Statistical Gambling System (High Winrate, but NO EDGE) | Page 18 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Made a 100% Winrate EA 26 replies

what can we do with 90% winrate? 130 replies

MACD +60% Winrate 1:1.5 RR 63 replies

Does anyone use a statistical edge 0 replies

How do you gain, and measure, your trading system's statistical edge? 29 replies

Attachments: Statistical Gambling System (High Winrate, but NO EDGE)
Exit Attachments

Statistical Gambling System (High Winrate, but NO EDGE)

  • #341
  • Jan 31, 2014 11:48am Jan 31, 2014 11:48am
Quoting Proximus
{quote} Thanks for the idea, i`ve never heard of that method, but i`ll look into it.At the moment i`m trying to analyze the cycles of the market, even though we try to build a mathematical model of the market, we still cant ignore the human aspect of it. It is a known fact that when new market sessions open, more volatility can be expected, and in this system we try to avoid that.So far i`ve found out the the best time to trade is between17:30 -> 5:00 GMT+0, because in this time range the volatility is the lowest.But still researching. ________...
Hi Proxy, surely you want the volatility its a gift for grid systems, up down every thing gets triggered
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #342
  • Feb 1, 2014 7:52am Feb 1, 2014 7:52am
Quoting billytt
{quote} Hi Proxy, surely you want the volatility its a gift for grid systems, up down every thing gets triggered
This system avoids volatility events, because we only scalp the small deviations from the daily MID price.Sure a breakout from the mid price can be helpful, but look the other way if it goes against us, usually the price tends to jump to that new level, and then it will stay there for a while.

For our NO-SL system, that can be very harmful
"There's a sucker born every minute" - P.T. Barnum
  • #343
  • Feb 2, 2014 1:19pm Feb 2, 2014 1:19pm
hello Proximus, how do you calculate your mid high and low levels please ?
  • #344
  • Feb 2, 2014 4:23pm Feb 2, 2014 4:23pm
Quoting metta87
hello Proximus, how do you calculate your mid high and low levels please ?

Its easy, you can get that from the indicator, but for simplicity here is the formula:

HIGH = previous day's high
LOW = previous day's low
CLOSE =previous day's close

pivot = (HIGH + LOW + CLOSE) / 3;

DAILYMAXHIGH= (2 * pivot) - LOW;
DAILYMAXLOW = (2 * pivot) - HIGH;
MID = ((4 * pivot) - LOW - HIGH)/2 ;
DAILY_MEDIUM_HIGH= (MID + (2 * pivot) - LOW) / 2;
DAILY_MEDIUM_LOW = (MID + (2 * pivot) - HIGH) / 2;
"There's a sucker born every minute" - P.T. Barnum
  • #345
  • Feb 2, 2014 4:46pm Feb 2, 2014 4:46pm
Hi proxy. Your daily pivots. Take it one step further calculate the pivot from example the uk session or/and us session and play them against each other. My girl used to calculate them for me
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #346
  • Feb 2, 2014 9:38pm Feb 2, 2014 9:38pm
Quoting Proximus
Wow Magix WTF is this? {image} Looks like your EA is not doing well, i have to remove it from this thread since its ruining my system's reputation. Besides my new 1.5 system will probably be better than the current forward test , that i`m doing. Why are you suspended again Magix ?
Prox...

My system was designed to take a loss as needed.

Looking at it...I'm still ahead in profit.

Best not put it back up, it will only embarrass you further.

BTW...this new EA of yours, is it designed to beat the Market "Statistically", by never trading??
Statistically, that is the only edge.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #347
  • Feb 3, 2014 7:41am Feb 3, 2014 7:41am
Quoting Magix
{quote} Prox... My system was designed to take a loss as needed. Looking at it...I'm still ahead in profit. Best not put it back up, it will only embarrass you further. BTW...this new EA of yours, is it designed to beat the Market "Statistically", by never trading?? Statistically, that is the only edge.
Welcome back Magix

Nope i consciously didnt traded for a week, i`m too busy now with another project, that is why i asked that anyone else can test it for me, the EA no matter what will have atleast 4-5 trades/day if its ran 24/7 , i will put the EA here and perhaps if someone is interested can do the forward test for me on a demo account.Because i dont really have time now
"There's a sucker born every minute" - P.T. Barnum
  • #348
  • Feb 3, 2014 9:02am Feb 3, 2014 9:02am
Quoting Proximus
{quote} Welcome back Magix Nope i consciously didnt traded for a week, i`m too busy now with another project, that is why i asked that anyone else can test it for me, the EA no matter what will have atleast 4-5 trades/day if its ran 24/7 , i will put the EA here and perhaps if someone is interested can do the forward test for me on a demo account.Because i dont really have time now
Prox,

I fail to see how setting an EA on an account consumes any time...given the basic logic, I don't even think your EA would consume much by way of resources.

There is an infinite number of free accounts you can open, and even for my personal PC, albeit a little dated, can and does run 7 platforms at the same time without hiccups.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #349
  • Feb 3, 2014 10:54am Feb 3, 2014 10:54am
Quoting Magix
{quote} Prox, I fail to see how setting an EA on an account consumes any time...given the basic logic, I don't even think your EA would consume much by way of resources. There is an infinite number of free accounts you can open, and even for my personal PC, albeit a little dated, can and does run 7 platforms at the same time without hiccups.

Its not about time, demo accounts or anything, i have 3 PC's and i run them all, and need all of their computational power to backtest my theories.I dont have an extra VPS to run an EA 24/7.

Its very time and resource consuming to run my backtests, with my high quality data that i have it usually takes 2-3 hours /test just to backtest only 1 configuration on my entire database.And thats on an I7 processor...

So all my computational power is required elsewhere at the moment, but when my research is done then i will continue my forward test.You must understand that

So in the meantime if anyone is interested to run my EA on a demo account , feel free to do so, but if you do, please create a TE and post the name of it here so i can attach it to this thread and see how well it goes

I`ll upload the EA in a few minutes , just making the last checks on it.
"There's a sucker born every minute" - P.T. Barnum
  • #350
  • Feb 3, 2014 11:14am Feb 3, 2014 11:14am
Quoting Proximus
Its very time and resource consuming to run my backtests, with my high quality data that i have it usually takes 2-3 hours /test just to backtest only 1 configuration on my entire database
I know you will refuse again to listen to actual fact on this, in me stating that your time is completely and 100% wasted on back test results no matter what quality you get...so ask yourself the simple question:

If backtests worked, why aren't there more successful EAs available?

Forward tests are the ONLY way to gain any sort of market semblance and understanding of true performance.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #351
  • Feb 3, 2014 11:21am Feb 3, 2014 11:21am
Quoting Magix
{quote} I know you will refuse again to listen to actual fact on this, in me stating that your time is completely and 100% wasted on back test results no matter what quality you get...so ask yourself the simple question: If backtests worked, why aren't there more successful EAs available? Forward tests are the ONLY way to gain any sort of market semblance and understanding of true performance.

You have a point there, but i`m not backtesting in the conventional way, i`m more like data mining, finding when the market is good for a trading purpose, relative to a time period.So i`ll have a better understanding about the flow of price and when or at which time the traders are more biased than other.

Get ready for my EA

PS: I also edited the indicator, and now we will take into consideration the monthly highs and lows also, as promised, sorry for the delayed update
"There's a sucker born every minute" - P.T. Barnum
  • #352
  • Feb 3, 2014 12:17pm Feb 3, 2014 12:17pm

STATISTICAL GAMBLING EA V1.5

(Based mostly on the V1.5 of the system, but modified a little bit)

Features:

  1. Configurable TP (35 pipette default as the system rules of v1.5 )
  2. Configurable maximum slippage & spread (20 pipette spread & 10 pipette slippage as default)
  3. Configurable risk (set to 1, and its recommended to not change it)
  4. Graphical Interface (to show additional information on the chart)
  5. Updated APROXIMATOR indicator , we will take into consideration the monthly high & lows aswell, and also changed the formula of the pivot, we will now focus on the median price, instead of the typical price.

-Use it on a DEMO account for the purpose to forward test the new strategy and the new indicator, if you decide to use it on a REAL account, then i`m not responsible for the losses it could make.


Here is a quick backtest for year 2013 (until september for the price to have time to reverse):

Attached Image (click to enlarge)
Click to Enlarge Name: a.gif Size: 12 KB


Attached File(s)
File Type: rar backtest v1.5.rar   15 KB | 589 downloads


Instalation: move the APROXIMATOR V1.1 in the indicator folder, and the STATISTICAL GAMBLING EA V1.5 EA in the experts folder.

If you forward test my system on demo acc, please make a TE and post the name/link of it here, so that i can attach it to this thread and so that we can follow the progress of my EA.Thanks!
Attached File(s)
File Type: mq4 APROXIMATOR V1.1.mq4   5 KB | 1,180 downloads
File Type: mq4 STATISTICAL GAMBLING EA V1.5.mq4   9 KB | 1,226 downloads
"There's a sucker born every minute" - P.T. Barnum
  • #353
  • Feb 3, 2014 3:25pm Feb 3, 2014 3:25pm
Quoting Proximus
STATISTICAL GAMBLING EA V1.5 (Based mostly on the V1.5 of the system, but modified a little bit) Features: Configurable TP (35 pipette default as the system rules of v1.5 ) Configurable maximum slippage & spread (20 pipette spread & 10 pipette slippage as default) Configurable risk (set to 1, and its recommended to not change it) Graphical Interface (to show additional information on the chart) Updated APROXIMATOR indicator , we will take into consideration the monthly high & lows aswell, and also changed the formula of the pivot, we will now focus...
If we assume that you have optimized it to 3.5 pips...

According to your tests, this has taken 141 trades, all of which were profitable, with exception of the last.
In an average year, there are 5 trading days per week, 4.2 weeks per month and of course, 12 months per year. From there, we are going to subtract the 2 biggie holidays that nobody trades, being Christmas and New Years and assume that this will work the rest of the time.
141 trades/250 days = .564 = 1 trade every 2 or so days...provided that it doesn't go into a week long DD like we have already seen.

To generate $36 worth of profit.


Now given that somebody that wanted to attempt this, should in fact allow for long draw down periods, and statistically, it is possible to catch upto 16 weeks of Draw Down, and knowing that Back tests do not actually calculate in swap fees, it would seem that the cost of running a statistical gambler would be done at the cost of power and resource consumption...

Meaning, if you had to pay for your power watched left $100 bucks in play for a year and dwindled away this account on broker fees, there is an incredible likelihood that you would actually come up at a loss.

I could be wrong...but I'm not.

It would seem that selling your hair or collecting pop cans would actually be more prosperous than this venture...as far as entertainment value, it'd come up as a zero.

Hello forest, I'm sorry we missed you by looking at all of these trees...

But good luck in testing it!
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #354
  • Feb 3, 2014 3:35pm Feb 3, 2014 3:35pm
Wait...

Hold the presses...

Something I may have forgot to mention.

The above might seem a little negative, but that is actually, statistically, the Best Case Scenario.



There is the actual possibility of holding on forever below your break even because of continued loss on swaps during long periods of DD.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #355
  • Feb 3, 2014 6:02pm Feb 3, 2014 6:02pm
Uh...yeah...
Attached Image
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #356
  • Feb 4, 2014 10:20am Feb 4, 2014 10:20am
Quoting Magix
{quote} If we assume that you have optimized it to 3.5 pips... According to your tests, this has taken 141 trades, all of which were profitable, with exception of the last. In an average year, there are 5 trading days per week, 4.2 weeks per month and of course, 12 months per year. From there, we are going to subtract the 2 biggie holidays that nobody trades, being Christmas and New Years and assume that this will work the rest of the time. 141 trades/250 days = .564 = 1 trade every 2 or so days...provided that it doesn't go into a week long DD...

The last one is unprofitable because it hasnt got time to reverse, so i didnt included that one in the backtest, but definitely it didnt got a margin call, so it was proof of 1 year without a margin call.

Quoting Magix
Wait... Hold the presses... Something I may have forgot to mention. The above might seem a little negative, but that is actually, statistically, the Best Case Scenario. There is the actual possibility of holding on forever below your break even because of continued loss on swaps during long periods of DD.
Yes the swaps could f*ck us up, i havent calculated this, but i have no way of predicting what could happen ahead, so that is why a demo forward test is good.

Maybe if you want you can do the forward test , you already have 4-5 demo accounts on forward test so why not add 1 more there


__

Anyway, new version of MT4, time to check it out.
"There's a sucker born every minute" - P.T. Barnum
  • #357
  • Feb 4, 2014 1:18pm Feb 4, 2014 1:18pm
<p>
Quoting Proximus
{quote} The last one is unprofitable because it hasnt got time to reverse, so i didnt included that one in the backtest, but definitely it didnt got a margin call, so it was proof of 1 year without a margin call. <img src="http://assets.forexfactory.net/images/smilies/yim/happy.gif" border="0" alt="" title="Smile" smilieid="82" class="inlineimg" /> {quote} Yes the swaps could f*ck us up, i havent calculated this, but i have no way of predicting what could happen ahead, so that is why a demo forward test is good. Maybe if you want you can do the...
</p>

<p></p>
Bud your a star. Try listening to someone. Magix is doing his best to help and your just wasting part of your life. Be good and I hope you prove us all wrong
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #358
  • Feb 5, 2014 1:53am Feb 5, 2014 1:53am
Proximus,

Aproximator v1.1 sets the Monthly Max Low for EURUSD to 1.3553 (CET based broker) but the Day Mid is at 1.3515 so there's no possible trades for at least another day using your EA. Day Mid being below the Monthly 'Low' line is true for GBPUSD, EURJPY and USDJPY as well. EURGBP has the chance for a trade today but only if it goes up another 15 pips but given the small range I don't like it's chances with this system even if it does trigger a trade.
  • #359
  • Feb 5, 2014 1:50pm Feb 5, 2014 1:50pm
Quoting Proximus
Maybe if you want you can do the forward test , you already have 4-5 demo accounts on forward test so why not add 1 more there __
In the grandest scheme of likelihoods, this suggestion doesn't even show up in the shadows.

Having an idea of market movement and a base understanding of what you have devised as a system makes no relevant sense. Again, because your ideas of trading are green, to say the least, and because you stats are based on BACKTEST reports, I don't need to undertake this, for the sake of proving myself correct.

Again, even at optimized settings, with my fingers crossed and having four leaf clovers surrounding my pc, even if this was the inevitable outcome, I still don't like the results.

Last week, my Gambler made 30% and wiped back to 11% for week close and is again back up 30%.

This is my idea of a gamble. Consistent profits, with a relatively low DD and if strategically tested Monday - Wednesday only, if future results are the same, could possibly show 120% gain in a month.

But, I will leave it run, see how it continues to drive and see if effectively trace the market dynamically.
Attached Image (click to enlarge)
Click to Enlarge Name: 32.png Size: 38 KB
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #360
  • Feb 5, 2014 7:51pm Feb 5, 2014 7:51pm
Quoting fx_malloc
Proximus, Aproximator v1.1 sets the Monthly Max Low for EURUSD to 1.3553 (CET based broker) but the Day Mid is at 1.3515 so there's no possible trades for at least another day using your EA. Day Mid being below the Monthly 'Low' line is true for GBPUSD, EURJPY and USDJPY as well. EURGBP has the chance for a trade today but only if it goes up another 15 pips but given the small range I don't like it's chances with this system even if it does trigger a trade.
I thought thats obvious, but i guess its not.Its not wise to trade when the daily stuff is above or below the monthly high/low.

So we only trade when the daily mid is between MONTHLY & DAILY HIGH/LOW, and of course the price just crossed the DAILY MID from a direction.

This way we will be close to the monthly median price and have a very high chance of reversal at any moment.
"There's a sucker born every minute" - P.T. Barnum
0 traders viewing now