While I can forecast reasonably well I think, my big question I've been thinking about is beyond increasing my score on this website, what tips do people here have on using forecasting to add value to ones life?
I think forecasting adds value by forcing us to focus on accountability and keeping ourselves humble. With increasing polarization it's easy and common to make wrong and overconfident predictions, so by taking a step back and closely tracking our predictions and how accurate they were we can be a force in the opposite direction. I think that accountability is valuable in our personal lives but it's valuable to the communities we're part of as well.
@Ryan-Beck Perfectly true. By being forced to live with the consequences of your prediction concerning your Brier you learn the same lesson again and again with each question: avoid only having an opinion. Go for the truth as much as you can. Look at all angles. Look for angles you didn´t consider possible so far. Let that become a habit. Here it might just look like "brier vs. median". In the outside world it´s about developing better judgment for the benefit of your life and your community.
Thinking analytically from all viewpoints and asking questions of what you see without your own biased assumption will prevent you from being disappointed when things don't go your way. You will also stray from emotional decision making based on other peoples personalities.
Thanks for your post. I think this is a very good contribution.
Effectively, it continues the theme that forecasting is more of a self-improvement tool, but your link provided a number of concrete ways to incorporate this self-improvement into ones life to make better decisions.
However, by focussing on the self-improvement angle (as per most replies) there are 2 issues. 1) Once the skills have been aquired and we know pretty much how to make a good decision, there are diminishing returns from continuing to forecast or becoming better at it. It's kind of binary, you have the skills or you don't. Taking that extra effort to be a grandmaster will not help one's life much better from a self-improvement prospective than the keen amateur. And 2) It suggests the core primary product of our work, the prediction itself, has no obvious use. So it then becomes more like a crossword puzzle than something of real world practical importance.
If we could make more of a case for 2, then we make more of a case for becoming far stronger at this, rather than simply working until you get the gist of most of the skills and then that's pretty much enough in terms of value per unit time invested.
You are right that forecasting requires the construction of a world that once made has a life of itself.
I have been looking at the dizzying heights of that world for many years now, seemingly alone until I found out about this website.
So yes, it's like you first answer a few questions and then the whole world opens up.
Thinking in correlations is also important. For example, I used this to make a legal decision, where my underlying case was strong (95%) and my case for not having to pay costs if I lost was strong (95%) but there was a very high correlation between losing one and losing the other even though both were based on different facts and law because both are tried infront of the same judge. So I chose to not assume I had approx 0% costs risk (1-0.05*0.05) and assumed it was closer to 95% and settle because I couldn't risk paying the enormous costs of being unlucky. As you say, correlations (or more generically co-dependence).
But you then said "Understanding things like that is very useful." When I read your post, I thought this was the intellectual leap.
I agree seeing the world from that place is beautiful, satisfying even. It's certainly intellectually stimulating. But I don't follow it's actual use. That being said, are you suggesting that my decision to settle while others might have proceeded gung ho via a naive 1-0.05*0.05 calculation is the full value, that I might now be financially ruined had I continued while now I have no risk of that and quit at the right time?
This argument even if very well made remains solely the self-improvement angle, and as per my above post, it suffers from 1) diminishing returns per time invested; and 2) the actual forecasts remain of no value.
I suppose, what I am really seeking, is here we all are, having worked one way or another to get these forecasting skills and we can see the wood from the trees from the clouds from the birds from the bats etc. But beyond just helping ourselves in doing better in little and big choices we make each day or decade, can we not achieve anything more meaningful for humanity itself, enhance one's career or perhaps obtain financial gain directly as a result of our forecasting skills?
@sebbysteiny Concerning substential financial gain as an indirect effect can be achieved by just making better decisions in life as described. Not playing russian roulette is a good idea, even if the chance of losing is "only" 5%, so that might have saved you from serious trouble. I think, you were absolutely right with that decision. With that in mind people wouldn´t marry persons who don´t fit, buy houses they can´t afford ore have other expensive habits, that will cause trouble in a "russian roulette style".
But you wanted a practical example, so her is one. I decided do reduce my consulting and sales work drastically for a few months to avoid any risk of getting my father infected through me and my contacts (he would be a risk patient, so it would have been another case of russian roulette). Why did I do so? Because I´ve been active in Covid-19 questions on GJO since February (reading every available info very early), so I knew the risks. When everybody else was still partying on Carnival my closer contacts had food for a month of potential quarantine in their homes and reduced their contacts to a minimum - because I knew, what consequences this could have. I had the right information before most others and applied it to my personal life. That was a direct positive consequence of this.
About direct financial aspects: if you want to make money directly by forecasting I consider any kind of betting no very good idea. Even the stock market is not as rational and predictable as it should be, so that´s quite risky. You might find a job, where it´s other people´s money you play with, but I don´t think there are many legal ways to do so, unless you do it for banks perhaps.
Last version and perhaps the best for you: selling forecasting knowhow is something at least two people in my homecountry do full time for a living. Both started their careers here, so I´m quite sure, that they still work for Good Judgment Inc.. I don´t know, how well it´s paid, but with all the other learning benefits involved there wouldn´t be almost 200 superbright and capable people working as professional Supers for them, if it wasn´t worth it. Just give it a try to find out yourself. It´s hard (I´ve been working on it without being selected so far), but the harder a way, the more there is to be learned on it...
Those two mentioned also sell forecasting knowhow to companies, government agencies, holding workshops (I don´t know, if they also do so directly or over GJ Inc only) and writing books about it (not my cup of tea), but of course you need a superb track record to have enough knowlege and credibility for that. Guess, where you can get that. Yes, here. And yes, both suggest to start here like they did.
Thanks again for your considered reply. I don't know this superforecasting land very well as I've joined very recently.
In terms of value beyond 1) self-help, you have identified the following:
2) Self-Defence / Insurance vs events in immediate neighbourhood;
3) Betting e.g. on stocks (which you discouraged);
4) Teaching forecasting.
and I add a new use:
5) Political activism (direct use of forecasting)
By understing how politics unfolds, it enables identification of the parts that really matter towards determining the outcome. This empowers you to use your voluntary resources (time and possibly money) to maximum effect. You won't guarantee winning of course, but have done more for your cause than the same resources without forecasting.
4) Teaching (indirect - self-help)
I was aware teaching happened (after googling superforecasting jobs) but I wasn't aware it was a big thing. And it seems the competition is fierce, and probably the only route to this approach is large amounts of work trying to get your Brier score up on this website. As a late starter, I'm well behind the game relative to others on that. In my case, I already have a well paid quant job so it's hard to see a move to superforecasting teaching as value per unit time invested in my case.
2) Self-Defence / Insurance (direct use of forecasting)
This was in fact the main reason I started doing it, and probably still do. When things go bad, they can go very bad. By forecasting Covid 19, it may have helped me be on top of the curve at least sometimes. But there are events in stable democracies that can cause it to become destablised posing a clear and present danager to my life and that of my family and friends. By seeing this early, I might be able to e.g. get out in time. Will such an event happen "on my watch"? Don't know. That is definitely value in my view and can payoff bigtime (by saving my life), but it is unlikely to pay out much over my life time. So is it possible to get more value on a more regular basis?
3) Betting / investing (direct use of forecasting)
It is my view that every skill one has has value that as proof of value can be monitised. If you spend ages learning magic for fun, you could sell it as entertainment at weddings and parties, or use it on dates or at work to obtain social value (with a similar effect as expensive shoes). I am passionate about wine tasting. I get value by buying better wine for less at supermarkets, wine shops and restaurants and by buying cheap at vineyards in certain places while on holiday, along with helping my friends do the same. In office parties, I often help with the wine thereby giving me a career boost and I have found wine "chat" also has some social value.
So I come to forecasting. Where is the value in that? The obvious answer to me is if you can see how things work out better than others, then you should be able to make money from it.
I would never invest in a single share because I don't see how I could get enough data to beat those who do this full time even if they are not as good at forecasting as me.
Funds don't work either, because all I can do is do sensible investing stuff like diversify as much as possible and not sell a holding just because it went down.
Which leaves me with direct betting on ones forecasts. The key here would be to determine how much of what you have seen has been priced in and only bet where you know there is a big difference between your forcecast and the current odds. For betting in betfair, this is easy to do. But you can bet via proxies too (e.g. if there is a Brexit deal, GBP will go up) but you are taking a risk not just on your original forecast but your secondary forecast that your proxy will behave how you hope in that event.
You say "I consider any kind of bettin no very good idea". Now I thought this too for years, but then I came accross the real financial risk called "loss of opportunity risk" (https://en.wikipedia.org/wiki/Opportunity_cost), which essentially means by not investing you are still taking a risk, namely risking missing out on the benefit.
There is nothing special about 0 in terms of P&L. If you know Macron will lose the next election but don't bet on it, you have lost out to the same extent than if you bet on Macron and lost.
So the optimal is not to do nothing or you get sunk by loss of opportunity risk.
On the other hand, putting all your money on Belarous' president falling is also a bad idea.
To make the point further, remember buying insurance is essentially just a bet. But not buying insurance is also a bet, albeit with better odds but higher possible downside.
The goldilocks zone here is balanced investment. If you have a pension or ISA or other form of sensible savings, you should have no doubt heard about the need to diversify it as much as possible. In which case, consider your superforecasting bets as simply another pot. So if your investment portfolio is $100 in US, $100 in EU, $50 in UK, $100 in China, $100 in rest of Asia, $100 in Latin America etc [DISCLAIMER: THIS IS NOT INVESTMENT ADVICE, JUST AN ILLUSTRATIVE EXAMPLE OF WHAT DIVERSIFICATION MEANS, FOR INVESTMENT ADVICE, SEE AN INDEPENDENT FINANCIAL ADVISOR], why not then create another £100 "super forecasting" pot to bet with?
You then want to do as many small bets on different forecasts as you can limited by your pot size, focusing only when the betting markets differ significantly from your forecast.
The key here is to only bet on things you have already forecast e.g. on this website so there is no additional time taken, you are just monitising what you have already done . It may also make you more determined to get your forecast right if you know it will likely translate into your personal investment in your talents.
I haven't done this. This is just for a discussion of ideas. If I were to do it for real, I would probably do something similar to that, unless of course you say something that makes me think this is a dumb idea.
@sebbysteiny This really seems to be a good summary of "what to take from it beyond making better decisions" (while I still consider that being the key point).
Very little time today, so just short comments: yes, understanding political develompents of course helps a lot with investment. And yes, of course you´re right about "the biggest risk is not taking risks" in that field. On the long run stocks (the right ones) have always been the best investment. And not doing anything guarantees you some 2% negative profit (inflation), so you need to do it, and when you do it, it´s helpful to see trends others don´t see yet to not slip away chance you could have taken to your advantage. With closer analysis you will find, that many questions here are small pieces in a big puzzle being very much connected to the big picture. Most are definitely not randomly chosen. From thinking about, why US Intelligence Agencies or companies might be interested in certain things you can learn a lot about the world.
You might also find this page interesting:
https://goodjudgment.com/
This is a great thread of comments. Thank you for starting this @sebbysteiny
It is an interesting topic, the search for value.
From my perspective as a chemist, honing my forecasts has helped me discern the signal to noise ratio in aspects of many topics I had no formal expertise in (politics, finance etc.). At the beginning it was a bit frustrating how entangled this two are in public/social topics but I came to appreciate the presence of this ratio and its ambiguity. It's part of the "wider picture of the world" that @pstamp mentions and accepting there will always be "unknown unknowns" (which we shouldn't replace with "biased knowns").
Understanding features of this picture and parts of social and economic dynamics is very valuable and it translates to less time spent on over-biased debates and decisions. I have many friends who have a very scientific approach in their life but routinely fail to see outside their perspective.
Reading your post carefully, it seems you are adding more into the 1) self-help category because by improving your understanding about the world, you make better quicker decisions. Is this a fair summary?
"I have many friends who have a very scientific approach in their life but routinely fail to see outside their perspective."
Don't I know it!! I never stop being surprised by how people who are so clued up scientifically during their main job and education who for no obvious reason simply turn their training off when dealing with social science topics and even topics in science outside of their immediate expertise and knowledge.
I have a tips request.
While I can forecast reasonably well I think, my big question I've been thinking about is beyond increasing my score on this website, what tips do people here have on using forecasting to add value to ones life?
I think forecasting adds value by forcing us to focus on accountability and keeping ourselves humble. With increasing polarization it's easy and common to make wrong and overconfident predictions, so by taking a step back and closely tracking our predictions and how accurate they were we can be a force in the opposite direction. I think that accountability is valuable in our personal lives but it's valuable to the communities we're part of as well.
@Ryan-Beck
Perfectly true. By being forced to live with the consequences of your prediction concerning your Brier you learn the same lesson again and again with each question: avoid only having an opinion. Go for the truth as much as you can. Look at all angles. Look for angles you didn´t consider possible so far. Let that become a habit. Here it might just look like "brier vs. median". In the outside world it´s about developing better judgment for the benefit of your life and your community.
Thinking analytically from all viewpoints and asking questions of what you see without your own biased assumption will prevent you from being disappointed when things don't go your way. You will also stray from emotional decision making based on other peoples personalities.
@yagudin
Thanks for your post. I think this is a very good contribution.
Effectively, it continues the theme that forecasting is more of a self-improvement tool, but your link provided a number of concrete ways to incorporate this self-improvement into ones life to make better decisions.
However, by focussing on the self-improvement angle (as per most replies) there are 2 issues. 1) Once the skills have been aquired and we know pretty much how to make a good decision, there are diminishing returns from continuing to forecast or becoming better at it. It's kind of binary, you have the skills or you don't. Taking that extra effort to be a grandmaster will not help one's life much better from a self-improvement prospective than the keen amateur. And 2) It suggests the core primary product of our work, the prediction itself, has no obvious use. So it then becomes more like a crossword puzzle than something of real world practical importance.
If we could make more of a case for 2, then we make more of a case for becoming far stronger at this, rather than simply working until you get the gist of most of the skills and then that's pretty much enough in terms of value per unit time invested.
@pstamp
You are right that forecasting requires the construction of a world that once made has a life of itself.
I have been looking at the dizzying heights of that world for many years now, seemingly alone until I found out about this website.
So yes, it's like you first answer a few questions and then the whole world opens up.
Thinking in correlations is also important. For example, I used this to make a legal decision, where my underlying case was strong (95%) and my case for not having to pay costs if I lost was strong (95%) but there was a very high correlation between losing one and losing the other even though both were based on different facts and law because both are tried infront of the same judge. So I chose to not assume I had approx 0% costs risk (1-0.05*0.05) and assumed it was closer to 95% and settle because I couldn't risk paying the enormous costs of being unlucky. As you say, correlations (or more generically co-dependence).
But you then said "Understanding things like that is very useful." When I read your post, I thought this was the intellectual leap.
I agree seeing the world from that place is beautiful, satisfying even. It's certainly intellectually stimulating. But I don't follow it's actual use. That being said, are you suggesting that my decision to settle while others might have proceeded gung ho via a naive 1-0.05*0.05 calculation is the full value, that I might now be financially ruined had I continued while now I have no risk of that and quit at the right time?
This argument even if very well made remains solely the self-improvement angle, and as per my above post, it suffers from 1) diminishing returns per time invested; and 2) the actual forecasts remain of no value.
I suppose, what I am really seeking, is here we all are, having worked one way or another to get these forecasting skills and we can see the wood from the trees from the clouds from the birds from the bats etc. But beyond just helping ourselves in doing better in little and big choices we make each day or decade, can we not achieve anything more meaningful for humanity itself, enhance one's career or perhaps obtain financial gain directly as a result of our forecasting skills?
@sebbysteiny
Concerning substential financial gain as an indirect effect can be achieved by just making better decisions in life as described. Not playing russian roulette is a good idea, even if the chance of losing is "only" 5%, so that might have saved you from serious trouble. I think, you were absolutely right with that decision. With that in mind people wouldn´t marry persons who don´t fit, buy houses they can´t afford ore have other expensive habits, that will cause trouble in a "russian roulette style".
But you wanted a practical example, so her is one. I decided do reduce my consulting and sales work drastically for a few months to avoid any risk of getting my father infected through me and my contacts (he would be a risk patient, so it would have been another case of russian roulette). Why did I do so? Because I´ve been active in Covid-19 questions on GJO since February (reading every available info very early), so I knew the risks. When everybody else was still partying on Carnival my closer contacts had food for a month of potential quarantine in their homes and reduced their contacts to a minimum - because I knew, what consequences this could have. I had the right information before most others and applied it to my personal life. That was a direct positive consequence of this.
About direct financial aspects: if you want to make money directly by forecasting I consider any kind of betting no very good idea. Even the stock market is not as rational and predictable as it should be, so that´s quite risky. You might find a job, where it´s other people´s money you play with, but I don´t think there are many legal ways to do so, unless you do it for banks perhaps.
Last version and perhaps the best for you: selling forecasting knowhow is something at least two people in my homecountry do full time for a living. Both started their careers here, so I´m quite sure, that they still work for Good Judgment Inc.. I don´t know, how well it´s paid, but with all the other learning benefits involved there wouldn´t be almost 200 superbright and capable people working as professional Supers for them, if it wasn´t worth it. Just give it a try to find out yourself. It´s hard (I´ve been working on it without being selected so far), but the harder a way, the more there is to be learned on it...
Those two mentioned also sell forecasting knowhow to companies, government agencies, holding workshops (I don´t know, if they also do so directly or over GJ Inc only) and writing books about it (not my cup of tea), but of course you need a superb track record to have enough knowlege and credibility for that. Guess, where you can get that. Yes, here. And yes, both suggest to start here like they did.
@pstamp
Thanks again for your considered reply. I don't know this superforecasting land very well as I've joined very recently.
In terms of value beyond 1) self-help, you have identified the following:
2) Self-Defence / Insurance vs events in immediate neighbourhood;
3) Betting e.g. on stocks (which you discouraged);
4) Teaching forecasting.
and I add a new use:
5) Political activism (direct use of forecasting)
By understing how politics unfolds, it enables identification of the parts that really matter towards determining the outcome. This empowers you to use your voluntary resources (time and possibly money) to maximum effect. You won't guarantee winning of course, but have done more for your cause than the same resources without forecasting.
4) Teaching (indirect - self-help)
I was aware teaching happened (after googling superforecasting jobs) but I wasn't aware it was a big thing. And it seems the competition is fierce, and probably the only route to this approach is large amounts of work trying to get your Brier score up on this website. As a late starter, I'm well behind the game relative to others on that. In my case, I already have a well paid quant job so it's hard to see a move to superforecasting teaching as value per unit time invested in my case.
2) Self-Defence / Insurance (direct use of forecasting)
This was in fact the main reason I started doing it, and probably still do. When things go bad, they can go very bad. By forecasting Covid 19, it may have helped me be on top of the curve at least sometimes. But there are events in stable democracies that can cause it to become destablised posing a clear and present danager to my life and that of my family and friends. By seeing this early, I might be able to e.g. get out in time. Will such an event happen "on my watch"? Don't know. That is definitely value in my view and can payoff bigtime (by saving my life), but it is unlikely to pay out much over my life time. So is it possible to get more value on a more regular basis?
3) Betting / investing (direct use of forecasting)
It is my view that every skill one has has value that as proof of value can be monitised. If you spend ages learning magic for fun, you could sell it as entertainment at weddings and parties, or use it on dates or at work to obtain social value (with a similar effect as expensive shoes). I am passionate about wine tasting. I get value by buying better wine for less at supermarkets, wine shops and restaurants and by buying cheap at vineyards in certain places while on holiday, along with helping my friends do the same. In office parties, I often help with the wine thereby giving me a career boost and I have found wine "chat" also has some social value.
So I come to forecasting. Where is the value in that? The obvious answer to me is if you can see how things work out better than others, then you should be able to make money from it.
I would never invest in a single share because I don't see how I could get enough data to beat those who do this full time even if they are not as good at forecasting as me.
Funds don't work either, because all I can do is do sensible investing stuff like diversify as much as possible and not sell a holding just because it went down.
Which leaves me with direct betting on ones forecasts. The key here would be to determine how much of what you have seen has been priced in and only bet where you know there is a big difference between your forcecast and the current odds. For betting in betfair, this is easy to do. But you can bet via proxies too (e.g. if there is a Brexit deal, GBP will go up) but you are taking a risk not just on your original forecast but your secondary forecast that your proxy will behave how you hope in that event.
You say "I consider any kind of bettin no very good idea". Now I thought this too for years, but then I came accross the real financial risk called "loss of opportunity risk" (https://en.wikipedia.org/wiki/Opportunity_cost), which essentially means by not investing you are still taking a risk, namely risking missing out on the benefit.
There is nothing special about 0 in terms of P&L. If you know Macron will lose the next election but don't bet on it, you have lost out to the same extent than if you bet on Macron and lost.
So the optimal is not to do nothing or you get sunk by loss of opportunity risk.
On the other hand, putting all your money on Belarous' president falling is also a bad idea.
To make the point further, remember buying insurance is essentially just a bet. But not buying insurance is also a bet, albeit with better odds but higher possible downside.
The goldilocks zone here is balanced investment. If you have a pension or ISA or other form of sensible savings, you should have no doubt heard about the need to diversify it as much as possible. In which case, consider your superforecasting bets as simply another pot. So if your investment portfolio is $100 in US, $100 in EU, $50 in UK, $100 in China, $100 in rest of Asia, $100 in Latin America etc [DISCLAIMER: THIS IS NOT INVESTMENT ADVICE, JUST AN ILLUSTRATIVE EXAMPLE OF WHAT DIVERSIFICATION MEANS, FOR INVESTMENT ADVICE, SEE AN INDEPENDENT FINANCIAL ADVISOR], why not then create another £100 "super forecasting" pot to bet with?
You then want to do as many small bets on different forecasts as you can limited by your pot size, focusing only when the betting markets differ significantly from your forecast.
The key here is to only bet on things you have already forecast e.g. on this website so there is no additional time taken, you are just monitising what you have already done . It may also make you more determined to get your forecast right if you know it will likely translate into your personal investment in your talents.
I haven't done this. This is just for a discussion of ideas. If I were to do it for real, I would probably do something similar to that, unless of course you say something that makes me think this is a dumb idea.
@sebbysteiny
This really seems to be a good summary of "what to take from it beyond making better decisions" (while I still consider that being the key point).
Very little time today, so just short comments: yes, understanding political develompents of course helps a lot with investment. And yes, of course you´re right about "the biggest risk is not taking risks" in that field. On the long run stocks (the right ones) have always been the best investment. And not doing anything guarantees you some 2% negative profit (inflation), so you need to do it, and when you do it, it´s helpful to see trends others don´t see yet to not slip away chance you could have taken to your advantage. With closer analysis you will find, that many questions here are small pieces in a big puzzle being very much connected to the big picture. Most are definitely not randomly chosen. From thinking about, why US Intelligence Agencies or companies might be interested in certain things you can learn a lot about the world.
You might also find this page interesting:
https://goodjudgment.com/
This is a great thread of comments. Thank you for starting this @sebbysteiny
It is an interesting topic, the search for value.
From my perspective as a chemist, honing my forecasts has helped me discern the signal to noise ratio in aspects of many topics I had no formal expertise in (politics, finance etc.). At the beginning it was a bit frustrating how entangled this two are in public/social topics but I came to appreciate the presence of this ratio and its ambiguity. It's part of the "wider picture of the world" that @pstamp mentions and accepting there will always be "unknown unknowns" (which we shouldn't replace with "biased knowns").
Understanding features of this picture and parts of social and economic dynamics is very valuable and it translates to less time spent on over-biased debates and decisions. I have many friends who have a very scientific approach in their life but routinely fail to see outside their perspective.
Thanks Averroes for your contribution.
Reading your post carefully, it seems you are adding more into the 1) self-help category because by improving your understanding about the world, you make better quicker decisions. Is this a fair summary?
"I have many friends who have a very scientific approach in their life but routinely fail to see outside their perspective."
Don't I know it!! I never stop being surprised by how people who are so clued up scientifically during their main job and education who for no obvious reason simply turn their training off when dealing with social science topics and even topics in science outside of their immediate expertise and knowledge.