Statistical Gambling System (High Winrate, but NO EDGE) | Page 17 | 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)

  • #321
  • Jan 29, 2014 12:43pm Jan 29, 2014 12:43pm
Quoting Magix
Houston... We have a problem! The EA has seemingly taken 2 identical positions, same lot size and same price. Though it is a good position, this will skew the overall results. Weird, I have checked through everything and cannot understand why this additional position was entered.... DAMMIT! {image}
Use a time filter, to not enter twice in a position at the same bar.It is very annoying and its 2x risk, of course the reward is also twice, but its better to diversify than just risk 2x more on the same trade, its not rational.
"There's a sucker born every minute" - P.T. Barnum
  • #322
  • Jan 29, 2014 12:55pm Jan 29, 2014 12:55pm
Quoting Proximus
{quote} Use a time filter, to not enter twice in a position at the same bar.It is very annoying and its 2x risk, of course the reward is also twice, but its better to diversify than just risk 2x more on the same trade, its not rational.
Problem is...

The order was sent 2x and processed 2x.

The grid was not respected, there is no reason for me to believe that a time filter would have been either.

This is the same basic core code I have been working with for over a year now and have actually never seen this happen.

This would lead me to believe that it is a broker issue, more than issue with code or trigger.

If order was originally rejected and resent then accepted both times, filling at the exact same time and rate is a possibility...

Really the only thing I could think of that makes sense.

Additionally, I am not typically interested in time as I am in the orders to fill at specific trigger and with the Trigger on Gridded step.

But thanks for the idea on time filter.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #323
  • Jan 29, 2014 2:37pm Jan 29, 2014 2:37pm
Quoting Magix
{quote} Problem is... The order was sent 2x and processed 2x. The grid was not respected, there is no reason for me to believe that a time filter would have been either. This is the same basic core code I have been working with for over a year now and have actually never seen this happen. This would lead me to believe that it is a broker issue, more than issue with code or trigger. If order was originally rejected and resent then accepted both times, filling at the exact same time and rate is a possibility... Really the only thing I could think...
Oh, then its a different issue.I thought it was a common error that newbie EA programmers do, which is that they dont put a filter to not place 2 identical orders at the same bar/time.

If it was a broker issue, you should contact your broker and ask them what happened.Maybe it was a server/lagg issue that caused it to trigger it twice.
"There's a sucker born every minute" - P.T. Barnum
  • #324
  • Jan 29, 2014 2:49pm Jan 29, 2014 2:49pm
Quoting Proximus
{quote} Oh, then its a different issue.I thought it was a common error that newbie EA programmers do, which is that they dont put a filter to not place 2 identical orders at the same bar/time. If it was a broker issue, you should contact your broker and ask them what happened.Maybe it was a server/lagg issue that caused it to trigger it twice.
If it were real money...I'd actually be on the horn with tech support hitting them with my best WTF?!?!?

But, as this is a Demo account, hard to say where the server is, what lag or latency issue might have been present and I wouldn't expect that I would ever really get this solved by some geek sitting in a chair, making minimum wage to be help desk.

I think that there are still modifications I should be making to this, especially on the profit side. I don't like it holding onto profit for too long as this can easily turn to loss, so I think I have some work to do on the exits.

I am still infinitely happier with the Dynamic Range, given the weak ass market week, there has still been enough range to churn 300 pips on EURUSD, and this with it still missing trades.

Just a trial run. At least I know there's a little hope for this one.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #325
  • Jan 29, 2014 6:46pm Jan 29, 2014 6:46pm
And yet...

It is hard not to be satisfied with an EA that even in a flat market can detect Overbought and take exit...

The whipsaw earlier, the EA didn't see this as being the top, and held out for more. When more wasn't on the table and the throttled back trigger was in effect...SHAZAM!

Took the money and ran!

Pretty fucking kewl, actually...

Sure, might be patting myself on the back...but seriously...very nice.

This is an M5 chart, just to give enough room to show trades on the week. The actual trader works off of the M1.
Attached Image (click to enlarge)
Click to Enlarge Name: os.png Size: 38 KB
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #326
  • Jan 29, 2014 7:05pm Jan 29, 2014 7:05pm
Stats through the Myfxbook Optics.
Attached Image (click to enlarge)
Click to Enlarge Name: mfxb.png Size: 40 KB
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #327
  • Jan 29, 2014 7:26pm Jan 29, 2014 7:26pm
Looking real good, Magix. You could be on to something, here.

Rob
  • #328
  • Jan 29, 2014 7:43pm Jan 29, 2014 7:43pm
Quoting ChicagoRob
Looking real good, Magix. You could be on to something, here. Rob
I think we both have a good idea how this is going to turn out...

But all the same, it is fun to watch.

Nice to see ya in print again, Rob.

It's been a while.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #329
  • Jan 30, 2014 5:01am Jan 30, 2014 5:01am
Quoting Magix
{quote} If it were real money...I'd actually be on the horn with tech support hitting them with my best WTF?!?!? But, as this is a Demo account, hard to say where the server is, what lag or latency issue might have been present and I wouldn't expect that I would ever really get this solved by some geek sitting in a chair, making minimum wage to be help desk.
Demo accounts usually dont lagg much, its the big lagg on real accounts and the requotes that could mess up the strategy.It could be a more serious issue maybe with your code.Even if you use the code for years, that doesnt mean there was not a hidden bug inside it.You might wanna fix that before going live with it.

EDIT: Looks like my losing position is going to turn around, after 2 weeks ,well anyway lesson learned, never going to buy or sell again near monthly or weekly low or high.
"There's a sucker born every minute" - P.T. Barnum
  • #330
  • Edited 7:25am Jan 30, 2014 7:05am | Edited 7:25am
New version of the system, i improved it in the last 2 weeks using your ideas and suggestions guys, now its way better and we wont make the same mistakes twice

UPDATE: VERSION 1.5 (NEVER MAKE THE SAME MISTAKES TWICE)

  1. We will use now my custom daily pivot point indicator called APROXIMATOR 1.0 , which is specially modified for this purpose, this way we wont trade at daily-weekly lows or highs, we will always stay in the middle
  2. I will try this new version of the system as soon as my losing position gets closed
  3. Stay away from news or high volatility situations as usual
  4. I replaced the Moving Average with the DAILY_MID pivot point, because the daily mid pivot even though its not dynamic , it shows better and more accurately the center-mid price where the daily price movement tends to fluctuate towards.The DAILY_MID pivot point is a much more realistic daily mid price than a daily moving average for example.

Attached Image (click to enlarge)
Click to Enlarge Name: eurusdm1.png Size: 22 KB


Pair: EUR/USD
Timeframe: M1 (1 Minute chart)
Indicators used: APROXIMATOR 1.0 (Custom daily pivot point indicator made by me)
BUY ENTRY: The price must be between DAILY_MEDIUM_HIGH and DAILY_MEDIUM_LOW, we enter BUY if the price 1 candle ago was below or equal the DAILY_MID, but now ,at the current candle,its above it
SELL ENTRY: The price must be between DAILY_MEDIUM_HIGH and DAILY_MEDIUM_LOW, we enter SELL if the price 1 candle ago was above or equal the DAILY_MID, but now ,at the current candle,its below it
TAKEPROFIT: 35-40 pipettes on 5 digit broker or 3-4 pips on 4 digit broker
STOPLOSS: None, yet (still experimenting with this new system)
LOT SIZE: we will use now 1% of the account balance
LEVERAGE: 1:1 is the best, the more the leverage the higher chance of the margin call.

LOT SIZE (in MT4) = ACCOUNTBALANCEINCASH * RISKPERCENT/10000

Only 1 order open at a time, dont move TP, and dont close orders manually, wait for rebounce to your TP.Also dont forget to withdraw the profits frequently in case the margin call will wipe out the account.

Here is the new template & indicator:

Attached File(s)
File Type: mq4 APROXIMATOR V1.0.mq4   5 KB | 3,462 downloads

Attached File(s)
File Type: tpl gambling system 1.5.tpl   2 KB | 727 downloads
"There's a sucker born every minute" - P.T. Barnum
  • #331
  • Jan 30, 2014 7:29am Jan 30, 2014 7:29am
Here is a little more explanation for those who are a little bit confused.

Attached Image (click to enlarge)
Click to Enlarge Name: ggg.png Size: 33 KB


You can clearly see that today the EUR/USD on was only tradeable between 0:00 gmt+0 and 7:00 gmt+0, because only then the price was between the DAILY_MEDIUM_HIGH and DAILY_MEDIUM_LOW, right after 7:00 when the London session opened, the price got down below the DAILY_MEDIUM_LOW and its still lower than that, so its not tradeable at the moment.
"There's a sucker born every minute" - P.T. Barnum
  • #332
  • Jan 30, 2014 7:55am Jan 30, 2014 7:55am
So for example,according to V1.5 rules, today our last trade should have been at 7:07 GMT+0, because the price was above the DAILY_MID before that, but it just pierced trough and went down at 7:07 so according to the 1.5 rules, we should have entered our last SELL trade at 7:07, and we have won that.

Attached Image (click to enlarge)
Click to Enlarge Name: eurusdm1333.png Size: 24 KB


Of course, before that 0:00 and 7:07 we would have had 7 hours to trade the sh*t out of the market, and all of our trades for today would have been a winner.

I`m very excited to present this new system, unless there is an unexpected major news with high volatility or something like that, its probably a 99.9999% winrate system
"There's a sucker born every minute" - P.T. Barnum
  • #333
  • Jan 30, 2014 8:19am Jan 30, 2014 8:19am
now this looks a lot better,lines in the sand are clear and defined,like it .good luck
  • #334
  • Jan 30, 2014 8:24am Jan 30, 2014 8:24am
Hi Proxy, hope your good. Keep the trade life alive. I only pop in now and again.
I am half busy developing trading strategies (or one) I was busted earlier yet again but realised it was my fault as I set up the scanner on pc wrong, Be good
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #335
  • Jan 30, 2014 9:32am Jan 30, 2014 9:32am
Quoting diceman555
now this looks a lot better,lines in the sand are clear and defined,like it .good luck
Quoting billytt
Hi Proxy, hope your good. Keep the trade life alive. I only pop in now and again. I am half busy developing trading strategies (or one) I was busted earlier yet again but realised it was my fault as I set up the scanner on pc wrong, Be good
Thanks, its not the best yet, but its still considerably better than the previous versions where we didnt take into consideration the daily low-highs.

I`m still tweaking the system, and at the moment i`m backtesting it.
"There's a sucker born every minute" - P.T. Barnum
  • #336
  • Jan 30, 2014 2:57pm Jan 30, 2014 2:57pm
Wow Magix WTF is this?

Attached Image (click to enlarge)
Click to Enlarge Name: MAG.png Size: 74 KB


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 ?
"There's a sucker born every minute" - P.T. Barnum
  • #337
  • Jan 30, 2014 6:28pm Jan 30, 2014 6:28pm
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 ?
Hi Proxy, good to see your newer method is holding up.
Regarding the suspension of Pinocchio, the guy isn't here to defend himself so I would advise you in a friendly fatherly way to man up and don't mock him.
Be good bro
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #338
  • Jan 30, 2014 6:59pm Jan 30, 2014 6:59pm
M, thanks for the mail. Hope things plan out for you, Funny thing is I am now getting interested in the gambling side of forex .I'll probably end up on the streets , started there end there be good n take care
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #339
  • Jan 30, 2014 7:49pm Jan 30, 2014 7:49pm
Glad to see some science on this forum, just got my master degree @economics.

@Proximus: Have u consider using a diffrent approach for the distribution, like multiple density. Dig around for markov switching, it's been proven to be better than ma. I still have to submit revision for my thesis, and required to write a journal for publication so I still won't touch fx for a while.

MSM mainly used in central banks, but I think it is viable to use in fx. There was already a works in fx, and it was quite good. I have the journal on my pc -I'll share it later. MSM use internal dummy to separate regime(i.e bullish and bearish) and have different parameter for each regime. The dummy formed by a markov process depends on previous period state, switching from a regime to another by calculating of transition probabilities.

In my works it was consistent that switching happened after there was unusually high volatility observed. Growth/bull period is when there is low variance/volatility, contraction/bear happened with high variance/volatility. Hope this would be useful for u :-) .
  • #340
  • Jan 31, 2014 6:01am Jan 31, 2014 6:01am
Quoting Passerby
Glad to see some science on this forum, just got my master degree @economics. @Proximus: Have u consider using a diffrent approach for the distribution, like multiple density. Dig around for markov switching, it's been proven to be better than ma. I still have to submit revision for my thesis, and required to write a journal for publication so I still won't touch fx for a while. MSM mainly used in central banks, but I think it is viable to use in fx. There was already a works in fx, and it was quite good. I have the journal on my pc -I'll share...
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.

________

But until i research, i would be happy if anyone else would try my system on a demo account, and make a TE so i can link that into this thread.It would be a great comparison chance on different brokers.

So if anyone wants to try my 1.5 version of the system on a DEMO forward test, post your TE here, and i will link it to the thread
"There's a sucker born every minute" - P.T. Barnum
1 trader viewing now