Posted by tosh 3 hours ago
>Pattern matching, on the other hand, can drive you to make moves like an advanced chess engine. Moves that others find alien and incomprehensible.
Sure. Index funds is a viral story which has a lot of rigorous data behind it. As for the advanced chess engine of intuition... ever heard of Fooled By Randomness?
To be honest i cant be bothered is the real reason I don’t play but maybe it’s the winning move?
It's like arguing with God: he's got better info that you.
Big Black Candle, Big White Candle, Black Body, White Body, Doji, Long-Legged Doji, Dragonfly Doji, Gravestone Doji, Hammer, Hanging Man, Inverted Hammer, Shooting Star, Long Upper Shadow, Long Lower Shadow, Marubozu, Spinning Top, Shaven Head, Shaven Bottom, Bearish Harami, Bullish Harami, Bearish Harami Cross, Bullish Harami Cross, Engulfing Bearish Line, Engulfing Bullish, Bearish 3-Method Formation, Bullish 3-Method Formation, Evening Star, Evening Doji Starv Morning Star, Morning Doji Star, Falling Window, Rising Window, Three black crows, Three White Soldiers, On Neckline, Doji Star, Tweezer Tops, Tweezer Bottoms, Dark Cloud Cover, Piercing Line, Darth Maul, Judas Candle, Island reversal
It's great that people think these are real.
Where I think pattern matching is essential is in debugging. Software is the easy one to explain, but problem-solving in life in general, too.
An experienced dev who's operated some systems can see odd behavior and be like "ah, feels like a race condition between X and Y" far faster than anyone could construct the narrative of that root cause from first principles.
But good luck doing that for a several decades in the financial markets... Maybe you don't go broke, but... there's a a whole lot of dumb strategies that can work better than average for a while, and it's actually pretty hard to do so devastatingly bad that you question your ability if you think you're the pattern-matching investor genius.
And the shot at index funds are silly. The beauty of index funds is that they raised the floor and stability for the average middle-income+ person investing for their future vs speculating to try to get a big edge. While bleeding less to overpaid advisors. That's real value, regardless of if some people do better or if you believe that you can consistently do better.
Investors literally can’t stop talking about “pattern recognition,” because it’s comforting to think that the past can predict the future if you’re clever enough. But the best investors are pattern breakers, not matchers.
This is false on so many levels. First of all illegibility also soothes. Many people prefer the “illegibility” of natural landscapes in the great outdoors over cold corporate offices with straight lines.
Second illegible systems are less predictable and harder to analyze. For example why do we have to build airliners in orderly ways using clean surfaces and simple shapes? Why not messy geometry and chaotic inconsistency? Because the math then becomes to hard to model and predict. Non linear chaotic systems are hard to predict using our quantitative modelling techniques.
In fact our modern engineering is built on top of quantitative modeling using applied mathematics and the only things we can really model are ordered and simple systems. Engineering is equivalent to prediction and it is our best form of prediction in the sense that you can get on a metal tube called an airplane and get hurled across the ocean and feel relatively safe because our predictive powers are so high that we know you’re incredibly safe.
Order is preferred and it is not bias. It’s because order has intrinsic advantages with analysis.
Performance is a different story. It may be the case that chaotic systems can lead to better results but we simply don’t have the tools to predict or understand how these systems work.
We are however hitting a new era Of engineering where we use blackbox neural nets to analyze illegible systems. It has downsides in the sense we can never truly fully understand the analysis nor can we fully trust it, but such an analysis is now possible to an extent. It also suffers from the fact that we need massive amounts of trial and error attempts to even get the data that enables building these models.
>Pattern matching, on the other hand, can drive you to make moves like an advanced chess engine. Moves that others find alien and incomprehensible.
Wow, this guy has found a way to consistently beat the market, he must be a billionaire at this point surely?