Trading made simple with added BS | Page 8 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading Made Simple 151,868 replies

Trading Made Simple the right way, and other trading lessons 334 replies

Indicators for the 'Trading Made Simple' system 343 replies

MTF Indicators for the 'Trading Made Simple' system 241 replies

Need Simple Alert Added To This Indicator Please 3 replies

Attachments: Trading made simple with added BS
Exit Attachments

Trading made simple with added BS

  • #141
  • Dec 12, 2015 11:23am Dec 12, 2015 11:23am
If you think you can trade without sometimes being in draw I'd give up now tbh. But yes, the length of drawdown period in your example would be tricky mentally for sure but trading is tricky mentally.

For any edge the distribution of results is random. You need to understand that.

The casino will always win but it does not win always.

I hope that's a manual backtest in Forex Tester or something and not some automated POS?

In 175 days you ended 20% up without compounding and you think that's bad? That's good m8. You didn't blow it. 40%+ a year without compounding's bad? You're kidding me right? You could always stick it in a pension I guess and made about 0.5% at the moment.

Forget the 300% a year BS you see on here, it's fantasy.
  • #142
  • Dec 12, 2015 11:47am Dec 12, 2015 11:47am
Quoting nunrgguy
If you think you can trade without sometimes being in draw I'd give up now tbh. But yes, the length of drawdown period in your example would be tricky mentally for sure but trading is tricky mentally. For any edge the distribution of results is random. You need to understand that. The casino will always win but it does not win always. I hope that's a manual backtest in Forex Tester or something and not some automated POS? In 175 days you ended 20% up without compounding and you think that's bad? That's good m8. You didn't blow it. 40%+ a...

Yes, was a manual test using forex tester, not automated. Thanks for the positive response it gives me hope. I will work on refining this method. PS, I can see why you trade the 6 hour period, I think I may have to try a test on this also.

I think a great way to trade this also would to be if there was an mt4 indi that produced a rolling open/reference point for a fixed candle period. This way we could take advantage of more opportunities, because as you, cp and neio have mentioned, the opening is just a reference point, it is no more significant than any other line, however it does provide us with a measure of price movement.

Like you say the key here is taking advantage of current price movement,, removing the predictive mentality, then concentrating on trade management and over coming greed. Looking forward to more of your posts nunrgguy
  • #143
  • Dec 12, 2015 1:36pm Dec 12, 2015 1:36pm
  • #144
  • Dec 13, 2015 7:17pm Dec 13, 2015 7:17pm
Quoting nunrgguy
{quote} I did completely make the numbers up btw. OK. So given that edge. .....
Can I asked why you pick the %'s the way you do? With picking the bottom wick distribution that lands below x-tics 50% of the time, A Close > y-tics 40% of the time?

I see the edge is dependent upon the y-tics outnumbering the x-tics.

But is there any reason the percentages where chosen this way?
  • #145
  • Dec 14, 2015 7:03am Dec 14, 2015 7:03am
those numbers were just picked out of my @rse as an example. I keep saying this but you need to run a frequency distribution to get 'the' numbers. Then all should be clear, possibly.
  • #146
  • Dec 14, 2015 11:59am Dec 14, 2015 11:59am
Great discussion in this thread.

I understand the distances are just made up. I.e. 50 tics is just a number you threw out there. It could be for any period over the last number of bars that your distances could be 70 tics, 98 tics or some other number.

But what I mean is why the percentages are what they are? Or do those not even matter either as long as you trade those percentages? What if you pick the length that is less than x-tics 90% of the time for the bottom wick instead of 50%? Or does it matter at all?

I see how the distribution tells you what the distance is based on the percentage you choose. But my question is why a particular % matters and not another in trying to define an edge?

Does that make sense?

For example, for this week the GU, looking at the past 250 up days, my stats are telling me the average bottom wick is ~86 pips below(blue dots) the open. Orange dots are the open.

Also I just chose a divider of 50% as to the bottom wicks, and this number my stats tell me is ~80 pips below the open (other blue dots).
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-12-14_115002.jpg Size: 213 KB
  • #147
  • Dec 14, 2015 11:42pm Dec 14, 2015 11:42pm
Something interested for those following...

So I've been thinking about the question I asked earlier pretty hard.

Attached in the picture, you can see some stats I've compiled. Where it is the rolling percentage of upWeek where the Low stayed within a 50% range of the samples. You can see that it is pretty steady, and the rate is higher than I expected, higher than 50%. For giggles and somewhat related to the question I asked earlier, and ran the same stats on an 80% bottom wick on upWeek . Again the hit rate is higher than the threshold tested on.

To be clear, this isn't a percentage of all the samples taken at the end, it is checking on EACH INTERVAL, whether or not the given week's Low stayed within the threshold given @ the beginning of the week. And the percentage is a rolling of wether or not the Low on an upWeek is within that threshold or not.

This is weird to me for some reason?

Time to ponder a bit more...
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-12-14_233143.jpg Size: 551 KB
  • #148
  • Dec 16, 2015 12:55am Dec 16, 2015 12:55am
It wholly depends on how YOU interpret the data as to what you find. I'm not wholly certain what you're presenting as yet - only read your message once.

At the simplest level. If you run a distribution over, say 1000 bars you should see something.
Total bars 1000.
red bars
open to low
0-20 pips% (i.e. % out of 1000)
20-40pips%
40-60pips%
60-80%
etc
open to close%
0-20
20-40
etc
open to high
etc

green candles
same as above yeah?


For 1 you SHOULD see that if the CURRENT candle went red that if it then turns green and the low was less than x% there's a reasonable chance of it staying green.
Likewise if it went red, went further than y% pips, then crosses the open, there's a good change it may not go very far above the open.

You should also see that for green candles, open to high will be greater than open to low, you will also see the % likelihood and what pips ranges. i.e. as a made up example (don't expect this to add up, it's an illustration):
green candle
open to high
20-40pips 5%
40-60pip 8%
60-80 pip 10%
80-100 pip 15%
100-120 pip 20%
120-150 pip 25%

open to low
20-40 pip 25%
40-60 20%
60-80 15%
80-100 6%
100-120 3%
120-150 1%


So those are made up number, they don't add up, but by simply eyeballing them you should see something?????
Red will be almost the same as green but the opposite way around. Split between red and green will be around 50%.
  • #149
  • Dec 19, 2015 2:46pm Dec 19, 2015 2:46pm
I can see some aggregations that I was not calculating in the way described above. Like previously taking a lookback period of some number, and looking at either up or down bars independently and creating stats based on that. So my stats would have the exact number of up vs. down.

I've recently taken a period, and THEN taking whatever distribution of Up vs. Down bars occurred within that period, and aggregated my stats for Red and Green bars by that. I hope that explanation makes sense?

But I've gotten a basic idea working as described in the post above.

I'm not sure about the sanity of all of my earlier data though, so I'm only using 775 bars in my calculation instead of 1000. So each calculation at the start of each weekly period is based on the prior 775 weeks of data. I will have to work on getting some more data.
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-12-19_143926.jpg Size: 941 KB
  • #150
  • Dec 20, 2015 5:03pm Dec 20, 2015 5:03pm
looking good
  • #151
  • Dec 20, 2015 6:36pm Dec 20, 2015 6:36pm
Thanks!
  • #152
  • Dec 21, 2015 1:12am Dec 21, 2015 1:12am
so you can see that you've roughly 50:50 red to green, slight edge to green.
Now, what do you notice re distributions of upper/lower wicks???
  • #153
  • Dec 21, 2015 2:21pm Dec 21, 2015 2:21pm
Possibly 2 things? I doubt either are the point your getting at.

Maybe the first being that the wicks are naturally the smallest entities on a bar. Shown how most of the samples wind up being in the lower tick range of the distributions. But this is intuitive as none of the weekly bars end up being dojis. And price typically moves and stays somewhere away from the Open, during that period. As the retracements are smaller than the body that forms in the direction of the Close.

The second thing being that the wick ranges wind up being similar in occurrence? I.e. that most of the green bars have wicks on either side that appear to be about the same length. You can see this by looking how the ranges have similar distributions of occurrences within them. Although the green bars seem to be a bit more skewed than the red bars.

I will think about this a bit more as I doubt the things mentioned above are that significant of an observation.
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-12-21_132751.jpg Size: 715 KB
  • #154
  • Dec 21, 2015 4:22pm Dec 21, 2015 4:22pm
and also the opposing range to wick. I.e for a green bar where are the vast majority of lower wicks (in terms of range): what range does the biggest % fall (60 pips or lower or thereabouts) how does this compare to range of open to high, where does the majority lie in this case?

So there is an x% chance of a lower wick of n pips or less as compared to a y% chance of m pips or more.
  • #155
  • Edited 12:24pm Dec 22, 2015 12:10pm | Edited 12:24pm
Quoting nunrgguy
...what range does the biggest % fall (60 pips or lower or thereabouts) ... So there is an x% chance of a lower wick of n pips or less as compared to a y% chance of m pips or more.
I'm not sure I understand the post? But it seems you could slice it up a lot of different ways to get the percentages you want?

For example, using the 50% OL and 80% OH used as an example earlier in the thread... for green bars my stats say that there is a 50% chance of a lower wick of 80 pips or less, as compared to a 80% chance of 121 pips or more... WHEN THE BAR HAPPENS TO BE GREEN.

If I want the OL to be 60 pips or less(as in your above post), based on my stats x = 40%. So 40% of the green bars have OL wicks that are 60 pips or less.

I might be reading this wrong, but what confuses me is that those numbers are ONLY for weeks where the bar happens to turn out GREEN? If I use the GREEN and RED bars, where the OL is less than 60 pips... 23.4% OF ALL bars have an OL of 60 pips or less.

Apologies to the thread as I might be posting nonsense @ this point.
  • #156
  • Dec 23, 2015 1:26am Dec 23, 2015 1:26am
it has nothing to do with what you WANT. All you've got to do is look at the pip ranges and see where the majority lie. What pipip amounts above and below can you see where the ratio upper to lower is 2:1, 3:1. Forget about probability. There are no probabilities. Just look at the stats. Forget any earlier bullshit number I posted as an example, they're irrelevant.
  • #157
  • Feb 27, 2016 12:39pm Feb 27, 2016 12:39pm
Had some time to work more on this 1 candle thing..

I've got a setup at the moment that is completely mechanical, requires no discretion... Only tested through 2012 with 134 samples, but so far but looks good. I will continue to test it but I'm wondering if this is too good to be true considering it could be easily coded...

Attached Image (click to enlarge)
Click to Enlarge Name: obKgdiX.png Size: 22 KB
  • #158
  • Feb 27, 2016 12:58pm Feb 27, 2016 12:58pm
Quoting NaughtyPip
Had some time to work more on this 1 candle thing.. I've got a setup at the moment that is completely mechanical, requires no discretion... Only tested through 2012 with 134 samples, but so far but looks good. I will continue to test it but I'm wondering if this is too good to be true considering it could be easily coded... {image}
Sounds interesting.... Please share more about what you are doing.
Live today like theres no tomorrow-- Prioritize like you will die today
  • #159
  • Feb 27, 2016 1:29pm Feb 27, 2016 1:29pm
Quoting iamsovereign
{quote} Sounds interesting.... Please share more about what you are doing.
There's not much to say really. Nunrgguy has said it all. I'd love to spell it out but it wouldn't be fair on the people who have put the time and effort in to trying to crack this... Besides, my sample size isn't big enough yet to be certain that this is a winning system. I need over a 1000 really and with forward testing...
  • #160
  • Feb 28, 2016 9:43am Feb 28, 2016 9:43am
Quoting NaughtyPip
Had some time to work more on this 1 candle thing.. I've got a setup at the moment that is completely mechanical, requires no discretion... Only tested through 2012 with 134 samples, but so far but looks good. I will continue to test it but I'm wondering if this is too good to be true considering it could be easily coded... {image}
350 samples...

As you can see its not immune to drawdown

think I need to go really far back and test this...

Attached Image (click to enlarge)
Click to Enlarge Name: SkvOYXW.png Size: 41 KB
0 traders viewing now