wot u think o these then?

darktone

Veteren member
Messages
4,019
Likes
1,086
Am interested to see what you guys can glean from these trades.

What i would like to know is:-
1) How you think the trader did in each example
2) What was the traders size
3) What was the method iyo
4) What was the objective
EDIT:-
5) what was the total risk the trader was at

And anything else you feel is relevant


example 1
171500d1392113229-wot-u-think-o-these-then-thurs-fri.png


example 2
171502d1392113229-wot-u-think-o-these-then-mon.png
 

Attachments

  • thurs-fri.png
    thurs-fri.png
    117.7 KB · Views: 2,737
  • mon.png
    mon.png
    101.2 KB · Views: 2,718
Last edited:
Am interested to see what you guys can glean from these trades.

What i would like to know is:-
1) How you think the trader did in each example
2) What was the traders size
3) What was the method iyo
4) What was the objective

And anything else you feel is relevant

example 1
171500d1392113229-wot-u-think-o-these-then-thurs-fri.png


example 2
171502d1392113229-wot-u-think-o-these-then-mon.png

No idea what's going on there.
and No I wouldn't run this EA.
 
DT:
1. Overall the trader lost money (see the P/L in the statement)
2. The size is shown in the statements (0.2 to 0.5)
3. & 4. Not really sure what you're asking re: method & objective...

I'll have a closer look in a bit...
Cheers for stepping up DJ
Re 3 & 4, Purely what you think (based on the statements) the trader was trying to do.
 
Are these your trades?
Only asking as you seem to already know its manual etc.

Sample size isn't any way near enough to gauge anything really.
If they are yours, try SIM testing in Ninja using the method I described here:
http://www.trade2win.com/boards/hom...ing-home-trading-dead-gone-2.html#post2278332

Get the replay data from here:
http://marketreplaydata.com/
Otherwise you need to record your own data, or buy tick data:
http://www.tickdata.com/
Note - if you buy tick data, you will have to use Sierra chart as ninja
will not replay tick data.

Last point, many say SIM is worthless.
Bear in mind most prop and IB do use SIM initially.
Also, SIM is only worthless if you don't treat it the same as live,
and that includes standing by losing trades and mistakes - no re-run.
 
Are these your trades?
Only asking as you seem to already know its manual etc.
Yep all mine mate.

Sample size isn't any way near enough to gauge anything really.
For sure but that dont mean we cant try :)
Theres more than enough data to answer 2&3 imo.
Granted 1&2 are subjective but still, fun to guess.

If they are yours, try SIM testing in Ninja using the method I described here:
http://www.trade2win.com/boards/hom...ing-home-trading-dead-gone-2.html#post2278332

Get the replay data from here:
http://marketreplaydata.com/
Otherwise you need to record your own data, or buy tick data:
http://www.tickdata.com/
Note - if you buy tick data, you will have to use Sierra chart as ninja
will not replay tick data.

Last point, many say SIM is worthless.
Bear in mind most prop and IB do use SIM initially.
Also, SIM is only worthless if you don't treat it the same as live,
and that includes standing by losing trades and mistakes - no re-run.
Sim is great and i share youre view, but sim (or nano size) imo cant test everything.
Go on LV, have a stab.
Cheers
 
Example 1 : Just shorts , size is fixed @ 0.5 , result : he lost .

Example 2: Just longs , size reduced from 0.5 to 0.2 after a series of consecutive winnings , then back to 0.5 after a series of consecutive losses , result : made some profits .

The trade sample is too small , however i like the idea that he did scalp in the same direction without switching over and over again . And i like that he did play with the trade size a bit - non martingale - after getting a series of winnings or losses .
 
DT, Ok just taken a look. I'm assuming this is someone's demo account and they're just taking the p*ss?

Edit: So to be more clear - I don't think you actually want anyone to analyse this - correct?
Nope, all live, real money, all mine.
 
Yep all mine mate.

For sure but that dont mean we cant try :)
Theres more than enough data to answer 2&3 imo.
Granted 1&2 are subjective but still, fun to guess.

Sim is great and i share youre view, but sim (or nano size) imo cant test everything.
Go on LV, have a stab.
Cheers
Really there is no point, you need a sample size of at least 500 as an absolute minimum.
Preferably I would say 1000+ to guage anything meaningful.
Thats why I suggested SIM - its quicker to build that sample size.
Provided you genuinely treat it the same as live otherwise no point.

I don't understand most people's attitude to SIM, backtesting after all
is just historical batch tested SIM.
In fact, although I don't think reading TITZ is anywhere near enough,
that is one point he does cover - the gulf between live and SIM,
and the reasons for that.
Bridge that gap and you have an effective tool for discretionary testing.
EDIT - live testing is obviously essential as well for slippage testing, fill speed etc.
 
Last edited:
first batch - trying to fight the mighty dax to the downside (from top of a range?) and getting cuffed until finally decided to go with it in the second batch. Looks like anticipating momentum bursts, killing them quick if it didn't come (better than in the first batch) and taking profit when momentum faltered.

There's more trades here in an hour than I'd be doing in a week :)
 
There's more trades here in an hour than I'd be doing in a week :)

Yeah the costs issue alone with retail spreads and comms won't be helping.
Trading with this kind of frequency, you really need to be earning the spread or
at least buy bid, exit market.
Discounted comms is another boost.

That alone is usually the reason most retailers fail with this kind of frequency,
the above factors are usually not an option in the first place.
 
  • Like
Reactions: tar
sample size is enough to determine eventual outcome. I only look ed at the 1st 10 trades in example 1. He continuously selling as price is rising. End result would be zero equity. Didn't bother looking at the others.

Peter
 
Example 1 : Just shorts , size is fixed @ 0.5 , result : he lost .

Example 2: Just longs , size reduced from 0.5 to 0.2 after a series of consecutive winnings , then back to 0.5 after a series of consecutive losses , result : made some profits .

The trade sample is too small , however i like the idea that he did scalp in the same direction without switching over and over again . And i like that he did play with the trade size a bit - non martingale - after getting a series of winnings or losses .
Yep, in example 1 I lost. iyo, did i lose well or badly?
Ok, in example 2, im ashamed to say that 0.2 was a ticket error. Was to busy reppin PL posts!
 
sample size is enough to determine eventual outcome. I only look ed at the 1st 10 trades in example 1. He continuously selling as price is rising. End result would be zero equity. Didn't bother looking at the others.

Peter

Can work with lower frequency mean reversion, although its not that efficient.
Although yeah, keyword is continuously selling and frequency - not good.

Although I don't think the sample size is large enough, if the approach you
have noted is applied consistently and long term, then yeah I agree,
one way ticket to zero.
 
Last edited:
I don't understand most people's attitude to SIM, backtesting after all
is just historical batch tested SIM.
EDIT - live testing is obviously essential as well for slippage testing, fill speed etc.
Yep, sim/demo imo is great. The only id add to your (slips and fills) is psych
 
Yep, in example 1 I lost. iyo, did i lose well or badly?
Ok, in example 2, im ashamed to say that 0.2 was a ticket error. Was to busy reppin PL posts!

1-Yes lost badly : 700 points in just 2 sessions . ( If you didn't use a SL you would have lost much less in the 6th and 7th of FEB ) .

2- ROFL so you didn't pay attention to your trading size for what ? 8 trades , nonetheless it did work :LOL: , what a coincidence to do that by mistake after a series of losses and winnings .
 
Last edited:
first batch - trying to fight the mighty dax to the downside (from top of a range?) and getting cuffed until finally decided to go with it in the second batch. Looks like anticipating momentum bursts, killing them quick if it didn't come (better than in the first batch) and taking profit when momentum faltered.

There's more trades here in an hour than I'd be doing in a week :)
Objectives are different in each example and am no anticipating anything.
Yep lots of ins and outs. Lots spread paid but if the trader had a longer term view is that such a bad thing?
 
Oh ok... well here's some quick analysis then:

Overall win rate: 32%
Overall avg loss 1.4x avg win
Overall expectancy on this sample: -4.56
(Win rate increases to 58%, and expectancy to 1.7 in second sample).

Trade risk - unlimited? (looks like no S/L set on any trades)
TP settings are "interesting"...
Again too little data to be really meaningful.

Not sure what the point is of trying to guess the trader's method & objective... if the above is a system, it is a system that loses money.
No stops! by the man a beer! :clap:
"TP settings are "interesting" arnt they just! :)
expectancy figs are meaningless unless the objectives are known
 
Top