Three Bar Scalp Method w Probability | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Scalp, scalp, scalp 5,689 replies

LUNC/USDT - Scalp Scalp Baby! 2 replies

3 Bar Fractal with mid-bar and bar on each side 11 replies

To scalp or not to scalp? (Images included) 18 replies

Pin bar, pin bar, pin bar 0 replies

Attachments: Three Bar Scalp Method w Probability
Exit Attachments

Three Bar Scalp Method w Probability

  • #61
  • Sep 18, 2007 2:42am Sep 18, 2007 2:42am
There is also a choice of filter. If someone does not like the RSI filter one can use an EMA say 18 periods and if the price opens above the EMA go long if price opens below EMA go short. But look the previous close should be above/below o fthe one before.

My granny undesrtands that
.
  • #62
  • Sep 18, 2007 5:29am Sep 18, 2007 5:29am
Quoting SynWind
i like this sytem thanks Bull rock for it.
reason
1) it tell you when to enter
2) it tell you when to exit and take profit
3) it tell you to put SL
4) it follow the trend... thus increase probalility on winining.
5) don't take up much time as it does only once a day.
6) most important it simple and easy for any more to follow.

GREAT SYSTEM!
Maybe twice if you want to reverse... If you have been chopped out on both ends you stop for the day. (If 1% each trade that is 2% for the day) Next day you might make 5% or more with this system. Or if you manage to stay in trend and you accumulate the lots along the line as per the rules... You can get even more.
.
  • #63
  • Sep 18, 2007 10:47am Sep 18, 2007 10:47am
Quoting bullrock
Method variation:

Pull to breakeven the SL when 25% of the ATR(14) is in gain.
Use 3 units.
Exit unit 1 when 50% af the ATR(14) is hit.
Exit unit 2 at the end of the day when in profit if not in profit leave until BE or SL is hit.
If signal is positive on the opening of the next candle leave unit 3 in play and move SL to breakeven and start with extra 3 units again.

Regards
Bullrock

I am reading this as saying bring 2 units to BE when 25% of the ATR is hit. Let the SL on third unit sit at original SL of 25% ATR until close of day. Then take it from there. Is that correct?

Chris
  • #64
  • Sep 18, 2007 11:09am Sep 18, 2007 11:09am
Quoting Cmartin371
I am reading this as saying bring 2 units to BE when 25% of the ATR is hit. Let the SL on third unit sit at original SL of 25% ATR until close of day. Then take it from there. Is that correct?

Chris
Yes, sort of. I have figured out that in case of the gu it would be around 40 pips. That is the optimal. It is reached 55%-60% of the cases. So there is the edge. You need to leave it to develop to there. I think we should take the first TP there and move the SL to be of the 2nd and 3rd one.
.
  • #65
  • Sep 18, 2007 11:13am Sep 18, 2007 11:13am
Quoting bullrock
Yes, sort of. I have figured out that in case of the gu it would be around 40 pips. That is the optimal. It is reached 55%-60% of the cases. So there is the edge. You need to leave it to develop to there. I think we should take the first TP there and move the SL to be of the 2nd and 3rd one.
Close the second if in profit (SL @ BE), and leave the third to the following day. Start over. Add 3 units.. So you will have 4 units in play. If this happens to work for 6 days. Then you will have 5 units from the previous days and 3 units from the 6th day... Imagine this cumulative cumulation... Say first unit 300 pips second 250 and so on. You could end up with over 1000 pips just using this simple approach.
.
  • #66
  • Sep 19, 2007 5:37pm Sep 19, 2007 5:37pm
Just wanted to say thanks for the work here. In the past few days I have done some backtesting of this using the original two bar above/below 50 on the RSI, using various MM ideas. Tesing one month at a time, I have yet to lose in a month. The best result was the most recently suggested one which I altered a bit, making the stop loss and first profit 50% of the ATR. 25% just seemed to close. Trading three lots, I would then close the second in profit at the close of the day, then re open three more the next day if it qualified, still carrying one from the original trade, moving it's stop up each day to the new stop loss until the stop was hit. Although there were some hefty losses during ranging markets, the trending more than made up for it, producing some terrific gains.

I am going to continue to tinker with this, and I look forward to hearing some more ideas.


Thanks
Richard
  • #67
  • Sep 20, 2007 1:36pm Sep 20, 2007 1:36pm
Quoting Richard7
Just wanted to say thanks for the work here. In the past few days I have done some backtesting of this using the original two bar above/below 50 on the RSI, using various MM ideas. Tesing one month at a time, I have yet to lose in a month. The best result was the most recently suggested one which I altered a bit, making the stop loss and first profit 50% of the ATR. 25% just seemed to close. Trading three lots, I would then close the second in profit at the close of the day, then re open three more the next day if it qualified, still carrying one from the original trade, moving it's stop up each day to the new stop loss until the stop was hit. Although there were some hefty losses during ranging markets, the trending more than made up for it, producing some terrific gains.

I am going to continue to tinker with this, and I look forward to hearing some more ideas.


Thanks
Richard
Great! Have fun.. Yes trends sometimes can give you multithousand pips (cumulatively)

Right yes.. My findings that for the GU the optimal would be around 40-50 pips.. The SL. Regards Bullrock If you have any ideas, questions, etc just keep posting. I am also working on various variations of these.
.
  • #68
  • Sep 20, 2007 1:38pm Sep 20, 2007 1:38pm
Richards
One more thing. Try to move the SL to BE when your TP has been hit. It will help a lot.

Regards
bullrock

Quoting Richard7
Just wanted to say thanks for the work here. In the past few days I have done some backtesting of this using the original two bar above/below 50 on the RSI, using various MM ideas. Tesing one month at a time, I have yet to lose in a month. The best result was the most recently suggested one which I altered a bit, making the stop loss and first profit 50% of the ATR. 25% just seemed to close. Trading three lots, I would then close the second in profit at the close of the day, then re open three more the next day if it qualified, still carrying one from the original trade, moving it's stop up each day to the new stop loss until the stop was hit. Although there were some hefty losses during ranging markets, the trending more than made up for it, producing some terrific gains.

I am going to continue to tinker with this, and I look forward to hearing some more ideas.



Thanks
Richard
.
  • #69
  • Sep 20, 2007 2:00pm Sep 20, 2007 2:00pm
Quoting bullrock
Richards
One more thing. Try to move the SL to BE when your TP has been hit. It will help a lot.

Regards
bullrock

I caught the USD/CAD signal last night. I am not at home but I think I am well over 100 pips on the final two lots so far. I used IBFX so my brokers day ends at 8pm est. I was stopped out on all other signals I have tried so far using 25% of ATR. With good MM this is real good stuff!
  • #70
  • Sep 20, 2007 6:16pm Sep 20, 2007 6:16pm
Very interesting backtest results above.

I'm interested in trying this system, but being new to forex, I don't know how. I've opened a demo account with MT4, and I know how to place trades. But I've been visually looking at the charts to determine the position of bars. Is there some way to do this "mechanically?" Which I suppose there is, since the system can be backtested via a computer? Or was it visually backtested, if there is such a thing?

Thanks in advance,
Marcy
  • #71
  • Edited 9:32am Sep 21, 2007 7:12am | Edited 9:32am
Hi, Is there any statistic reasons when you set up the SL to be 25% of the ATR(14)? If you know the winning probability is higher than 50%, is it more sensible to set up a larger SL? Particularly, if statistically there are more winning pips than losing pips, then perhaps no point setting up a SL at all. I am afraid that the positions may be closed by ST before the profits happen. If yes, it will reduce the winning probability.

Sorry for lots of questions and many thanks in advance!


Quoting bullrock
I've cooked up quickly a system. I've got the ideas from various places on the forum and then performed some calculations and certainly looks promising.
Edit:
  • #72
  • Sep 23, 2007 7:03am Sep 23, 2007 7:03am
Quoting dogberry
Hi, I am not sure if the RSI(8) you were using is based on bar 3 or bar 2 when you calculated the winning probability of bar 3. In practice we can not know bar 3's RSI(8) until the end of the day.

When I tested your system on both GBP/USD and EUR/USD over the last 8 years, based on bar 2's RSI(8). Both are very Negative, unless I use bar 3's RSI(8), which is very profitable, but unfortunately, as I mentioned, is practically impossible. So I am not sure if you made a mistake here.
Nope. I haven't. Just the description has problems To get serious... The optimal SL/TP combination for GU are 40 pips. the next TP could be anywhere you like. Say 100 pips. But make sure when your 40 TP has been hit move the SL to BE. There is approximately, in case of the GU, 60% chance that the TP will be hit and 54% chance that the SL will be hit.. But there is only 43% chance that say 80 or 100 TP will be hit so you MUST move the SL to BE because of the probability is below 50%. Now the third unit should be left till the end of the day. In case the GU wants to rocket to 300 pips that day.

Have fun, jope this helps.

PS we do not care intraday shit here like MACD, S/R, pivots and all that stuff. That just compluicates the picture unncessarily..Try out you will see.
.
  • #73
  • Sep 23, 2007 7:04am Sep 23, 2007 7:04am
Quoting dogberry
Hi, I am not sure if the RSI(8) you were using is based on bar 3 or bar 2 when you calculated the winning probability of bar 3. In practice we can not know bar 3's RSI(8) until the end of the day.

When I tested your system on both GBP/USD and EUR/USD over the last 8 years, based on bar 2's RSI(8). Both are very Negative, unless I use bar 3's RSI(8), which is very profitable, but unfortunately, as I mentioned, is practically impossible. So I am not sure if you made a mistake here.
Leave the ATR for now.
.
  • #74
  • Sep 23, 2007 10:40am Sep 23, 2007 10:40am
Taking September for example, I did some calculation. The probability of both the 3rd bar (trade day) closed higher or lower is 50% for the Gu.
Suppose we do not use any SL or TP, the profit/loss are based on the difference between close and open. The results are as follows,
Sept 6,7,10,12 buy, profit 32, 54, -5, -31 pips;
Sept 14 17 18 20 sell, profit -117, -119, 193, 92 pips.
Net profit 1 pip. Probability 50% / 50%.
I also calculated the probability over the last 8 years. Net profit is even poorer (-3000 pips without considering spread), with probability the same, 50% / 50%.
Where is my mistake?

[/quote]
Quoting bullrock
The probability that the 3rd bar (trade day) will close higher is 62.77% for the Gu and that the 3rd bar will close lower is 60.53% viewing the last 7 years.
Edit:
  • #75
  • Sep 23, 2007 4:55pm Sep 23, 2007 4:55pm
Quoting dogberry
Taking September for example, I did some calculation. The probability of both the 3rd bar (trade day) closed higher or lower is 50% for the Gu.
Suppose we do not use any SL or TP, the profit/loss are based on the difference between close and open. The results are as follows,
Sept 6,7,10,12 buy, profit 32, 54, -5, -31 pips;
Sept 14 17 18 20 sell, profit -117, -119, 193, 92 pips.
Net profit 1 pip. Probability 50% / 50%.
I also calculated the probability over the last 8 years. Net profit is even poorer (-3000 pips without considering spread), with probability the same, 50% / 50%.
Where is my mistake?
[/quote]I'll have a look ASAP.
.
  • #76
  • Sep 23, 2007 5:30pm Sep 23, 2007 5:30pm
Quoting dogberry
Taking September for example, I did some calculation. The probability of both the 3rd bar (trade day) closed higher or lower is 50% for the Gu.
Suppose we do not use any SL or TP, the profit/loss are based on the difference between close and open. The results are as follows,
Sept 6,7,10,12 buy, profit 32, 54, -5, -31 pips;
Sept 14 17 18 20 sell, profit -117, -119, 193, 92 pips.
Net profit 1 pip. Probability 50% / 50%.
I also calculated the probability over the last 8 years. Net profit is even poorer (-3000 pips without considering spread), with probability the same, 50% / 50%.
Where is my mistake?
[/quote]
I abviously made a mistake in my calculations. There was a bug in one of the formulas, that is I compared wrong things. I have included the Excel sheet this time. An extract of my originals and corrected. Obviously 60% looked to goog to be true. Damn.

Scroll to the bottom to see the results.

For long
53.35%/46.65% success/failure
For short
48.02/51.98 success/failure

Regards
Attached File(s)
File Type: zip GUProbabilities.zip   497 KB | 632 downloads
.
  • #77
  • Sep 23, 2007 6:58pm Sep 23, 2007 6:58pm
nice system bullrock, and i'm going to start demoing it as it fits in with my creteria of keeping things simple! one question though, for those using IBFX, isn't ATR invalid because it includes the sunday bar in its calculation?
  • #78
  • Sep 23, 2007 9:43pm Sep 23, 2007 9:43pm
Quoting Trumpcard
nice system bullrock, and i'm going to start demoing it as it fits in with my creteria of keeping things simple! one question though, for those using IBFX, isn't ATR invalid because it includes the sunday bar in its calculation?
Use a spreadsheet and calculate yourself. I think Metatrader uses a different formula.

Look here> http://stockcharts.com/school/doku.p...true_range_atr

It is very easy to calculate yourself. You can use the spreadsheet in my previous post.
.
  • #79
  • Sep 23, 2007 9:47pm Sep 23, 2007 9:47pm
Quoting Marcy
Very interesting backtest results above.

I'm interested in trying this system, but being new to forex, I don't know how. I've opened a demo account with MT4, and I know how to place trades. But I've been visually looking at the charts to determine the position of bars. Is there some way to do this "mechanically?" Which I suppose there is, since the system can be backtested via a computer? Or was it visually backtested, if there is such a thing?

Thanks in advance,
Marcy
Dont be lazy man. This system requires 10 minutes from you time anyway. My backtest was a plain Excel sheet.. hehe I tell you Forex can be simple but that simplicity will come to you through hard work.. Have fun. DONT BET THE FARM!!
.
  • #80
  • Sep 25, 2007 10:52am Sep 25, 2007 10:52am
nice and simple
bj
BJ
0 traders viewing now