I think I figured out why trading can never work for so many people

bwolinsky · Mar 31, 2013

Quote from cornix:

Comes from my own experience of what I saw: most trading algos and most algos in general still have relatively poor abilities of reading context (not just in trading, but anywhere). For example Google cars can't drive in winter, because surrounding reality look changes. Not a slightest problem for a human to recognize the same roads and other stuff in winter or when it rains.

And trading with ability to quickly recognize context is still something human brain is able to do very easy and very well compared to robots.

That's why "human algo" (trader who developed her/his skills to the stage of nearly perfect self-automation) is still superior.

Only in cases when discipline is perfect of course, we don't consider classic emotional lapses here.

Just my subjective experience. If someone created a trading robot, which can consider context to filter stuff that must be great.

P. S. I beg pardon for possibly Al Brooks like writing style today... yesterday's night tempranillo still in effect a bit.

I mean the results I just said had less than a 52% win percentage and had a 55% win percentage before that.

In day trading, supply and demand's the tell, so back testing the purest information the market gives you should be profitable.

cornix · Mar 31, 2013

Quote from bwolinsky:

I mean the results I just said had less than a 52% win percentage and had a 55% win percentage before that.

In day trading, supply and demand's the tell, so back testing the purest information the market gives you should be profitable.

I'm fine with about 50% winning trades too as long as R:R is good (and it usually is).

bwolinsky · Mar 31, 2013

Quote from cornix:

I'm fine with about 50% winning trades too as long as R:R is good (and it usually is).

Swing trades should have much higher win percentages though.

eurusdzn · Mar 31, 2013

Why cant a simple formula determine context better than a human. Or, maybe a human that gets stuck..ha.

Say I design bands , surrounding and slightly offest by a small factor from the highs and lows of price bars on both hourly and daily bars. Always in , I sell at a top band tag of both hourly and daly bars and buy at the bottom tag of the hourly and daily . Two timeframes in play.
Maybe in a day or two the net trading results indicate that in the present context it is better to go either long or short and I trade that way until the context filter changes.
Is this bullshit or not? Maybe I will test it.

bwolinsky · Mar 31, 2013

Quote from eurusdzn:

Why cant a simple formula determine context better than a human. Or, maybe a human that gets stuck..ha.

Say I design bands , surrounding and slightly offest by a small factor from the highs and lows of price bars on both hourly and daily bars. Always in , I sell at a top band tag of both hourly and daly bars and buy at the bottom tag of the hourly and daily . Two timeframes in play.
Maybe in a day or two the net trading results indicate that in the present context it is better to go either long or short and I trade that way until the context filter changes.
Is this bullshit or not? Maybe I will test it.

It would be bs.

Don't confuse a simple formula or set of rules with hard mathematical derivation. Something simplistic like writing rules versus deriving indicator then applying them to data sets should be more of a priority than thinking about how prices look versus where they've been and appear to be heading according to those proofs that are so much more complicated than people give them credit for.

Being well capitalized hft helps, too!

HurricaneUS · Mar 31, 2013

Quote from bwolinsky:

...I've been considering ignoring you for a long time for your arrogance and stupid non-chalance .....

wise decision....I added oldtime to my ignore list eons ago

unfortunately I'm still able to see his/hers/thing rambling posts whenever he/she/it is quoted

cornix · Mar 31, 2013

Quote from bwolinsky:

Swing trades should have much higher win percentages though.

Probably, I've heard many times that longer-term signals are more reliable. But somehow suck at swing trading, because ironically lack patience and discipline to wait for a proper setup, which may happen once in a few weeks/months per market.

cornix · Mar 31, 2013

Quote from eurusdzn:

Why cant a simple formula determine context better than a human. Or, maybe a human that gets stuck..ha.

Say I design bands , surrounding and slightly offest by a small factor from the highs and lows of price bars on both hourly and daily bars. Always in , I sell at a top band tag of both hourly and daly bars and buy at the bottom tag of the hourly and daily . Two timeframes in play.
Maybe in a day or two the net trading results indicate that in the present context it is better to go either long or short and I trade that way until the context filter changes.
Is this bullshit or not? Maybe I will test it.

Probably because context part is the most subjective part of trading. Maybe not subjective, but complex, consisting of many inputs to be processed simultaneously.