For problems with a finite set of outcomes (like poker, chess etc.), this is correct.
But up to now, the CFR results that have been published aren't very convincing
for financial markets. Living on other people's blunders was also possible there, but
not very attractive after commissions. One of the major problems besides competition
(also a finite set in poker) was the moving benchmark with an infinite (or at least unknown)
number of outcomes - ie.: CFR or other strategies have been rejected when tested on dominance.
(For poker, CRF is clearly dominant.)
The following user says Thank You to choke35 for this post:
Don't get me wrong, I love math - but I have come to learn trading is not a math problem. It's a game about herding cattle, getting them all wild and crazy and driving them off a cliff to perish - oh and by the way you get $100 for each one that goes off the cliff.
I think the math helps you gauge things, much like the tachometer of a car helps you understand how fast your car is going or can go. It does not tell you the "exact speed" but you can get an idea.
I see the math is a measurer, not a predictor.
Reminds me of the story, the old bean traders is shown a new fangled indicator by some hotshot developer. The developer says see that, the market will now go up now. To which the old bean trader says oh really, and then makes a call to his broker and says sell 100,00 now.
With regards to the math, I believe the missing component is the sentiment of the transaction - for example is the sell a sell to get in, a sell to get out, a sell to entice more buyers, a sell cause I'm stupid and pushed the sell button and all these applicable to the other side buys of these transaction; this list is boundless. To me it seems to solve as a discrete math problem you would need to know more about these input controls, information which is simply not available and ever changing. Then take into account the interrelationships of markets macro and/or micro.
Me, I prefer learning the game of how to see the cattle being herded, identify the cliff, and be ready to take action as required.
Last edited by tulanch; July 8th, 2015 at 02:44 AM.
The following 3 users say Thank You to tulanch for this post:
You are correct about sentiment, or motive, or intent. There are lots of reasons to sell, not the least of which is that the seller has already made so much money that if he makes any more he'll have to go lie down somewhere. But the "why", while interesting, and a good way to entertain oneself during the day, is not particularly relevant to the trading decision. Either price hits your entry stop or it doesn't. The "why" is not foremost in the trader's mind. Or if it is, it shouldn't be. Concerning oneself with "why" more often than not leads to hesitation, and hesitation is not your friend when it comes to daytrading.
The following 3 users say Thank You to DbPhoenix for this post:
I think the main difference is AI always operate off rules and axioms, core assumptions, whereas humans can question and change assumptions. In the end I think it boils down to one is conscious and can make choices (not operate off rules), and one isn't.
Understanding yourself is just as important as understanding markets.
The following user says Thank You to TickedOff for this post: