In machine learning, there is a technique in the data processing phase where you have to "vectorize" categorical fields. The reason this must be done is so a computer does not make false inferences based on arbitrary differences in labels. Or if someone incorrectly decided to encode labels with integers, a computer could make false inferences based on the numerical differences. For example, if you replaced New York with 1 and California with 2, a computer might conclude that California = 2*New York and generate inaccurate predictions based on this. As silly as this might sound, human beings are far worse at making false inferences -- look no further than the sports enthusiast who insists on wearing the same unwashed jersey while his team is on a winning streak.
In the book "Fooled by Randomness", these quack inferences also exist in the investment world and this is one of the major things the author criticized -- short term windfalls without an understanding of sample size and statistical analysis leads to bogus theories about how the market operates. Come up with enough investment strategies that you know are ridiculous, and one of them is bound to be successful in some small window of the market's history (or possibly in the future). It's worth noting that if you applied that strategy over a long enough period of time, it would approach it's actual return rate.
This is precisely where kepton is getting his inflated ideas about luck: because SOME investors get lucky in SMALL windows of time (just like some people win the lottery despite the odds being near zero), everything must be based on luck. One major problem with this (aside from the lack of understanding of conditional probability and sample size) is that it makes false equivalences based on arbitrary language usage. Just because we use the word investment to refer to buying stocks and for putting resources into a self-made business does not mean they both carry the same risks and rewards.
Create your own conditions which manipulates your individual probabilities (favorably or unfavorably) to make your own luck. Would you flip a coin for your success if you get to use a rigged coin that you made? I would. With the right conditions, success can be the mathematically expected outcome -- no luck required.
In the book "Fooled by Randomness", these quack inferences also exist in the investment world and this is one of the major things the author criticized -- short term windfalls without an understanding of sample size and statistical analysis leads to bogus theories about how the market operates. Come up with enough investment strategies that you know are ridiculous, and one of them is bound to be successful in some small window of the market's history (or possibly in the future). It's worth noting that if you applied that strategy over a long enough period of time, it would approach it's actual return rate.
This is precisely where kepton is getting his inflated ideas about luck: because SOME investors get lucky in SMALL windows of time (just like some people win the lottery despite the odds being near zero), everything must be based on luck. One major problem with this (aside from the lack of understanding of conditional probability and sample size) is that it makes false equivalences based on arbitrary language usage. Just because we use the word investment to refer to buying stocks and for putting resources into a self-made business does not mean they both carry the same risks and rewards.
Create your own conditions which manipulates your individual probabilities (favorably or unfavorably) to make your own luck. Would you flip a coin for your success if you get to use a rigged coin that you made? I would. With the right conditions, success can be the mathematically expected outcome -- no luck required.
Last edited: