Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
That is what I should have said in the first place...
I do not know what else to technically call it......higher chance, better trade sequence...whatever....I take 10 sequences into account, throw the 9 comfortable ones away, and take the one that scares me the most . Yes, yes...that's not exactly how it goes...it's freaking 10PM EST and I am still here...time for bed....
Can you help answer these questions from other members on NexusFi?
Then you cannot fairly discuss probability - because probability without market context is the only way to compare trading performance against pure "random chance alone."
This is why people are continually bringing up things like coin-toss probabilities when talking about trading, because it provides a context to talk about probability without opening a can of worms involving market context.
Many people believe that "random chance alone" for any given trade to succeed is somehow 50% - and that if your win-rate is less than that, then you are somehow doing worse than if you just bought-and-held, or threw darts, or some other silly non-strategy.
All I'm saying is that the actual "random chance alone" for any real-world trade to succeed is MUCH, MUCH lower than 50%.
If you flip a coin and go Long-on-Heads, Short-on-Tails, that is one parameter. Which has a 50% chance of success.
If you say "Long-on-heads, with a 10-pt stop", that is two parameters - with MUCH less chance of success than the one-parameter example.
Does this mean you have no probability of winning? Or a un-determinable probability of winning? In all reality, neither one of these things is true. Not only do you still have a probability of winning/losing, but it can be calculated within a reasonably accurate range.
A mechanical trader does have the advantage that their calculable ranges are probably tighter than a discretionary trader's, but that is all.
And in some cases, this could be a liability, rather than an asset.
"An approximate answer to the right problem is worth a good deal more than an exact answer to an approximate problem. " - John Tukey
Ok, I am still up...this is something I haven't thought about since school.....so I am thinking about it now.....Yes...correct.. technically speaking I am unfairly using the words with the concept... technically speaking.
I think we are getting carried away by definitions…….Probability as it pertains to pure random chance versus chance as it pertains to discretionary trading…..(if we were all systematic mechanical traders this entire forum at nexusfi.com (formerly BMT) would be the realm of mathematicians, statisticians and programmers and no one else)
If we take pure random trading….using independent normally distributed functions, each such added function would result in the exponential joint probability curve…defined as multivariate distribution
Each parameter (in itself normally distributed) at 50%, two would make it 25%, then 12.5%, then 6.25%...and so on……roughly speaking P(x1,x2,x3…) = P(x1) * P(x2)*P(x3)…..and so on…assuming normal distribution for each parameter…we could define this as you already stated….such as this…(the actual formula is much more dry….. here….but it essentially boils to the above and the plot below)
Yes, technically how can one argue against that? This is not important in discretionary trading at all…for one thing…we are now talking about normalized Gaussian distribution instead of multivariate joint exponential probability….so the curve immediately looks like this…..
On the other hand, when I take a short in an uptrend, I call that a low probability trade (maybe I should call it something else, like a low chance trade). When I take a short in a downtrend I would call that a higher probability (chance) trade……I think we are getting hung up on definitions……that wasn’t my point.
Call it whatever you want to call it, the point was that if one takes higher chance trades risk is automatically reduced of damage to account….identifying such higher chance trades is a different thing all together……Fair description or not, this is like missing the forest for the trees….this thread is about risk in trading, about mitigating and managing it. One prime way to do that is taking contextual trades, and increasing one’s read on the market to lower the risk of loss, not just by stops but more by not trading, trading smaller, scaling in…….passing on trades…etc etc…
So tomorrow morning, when I set up my sequences, I will still call them low or high probability trades based on my read…..whether that is technically correct or not…is probably not as important as getting my context correct.
It might be more interesting to discuss whether the markets are Gaussian or not, and how often do fat tail events squeeze the distribution, and how algos are changing the distribution curve..Although that’s way past me…
Edit - notice that i do not want to focus on chance percent of a sample of trades but the nature of the market itself and trades based on that.... because that to me is more relevant than a gaggle of normally distributed trades. So I am digressing from your response, and its on purpose.
Yes, excellent! That's all I was trying to say in the first place.
Then, it's the trader's job to try to make each parameter have the highest probability of winning - ie, taking with-trend trades, like you spoke of, etc. Dialing your stops up/down for greatest chances of success, & so on.
But I keep hearing this "50% chance of winning" idea, and wanted to point out that it's not correct ..
Now that it's been put to bed by your post, I think we're on the same page.
Do you believe a baseball player with 5 years of hitting stats has an inaccurate data set for probabilities just because he picks and chooses his pitches to hit?
Even though I am a big picture guy all the time, the devil is in the details. Many time I have ignored the small bits and they have come back to bite me...so thank you for making me think..things like this encourage me to come back to futures.io (formerly BMT)
Let's say I enter a trade here. My stop is 50 ticks away, and my first target is also 50 ticks away. I enter here because it is moving higher, and just bounced off a support level.
Maybe after I enter, price moves down some but not all the way down to my 50 tick stop, which was placed there because that is beyond the next support shelf level, and I concede that if price moves to this area, I was wrong about this trade. So it has moved down some -- lets say 20 ticks -- but then it stops and starts moving higher again.
This happens often, as it breaks through just enough of the prior support level to trap some shorts or to take out prior longs.
In this case, my trade is still valid because I had planned for this as a possible event. It is not enough to simply say this happens all the time, so wait for price to move down 20 ticks before entering in the first place. Sometimes that works, sometimes it does not. I prefer to take my position when I see my setup, and add to my position when I feel that the above scenario has happened, or the other scenario is when price makes a HH+HL combo in this case.
As for RTH session times, I am not following how that has any bearing. Sometimes I do hold trades overnight, but not most of the time.
Mechanical traders and automated traders base their probabilities on prior backtests or prior forward tests. Either way, it is on a set of sample trade data.
Discretionary traders do the same thing. As I explained before when you brought this up, if you take the prior 1,000 discretionary cash trades and analyze them, you can define threshold parameters for what the next few hundred trades may look like --- just like you would with a mechanical or automated system.
It is all apples. Not apples and oranges. Apples for my last 1,000 trades, and apples for my next 1,000. All apples and all consistent with the same approach.
Bob drives to work every day for the last 10 years using the exact same route from his home to his work place. He also drives home every day the exact same route from his work place to his house.
He drives himself. He does not have a robot driving for him and he is not using cruise control, GPS, or some other mechanical system. He is a discretionary driver. If he wants to change lanes, he does so, he does not need to wait for his rule book to say it is ok to change lanes.
Over the last 10 years, Bob has found it takes him between 30 and 40 minutes to get to work.
How long do you think it will take Bob to get to work for the next few months, on average?
A) 30-40 minutes
B) 5 minutes
C) 2 hours
. Yes, yes...that's not exactly how it goes...it's freaking 10PM 
