Is it worth selling!?

[_Aryan_]

I was thinking if there could be a mathematical way of deciding things. Basically if you are stuck in a situation where you want to know if you should take an action or not in a scenario. And a system came to my mind which takes data from you regarding a situation in 3 formats-
Past situations- data regarding similar situations that happened in the past.
Present situation- data regarding current situation.
Future certanities- data of future events certain to happen.

After getting all this. The program segregates the data and decides if a data statement is in your favor or not.

After all this computation. It gives you 2 percentages. One whether action must be taken and 2 if it shouldn't be taken.

3 of the merits of the program is that it will tell you the probability of events happening from the inside given data or something that never happened. It's editable if more data statements are added with time it changes the percentage accordingly. You can pick your weaknesses and work upon it to increase your chances and if against anyone else, you can branch out his weakness statements.

It's accuracy depends on the data. If you provide 100 percent data, you get 100 percent accuracy, if you provide 40 percent data, you get 40 percent accuracy.

This can be used in-
Small scale life decisions
Stocks analysis
Large scale decisions
Business decisions
.....etc

In a nutshell, it's a mathematical way of deciding things.

I've already programmed it and tested it on a small scenario to compute who has more winning chances and it came out to be right.

I don't know if it's a normal project I've made or it's worth selling.
Would like to hear your opinions!
Thanks!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Johnny boy]

The benefit of these things is that you can usually come up with a decision yourself, it's just when it asks for the data, you become conscious of the relevancy and accuracy of the data. It gives structure and clarity to your decisions before it even spits out an answer. I see a lot of foolish decisions because people lack the critical thinking to slow down and analyze the choices, they simple have an underlying emotion and let it guide their decisions.

More useful as an exercise in critical thought than it is as a tool to actually make your decisions.

No I would not 'sell' it. Perhaps it might work as a free and neat little online tool, with something on the side that is your real offer. This is a lead magnet if anything.

 

[WJK]

In a nutshell, it's a mathematical way of deciding things.

I've already programmed it and tested it on a small scenario to compute who has more winning chances and it came out to be right.

I don't know if it's a normal project I've made or it's worth selling.
Would like to hear your opinions!
Thanks!

How is this system working for you? Over time, how have those decisions worked out? It seems to be rather an impersonal system for making decisions. How are you weighing the data? That would seem to be the elephant in the room.

But, on the other hand, I go through a process when I'm making a major decision. I make a line down the middle of my paper. I put the plus items on one side and the negative items on the other. Then I weigh the two sides to help me decide. It is only an influencer, not a determinate. I must feel that the decision is right in my heart before I can get behind it.

 

[_Aryan_]

How is this system working for you? Over time, how have those decisions worked out? It seems to be rather an impersonal system for making decisions. How are you weighing the data? That would seem to be the elephant in the room.

But, on the other hand, I go through a process when I'm making a major decision. I make a line down the middle of my paper. I put the plus items on one side and the negative items on the other. Then I weigh the two sides to help me decide. It is only an influencer, not a determinate. I must feel that the decision is right in my heart before I can get behind it.

There are to aspects for a decision
Emotional and logical. We are designed to take good chunks of emotional data and analyse it but the same is not with logical thinking. We can miss out big parts while analysing big situations with huge data, like political circumstances. It's like asking a 32 GB computer to work with 1TB of data. It's impossible for a small computer to calculate 1TB but it's possible for a supercomputer for sure. Emotional aspect does not need such devices but i believe when we move up to the logical aspect this with huge data, this can help alot. Every data statement will be weighed according to its importance in your situation. You can take a data statement as base 1 and rate other statements accordingly to their importance in comparison. As i said this system will not let you skip the logical thinking part. Which happens frequently.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[_Aryan_]

The benefit of these things is that you can usually come up with a decision yourself, it's just when it asks for the data, you become conscious of the relevancy and accuracy of the data. It gives structure and clarity to your decisions before it even spits out an answer. I see a lot of foolish decisions because people lack the critical thinking to slow down and analyze the choices, they simple have an underlying emotion and let it guide their decisions.

More useful as an exercise in critical thought than it is as a tool to actually make your decisions.

No I would not 'sell' it. Perhaps it might work as a free and neat little online tool, with something on the side that is your real offer. This is a lead magnet if anything.

I would for sure agree with you on the first para of your reply. But you can't perform critical thinking in situations which are seemingly very big. You can't take all the aspects of a political situation in your skull or on a paper and apply mathematical thinking to it just like a normal laptop cannot handle 100TBs of data. Emotional aspect is easy for humans because we are designed to think that way. Machines can apply math and logic better than us. So I suppose this program can help us understand the logical picture of a very huge scenario which might not be possible for us to calculate.

 

[Jobless]

Provide an actual example and use case. Who has the input data, and who needs the output result? Saying it can perform 'stock analysis' or make 'small scale life decisions' is too general.

 

[Devilery]

Can you define it in a sentence or two? Currently, it sounds like a very general idea that may be extremely tricky to execute.

If can build it, can you guarantee consistent accuracy? How is the data sourced?

Data plays a role in every decision. There are "analysts" spread across all sectors. There are stock analysis tools. There are real estate market analysis tools. Product research tools. There are coaches, mentors, and consultants who help you make better decisions.

You say that you've already tested it. Share more on the backend process of that. Obviously, there are billions to be made if your service/software can accurately tell the "most right" decision, but I feel like everyone wonders about the mechanism that does the analysis part.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[_Aryan_]

Provide an actual example and use case. Who has the input data, and who needs the output result? Saying it can perform 'stock analysis' or make 'small scale life decisions' is too general.

I would for sure agree with you on the first para of your reply. But you can't perform critical thinking in situations which are seemingly very big. You can't take all the aspects of a political situation in your skull or on a paper and apply mathematical thinking to it just like a normal laptop cannot handle 100TBs of data. Emotional aspect is easy for humans because we are designed to think that way. Machines can apply math and logic better than us. So I suppose this program can help us understand the logical picture of a very huge scenario which might not be possible for us to calculate.

Ok I'll give a super simplified answer for example- 2 people are standing in a remote desert and only one of them can get off alive because one of them has a gun. Now this is a very unlikely and weird example but it's for the sake of simplifying, otherwise I could've taken complex examples like decision of picking up a college, political decision for a country etc.

Now suppose A has gun and B hasn't. So there is a 100 percent chance of A coming out alive. But how can I prove this with my program.

Taking data from exactly similar problems from the history and it's outcome.
Then segregating data into your positives, weaknesses and into your opponent's strengths, weaknesses. (Not necessary that you have an opponent). I use Set Theory for this segregation.

Now data from current situation or i can say in your case. Doing the same segregation.

And doing the same with future certanities.

Now, this input doesn't have to be in a line or even words. 5 letters are enough as i have programmed it to be. Now after all this input it will gives you two percentage as i mentioned in my post.

Now this example may seem dumb but, it's like saying I just did a 2 x 4 = 8. But you would need a computer for multiplying values in billions and trillions. This example is a 2 times 4. I use it for the latter case.
So summing it all up.
I segregate data into positives and negatives data statements from all the data and rate their importance. And also give you the probability that if the outcome will be based on previous historical situations or new.
Based on Set Theory and probability.

You can give me a situation and i can do it for you and explain it you if you want.

 

[_Aryan_]

Can you define it in a sentence or two? Currently, it sounds like a very general idea that may be extremely tricky to execute.

If can build it, can you guarantee consistent accuracy? How is the data sourced?

Data plays a role in every decision. There are "analysts" spread across all sectors. There are stock analysis tools. There are real estate market analysis tools. Product research tools. There are coaches, mentors, and consultants who help you make better decisions.

You say that you've already tested it. Share more on the backend process of that. Obviously, there are billions to be made if your service/software can accurately tell the "most right" decision, but I feel like everyone wonders about the mechanism that does the analysis part.

As i told If you provide all the data in the situation result will be accurate and if you provide half of it then result will be accordingly. Data is what you know about the situation, everything you possibly know.
I used Python for it. And mathematical concepts like Set Theory, Probability and other small calculations like percentage and segregation.

How bout you give me a decision and i give you the results.

 

[Guest-5ty5s4]

How about a use case with real numbers, like a financial transaction?

For instance, I've been living in a home for 1 year while repairing and upgrading it. A "live in flip." I now have somewhere between $100,000 and $150,000 in equity. I'll need to get an appraisal.

Now that it's the 1 year mark (15% capital gains tax), I can't decide if I should sell it and move, keep it and cash out refinance (more debt but also more cash in pocket), stay another year for no capital gains taxes and sell (is 15% saved really worth waiting another 365+ days?), etc.

This is definitely a problem that a computer program could analyze. I don't think there's a right answer, just pros and cons.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WJK]

There are to aspects for a decision
Emotional and logical. We are designed to take good chunks of emotional data and analyse it but the same is not with logical thinking. We can miss out big parts while analysing big situations with huge data, like political circumstances. It's like asking a 32 GB computer to work with 1TB of data. It's impossible for a small computer to calculate 1TB but it's possible for a supercomputer for sure. Emotional aspect does not need such devices but i believe when we move up to the logical aspect this with huge data, this can help alot. Every data statement will be weighed according to its importance in your situation. You can take a data statement as base 1 and rate other statements accordingly to their importance in comparison. As i said this system will not let you skip the logical thinking part. Which happens frequently.

I'm a retired commercial real estate appraiser. I've dealt with analyzing numbers and data for years -- laid out in graphs and on forms. Those numbers give a range and indicators. They do not give a bottom-line answer to the estimate of value. We all know that numbers can lie. People can make them say things that aren't necessarily true. I keep that truism in mind when I look at any studies or results.

 

[Johnny boy]

ok after hearing you explain it I know it's a terrible idea

 

[Jerma]

I'm sorry to say it, but your idea in its current form is doomed to fail.

I won't go into details, but you can google
The frame problem
Combinatorial explosion

On the other hand, you could check those things out:

There's a branch of computer science that specializes in this sort of problem. It's called operational research. They use mathematical modelling and heuristic ("good enough" algorithm) to side-step combinatorial explosion. It's great for problems like the one @thechosen1 talked about.

Let's say you have 3 factories that produce 30 products and you know the cost and profits of every product. You also have constraints: not all your employees can operate all the machines, some machine needs to be shared between different product lines or needs to be reset between uses, it takes time to transfer parts between factories, you have stocks limits, etc.. You want to know how to allocate your resources to maximize your profits.

You can model those constraints with mathematical equations and feed them into an algorithm (probably could use the "simplex algorithm" here), and it will spit out the optimal solutions. The problem is that this only works if the model is not too complicated (polynomial time-complexity). The other problem is that reality is always more complicated than any model you can construct.

Another branch that deals with those problems is AI. AlphaGo is a great example. Using deep learning, they side-stepped the combinatorial explosion of computing the optimal solution. As a result, AlphaGo doesn't "know" the optimal moves; it has something akin to intuition.

Edit: Good video about the frame problem

View: https://www.youtube.com/watch?v=IHE63fxpHCg

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WJK]

I'm sorry to say it, but your idea in its current form is doomed to fail.

I won't go into details, but you can google
The frame problem
Combinatorial explosion

On the other hand, you could check those things out:

There's a branch of computer science that specializes in this sort of problem. It's called operational research. They use mathematical modelling and heuristic ("good enough" algorithm) to side-step combinatorial explosion. It's great for problems like the one @thechosen1 talked about.

Let's say you have 3 factories that produce 30 products and you know the cost and profits of every product. You also have constraints: not all your employees can operate all the machines, some machine needs to be shared between different product lines or needs to be reset between uses, it takes time to transfer parts between factories, you have stocks limits, etc.. You want to know how to allocate your resources to maximize your profits.

You can model those constraints with mathematical equations and feed them into an algorithm (probably could use the "simplex algorithm" here), and it will spit out the optimal solutions. The problem is that this only works if the model is not too complicated (polynomial time-complexity). The other problem is that reality is always more complicated than any model you can construct.

Another branch that deals with those problems is AI. AlphaGo is a great example. Using deep learning, they side-stepped the combinatorial explosion of computing the optimal solution. As a result, AlphaGo doesn't "know" the optimal moves; it has something akin to intuition.

Edit: Good video about the frame problem

View: https://www.youtube.com/watch?v=IHE63fxpHCg

You are right. The video was so true. Like I said in my earlier post, I've done a lot of statistical analysis over the years. One thing I learned is that SOMETIMES, surprise, surprise, the answer to the question or the trend, is the outlier or black swan event. It takes common sense to see the relevance of that factor on the whole. In most analyses, those stand-alone pieces of data are discarded due to their non-homogeneous nature. But, they are critical to understanding the situation because humans are herd animals. We turn our collective direction on a dime. These pivots are roughly cyclic, but not precisely predictable. New events, inventions, and social movements can take them in diverse and unexpected directions. The statistics are one factor that must be reviewed and weighed. They are only a factor, not a determinate.

 

[WJK]

I was telling my husband at our lunchtime about this thread and the associated video. And I was struck by this thought. There are studies that have concluded that a novice in a discipline or sport tends to group data differently from an expert in the same type of situation. Part of the process of weighing the data used in a statistical analysis is to understand how that data is grouped and how that grouping affects the conclusion.

For example: A novice baseball player would group the hoop shots together, the dribbling together, etc.. An expert would group several activities together (the dribbling, the pass-offs, the hoop shot, etc.,) that are involved in the whole play to making the basket. It's a totally different point of view. We do the same in business. We don't segregate the initial contact from the selling presentation, from closing the deal, from the delivery of the goods or services, from the customer service, etc.. We see it as a seamless flow activity concerning servicing that customer.

A machine cannot assign those different weights to the data and make sure the data points are properly grouped. Again, this takes a human common-sense analysis to ensure that the end results have any validity.

 

[Raedrum]

As i told If you provide all the data in the situation result will be accurate and if you provide half of it then result will be accordingly. Data is what you know about the situation, everything you possibly know.
I used Python for it. And mathematical concepts like Set Theory, Probability and other small calculations like percentage and segregation.

How bout you give me a decision and i give you the results.

Your reasoning is indeed very pleasant on a theorical approach, but I don't think the ratio data/accuracy is an identity - or even constant - function. Having half of the data doesn't mean you'll have half accuracy.

Having a godlike understanding of things is not achievable physically, but don't misunderstand me: data analysis is actually an high-demand skill in the industry.

Risk management is a typical exemple of applied mathematics in business decisions making. And larges businesses are willing to pay big money for it.

MJ DeMarco give an exemple of it in The Millionaire Fastlane : his weighted decision matrix, which is a over-simplified exemple of your theory but has given greats results anyway.

So if you're interested in this field and good at it, you might be able to develop a very profitable skill.

(and yes python is definitely the tool to go for, at least for technical analysis)

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Jerma]

Part of the process of weighing the data used in a statistical analysis is to understand how that data is grouped and how that grouping affects the conclusion.

Yes, that's the essence of the frame problem. The world is infinitely complex. There's too much data to be able to process everything. There are near-infinite ways of grouping the data. So an important feature of intelligence is the ability to correctly frame things. If you have a good frame, you can use your resources more efficiently (pay attention to the right things) and achieve your goals more effectively.

The frame contains the information you can "work with," and everything else is outside of the frame. So the frame is a heuristic. We need heuristic because we don't have infinite time and processing power. Also, we don't have God-like perceptions, so we only have access to a fraction of the information out there.

A heuristic is a good enough algorithm that takes shortcuts to achieve an answer close to the actual answer.

For example, I'm tasked with counting the number of people in a packed stadium. One correct algorithm would be to count everybody one at a time, row by row. A good heuristic would be to count how many people are in a row, count how many row there is in a section, count the sections, and then do simple math. The answer won't be the actual number of people in the stadium, but it will be good enough for most situations.

In the baseball example, the expert won't pay attention to the players' socks' color in relation to their performance. However, a machine learning AI with a huge dataset will.

At the same time, why is it so obvious to me that the color of the socks is irrelevant? It seems obvious that the color of the sock has no impact on the mechanics of playing basketball, but I can't really explain WHY it's obvious. I just know it.

It's because of our build-in common sense. The human brain is fully pre-loaded with heuristics that help us know what to pay attention to and what to ignore. When those heuristics function correctly, we call them intuitions, and when they don't, we call them cognitive biases (confirmation bias, anchoring bias, availability bias, etc etc). The difference between the expert and the novice is that the expert has fine-tuned his frame, got rid of his bad intuitions and created new heuristics (like a good player should train at least 1h a day).

The process of building frames is automatic for humans, so we take it for granted. We haven't figured out how to teach machines how to build good frames by themselves. So if we want an intelligent machine right now, we need to do all the intelligent work upfront of picking the correct data and programming the right algorithms and heuristics for the situation.

To be honest, I don't feel we know how to teach it to other humans either. Some people have common sense, some people have.. less common sense. The expert can't really transmit his frame to the novice with words. Instead, he can guide the novice through his own learnings.

That was longer than I expected, sorry for the rambling!

 

[WJK]

Yes, that's the essence of the frame problem. The world is infinitely complex. There's too much data to be able to process everything. There are near-infinite ways of grouping the data. So an important feature of intelligence is the ability to correctly frame things. If you have a good frame, you can use your resources more efficiently (pay attention to the right things) and achieve your goals more effectively.

The frame contains the information you can "work with," and everything else is outside of the frame. So the frame is a heuristic. We need heuristic because we don't have infinite time and processing power. Also, we don't have God-like perceptions, so we only have access to a fraction of the information out there.

A heuristic is a good enough algorithm that takes shortcuts to achieve an answer close to the actual answer.

For example, I'm tasked with counting the number of people in a packed stadium. One correct algorithm would be to count everybody one at a time, row by row. A good heuristic would be to count how many people are in a row, count how many row there is in a section, count the sections, and then do simple math. The answer won't be the actual number of people in the stadium, but it will be good enough for most situations.

In the baseball example, the expert won't pay attention to the players' socks' color in relation to their performance. However, a machine learning AI with a huge dataset will.

At the same time, why is it so obvious to me that the color of the socks is irrelevant? It seems obvious that the color of the sock has no impact on the mechanics of playing basketball, but I can't really explain WHY it's obvious. I just know it.

It's because of our build-in common sense. The human brain is fully pre-loaded with heuristics that help us know what to pay attention to and what to ignore. When those heuristics function correctly, we call them intuitions, and when they don't, we call them cognitive biases (confirmation bias, anchoring bias, availability bias, etc etc). The difference between the expert and the novice is that the expert has fine-tuned his frame, got rid of his bad intuitions and created new heuristics (like a good player should train at least 1h a day).

The process of building frames is automatic for humans, so we take it for granted. We haven't figured out how to teach machines how to build good frames by themselves. So if we want an intelligent machine right now, we need to do all the intelligent work upfront of picking the correct data and programming the good algorithm and heuristics for the situation.

To be honest, I don't feel we know how to teach it to other humans either. Some people have common sense, some people have.. less common sense. The expert can't really transmit his frame to the novice with words. Instead, he can guide the novice through his own learnings.

That was longer than I expected, sorry for the rambling!

I too don't know how to teach common sense to others -- except through being an example. People ask me how I do certain things. Most of my activities have become intuitive rather than deliberate. These days I don't think through most things on a step-by-step basis. I just do them. You're right. I have most parts of my life in a "frame" that was built up over many years. And those pre-choreographed steps save me a world of hurt.

 

[CaptainAmerica]

It reminds me of Geek Logik: 50 Foolproof Equations for Everyday Life by Garth Sundem. One of my favorites is readily available for interactive use: The Can I Buy It? Calculator.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Frost]

If only life were that simple.
I suggest you read Fooled By Randomness by Nassim Nicolas Taleb