Is forex education misleading or everyone is stupid? | Page 27 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Forex Education 78 replies

forex education 24 replies

Education IRA (forex)? 2 replies

Just finished forex education, what should be my next path? 12 replies

Stocks vs Forex - Is everyone suited for forex? 18 replies

Attachments: Is forex education misleading or everyone is stupid?
Exit Attachments

Is forex education misleading or everyone is stupid?

  • #521
  • Jan 27, 2022 10:08am Jan 27, 2022 10:08am
Quoting ElmPie
{quote} Exactly it is decentralized so I'm not sure how you define 'when there is a lot of offers' since you can only see the offers on your broker.. Trying to think about what you mean with the lot size issue but not getting it.. yet.

Quotes tell us when the last trade was done.

Meaning if there was a trade, there must be an offer first right?

So, how close the quotes move?

Do they move smoothly or quickly but with a direction?
Offers are close together, there's liquidity then or its enough


If they move in a sudden manner means offers are far apart. Low liquidity.

This is how one can tell
  • #522
  • Jan 27, 2022 11:17am Jan 27, 2022 11:17am
Quoting captalTrader
{quote} Interestingly good quality ticks that are quantifiable accurately start coming in at 4 then fade off by 6 then start again 8 until 16. GMT +2. These calculations don't know it's now Asian, London or New York open. That's just how lovely mathematics is, so lovely See how all the trades have almost all been done during London and New York times? It's lovely honestly. There's nothing so satisfying than witnessing equations in real life. It's...
Yes every entry is random, there's nowhere in the code that say's trade at Asian open. If anytime there's enough incoming ticks to calculate.... And if probabilities are as defined, take a trade!

Uhhhh, no, that is not correct. You have a defined method for making an entry that is not based on making a random selection. You have stated 2 criteria:
1. a specific number of ticks (of some quality) must arrive;
2. the current probabilities must match defined probabilities.

The fact that it is unknown and undeterminable exactly when enough ticks will arrive does not mean that this time is a random selection. I suspect that there is actually a probability distribution over whatever time period is selected, such as the 24 hours of a trading day, that is not a random distribution.
To free Gazans of Hamas, use whatever it takes.
  • #523
  • Jan 27, 2022 11:31am Jan 27, 2022 11:31am
Quoting captalTrader
Someone on mql5 suggested we alter the distance between take profit and SL aka pips so to compensate the inaccuracies. At first I thought it might work but realised late that the lot size calculations are dependent to the same distance changing the distance will then have different lot calculations. With this problem, the EA is not profitable at all because the real edge of its logic it the ability to achieve 51% accuracy. For now it's flying there close to 60% which is just alreastic and extremely lucky but once reality sets...
Option 2. here would solve your issue. As the error in profits/losses will be marginal. It seems your EA logic is simply configured for well capitalized accounts.
  • #524
  • Edited 12:34pm Jan 27, 2022 12:14pm | Edited 12:34pm
Quoting captalTrader
{quote} Hhhha sir! How the code will be different than the free copy of the EA that cannot be re sold For now we have a very serious problem that you can try and help with. That is accurate lot size calculations. You see, say the algorithm calculations say 0.001224 lots to take this trade in order for the risk to be 5% or whatever preferable. Problem is minimum lot size is 0.01 and 0.01 will not be so accurate anymore. This one is somehow easy cause we can just skip this trade altogether....
say lot required is 0.213455,

You gave 6 significant figures here. Do you require that much precision? Have you determined what the minimum number of significant figures you need to have your result meet you minimum requirement?

After you determine the number of significant figures required, you can then determine what the minimum lot size is that can "fit" your chosen risk, i.e. the digits that you truncate aren't of value at that lot size. If your example using 6 signficiant figures is typical (maybe it is, because you claim that only 2 significant digits would introduce huge error in the long run), then you will need a much larger lot size than 0.2. And I suspect that would make your SL unrealistically small. So that means either a larger account balance is needed or you simply must skip a lot of potential entries
To free Gazans of Hamas, use whatever it takes.
  • #525
  • Jan 27, 2022 12:20pm Jan 27, 2022 12:20pm
Quoting DrDave
{quote} Yes every entry is random, there's nowhere in the code that say's trade at Asian open. If anytime there's enough incoming ticks to calculate.... And if probabilities are as defined, take a trade! Uhhhh, no, that is not correct. You have a defined method for making an entry that is not based on making a random selection. You have stated 2 criteria: 1. a specific number of ticks (of some quality) must arrive; 2. the current probabilities must match defined probabilities. The fact that it is unknown and undeterminable exactly when enough ticks...

Ok I understand and you're correct
  • #526
  • Jan 27, 2022 12:28pm Jan 27, 2022 12:28pm
Quoting sinqua
{quote} Option 2. here would solve your issue. As the error in profits/losses will be marginal. It seems your EA logic is simply configured for well capitalized accounts.
I figured, maybe with 1000 and more this will not affect the performance that much but still it would be great to solve a problem altogether than hoping for the best I know in time I'll find a solution
  • #527
  • Jan 27, 2022 12:56pm Jan 27, 2022 12:56pm
Quoting DrDave
{quote} say lot required is 0.213455, You gave 6 significant figures here. Do you require that much precision? Have you determined what the minimum number of significant figures you need to have your result meet you minimum requirement? After you determine the number of significant figures required, you can then determine what the minimum lot size is that can "fit" your chosen risk, i.e. the digits that you truncate aren't of value at that lot size. If your example using 6 signficiant figures is typical (maybe it is, because you claim that only...
Let's leave that abstract example.

Check this last trade.

It was supposed to be 21.184 to be 5% off 423.67 but this lot results was 20.75 even with positive slippages it's still not 21.184

Why is that, maybe I'm missing something and this can be explained
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot_2022_0127_204728.jpg Size: 53 KB
  • #528
  • Jan 27, 2022 2:31pm Jan 27, 2022 2:31pm
Quoting captalTrader
{quote} Let's leave that abstract example. Check this last trade. It was supposed to be 21.184 to be 5% off 423.67 but this lot results was 20.75 even with positive slippages it's still not 21.184 Why is that, maybe I'm missing something and this can be explained {image}
I don't see a problem. If your algo had come up with 0.22 lot, then the amount at risk would be more than the calculated 5% of 21.18. So 0.21 lot was the closest it could get without going over the calculated risk amount. As I already said, if you need better accuracy for matching the lot size to the calculated risk amount, then you have to scale up.
To free Gazans of Hamas, use whatever it takes.
  • #529
  • Jan 27, 2022 2:32pm Jan 27, 2022 2:32pm
Quoting captalTrader
{quote} Let's leave that abstract example. Check this last trade. It was supposed to be 21.184 to be 5% off 423.67 but this lot results was 20.75 even with positive slippages it's still not 21.184 Why is that, maybe I'm missing something and this can be explained {image}
Since the lot was 0.21 the 20.75 profit seems correct.

(115.362 - 115.248) / 0.001 = 114 (points)

(0.001/115.362) * 21000 = 0.182035679 (point value)

0.182035679 * 114 = 20.75206741 (profit)

With a 21.184 lot (if it would be possible) the profit would have been 20.93389504 .. so I guess one percent of the total amount difference
  • #530
  • Jan 27, 2022 2:37pm Jan 27, 2022 2:37pm
Quoting DrDave
{quote} I don't see a problem. If your algo had come up with 0.22 lot, then the amount at risk would be more than the calculated 5% of 21.18. So 0.21 lot was the closest it could get without going over the calculated risk amount. As I already said, if you need better accuracy for matching the lot size to the calculated risk amount, then you have to scale up.
I guess without knowing the algo it's difficult to understand why this would be a problem :/
  • #532
  • Jan 27, 2022 3:08pm Jan 27, 2022 3:08pm
Quoting Mpfreefx
Hidden
How is this supposed to add value to this discussion?
  • #533
  • Jan 27, 2022 3:15pm Jan 27, 2022 3:15pm
Attached Image (click to enlarge)
Click to Enlarge Name: 360_F_59067754_uuM6h2R1xAtJlqUJd7PDBRhxLz1KFkuz.jpg Size: 36 KB
  • #534
  • Jan 27, 2022 3:17pm Jan 27, 2022 3:17pm
Quoting elmpie
{quote} how is this supposed to add value to this discussion?
just an intersting trade i thought i would share. My apologies
"TIME AND PRICE IS ALL THERE IS"
  • #535
  • Jan 27, 2022 3:38pm Jan 27, 2022 3:38pm
Quoting Mpfreefx
{quote} just an intersting trade i thought i would share. My apologies
No clue what you are doing but it looks very nice!
  • #536
  • Jan 27, 2022 4:47pm Jan 27, 2022 4:47pm
Quoting ElmPie
{quote} No clue what you are doing but it looks very nice!
MUCH APPRECIATED
"TIME AND PRICE IS ALL THERE IS"
  • #537
  • Jan 27, 2022 9:00pm Jan 27, 2022 9:00pm
Quoting ElmPie
{quote} Since the lot was 0.21 the 20.75 profit seems correct. (115.362 - 115.248) / 0.001 = 114 (points) (0.001/115.362) * 21000 = 0.182035679 (point value) 0.182035679 * 114 = 20.75206741 (profit) With a 21.184 lot (if it would be possible) the profit would have been 20.93389504 .. so I guess one percent of the total amount difference
Perfect! Thank you
  • #538
  • Jan 27, 2022 9:03pm Jan 27, 2022 9:03pm
Quoting DrDave
{quote} I don't see a problem. If your algo had come up with 0.22 lot, then the amount at risk would be more than the calculated 5% of 21.18. So 0.21 lot was the closest it could get without going over the calculated risk amount. As I already said, if you need better accuracy for matching the lot size to the calculated risk amount, then you have to scale up.
  • #539
  • Jan 28, 2022 2:17am Jan 28, 2022 2:17am
Quoting DrDave
You gave 6 significant figures here. Do you require that much precision?
Yeah I love that "math" guys too. Let's complain that we have only 2 numbers after dot, but we will never calculate no spread no trading fee that can be up to 50%.
Just curious if his algo is so good why not trade on crypto. There are 8 number after dot on BTC pairs and 4 on USDT.
Observer effect
  • #540
  • Jan 28, 2022 5:07am Jan 28, 2022 5:07am
Quoting ryuryu
{quote} Yeah I love that "math" guys too. Let's complain that we have only 2 numbers after dot, but we will never calculate no spread no trading fee that can be up to 50%. Just curious if his algo is so good why not trade on crypto. There are 8 number after dot on BTC pairs and 4 on USDT.
My algo is not so good.

If it was, I wouldn't be here replying to negative throlls like you on a forum would I?
2 traders viewing now