not. A bit like a coin that is only weighted sometimes, and truly random the rest.

 

Markets only have to behave like a coin flip some of the time in order to make money from this observation. You just have to lose less when they are non-random than you make when they are random. And when they are non-random, because you assume random-ness, you'll still be correct 50% of the time - that is to say, if you flip a coin to decide up or down, and the market is only ever up, then you'll still be right as often as you are wrong.

 

BlueHorseshoe

[TheRumpledOne 14]

Let's try some random trading. For 10 trading days, starting Monday 7/22 thru Friday 8/2 we will open 1 trade on my Oanda demo account.

It will be 1 trade a day opened at the NY Open ( 09:30 am EST ) The size will be 1 dollar a pip. The trade will have a 20 pip TP and a 20 pip S/L If for some reason the trade has not closed by 5 pm of the same day it will be closed ( no holding trades over night ).

The pair will be EUR/USD.

The method we will use to determine the direction of the trade will be a coin toss. Heads is buy, tails is sell.

I will flip and call the first trade on Mon 7/22. I need 9 more members to register here in the thread to flip and call the remaining trades ( one each ).. First come first served. Please try to have the trade call in 12 hours before market open.

 

After the experiment we will decide if it was bullshit and have a good laugh, or maybe realize that with a little fine tuning ,,,, well who knows.

 

Have you read FOOLED BY RANDOMNESS, THE BLACK SWAN, or THE SIGNAL AND THE NOISE?

 

Your experiment will not prove a thing, IMHO.

 

How about flipping the coin ONCE for the entire experiment, heads go long, tails go short. Use the WEEKLY ATR of the EUR/USD for a stop loss and trade $.10 per pip.

[TheRumpledOne 14]

 

Position size based on stop loss placement can keep you in the game a lot longer.

 

Like playing Blackjack.. you are not going to win every hand, but if you bust, you are never going to win.