3 ms·

I think everyone would seriously benefit from learning poker. I used to play professionally and the idea of looking at things as probabilistic bets, and in term

I think everyone would seriously benefit from learning poker. I used to play professionally and the idea of looking at things as probabilistic bets, and in terms of expected value is so deeply rooted in my mind Investments are bets. Most sane investors aren't putting it all on one thing Startups are bets Applying for jobs. Sales. Dating. Health. Basically everything You risk $X money and time for a payoff of $Y that comes Z% You can make the best decision and have a bad outcome because there are so many unknowns. This isn't chess You can play everything wrong and still hit it out of the park I mean this is one of the range of outcomes that could've happened. You can't declare yourself a success or failure from one project Just keep making good decisions and don't risk it all, and you'll more than likely end up fine

teiferer 9mo ago
> You risk $X money and time for a payoff of $Y that comes Z% Unlike in poker where you know those 3 numbers if you are paying attention, in real life you know only X, and sometimes not even that. The rest is a guessing game, and that estimation is the hard part.
- foltik 9mo ago
  In poker you learn to strategically size $X based on your estimates of $Y, Z%, and current bankroll. Even if your estimates have large error bars, it can still be a good bet if you believe it’s +EV and have a large enough bankroll.
altmanaltman 9mo ago
I read about mathematical expectation in a poker book for the first time a long time ago and it's really an interesting way to think about the world. In real life, it's a bit different though since there are other factors than just raw percentages like poker hands. For example, you could do everything right and still fail while someone else (like a nepo hire) can do everything wrong and still succeed. Life's odds are rigged.
- diab0lic 9mo ago
  I think the metaphor is pretty extensible to handle the things you point out. > you could do everything right and still fail Sometimes even a high probability draw doesn't get there on the river. > while someone else (like a nepo hire) can do everything wrong and still succeed Some people always start with pairs, Axs, or suited connectors.
  - altmanaltman 9mo ago
    No, that's not what I meant. I meant more like you could flop a high probability draw that does get there on the river, but the nepo kid still wins even though they just had one pair. In poker, there are rules everyone has to follow. In real life, the rules can differ from person to person. Hence, the odds are rigged.
mhogers 9mo ago
One of my favorite star trek scenes: https://www.youtube.com/watch?v=t4A-Ml8YHyM https://www.youtube.com/watch?v=t4A-Ml8YHyM
PlatoIsADisease 9mo ago
If you want me to ruin this for you, look up Bertrand's Paradox and The Problem Of Priors. I recently had to deal with this, ugh, I just pick a prior or two and see what the outcome looks like. I'm not sure we can calculate probability of success, but rather use probability to find losing ideas that should never be done.
- rramadass 9mo ago
  Nice. What are some good books to build this sort of perspective on Probability/Randomness/Chance?
  - PlatoIsADisease 9mo ago
    These are ~1 layer removed, but it shook me from an epistemological perspective. Basically it destroys the idea of metaphysical truth. I'll start with the easy one that was written in native English, William James's Pragmatism. Its only 4 hours long, and teaches 1 of the 3 branches of metaphilosophy(Pragmatic). For ROI in philosophy, it doesn't get better than this. It teaches one of the 3 theories of Truth. Next is significantly more difficult, but everyone only understands 10% of Wittgenstein. Read both early and late Wittgenstein. If I remember correctly, its ~12 hours of pain as you will barely understand him. However, it shakes foundations of math and science as you will find these are mere constructs of language. This will also teach you 1 of the 3 branches of metaphilosophy(Analytical), and teach you how language works. You really need to read both early and late. But I imagine reading early Wittgenstein will convince you to read late. Its super painful, but you should deal with the pain. (the remaining branch of philosophy is Continental... that takes a long time to learn, but its mostly nonsensical rebuttals to Plato, take it from someone who read this for 8 years...) Regarding that initial thought on the Problem of Priors, here is a quick youtube video: https://www.youtube.com/watch?v=xy6xXEhbGa0 https://www.youtube.com/watch?v=xy6xXEhbGa0
    - rramadass 9mo ago
      William James i have read, both "Pragmatism"(not completely though) and "Talks to Teachers" (my favourite) but not much of Wittgenstein. There is a lot here which jives with ancient Hindu Schools of Philosophy, specifically; Nyaya (https://en.wikipedia.org/wiki/Nyaya https://en.wikipedia.org/wiki/Nyaya), Vaisheshika (https://en.wikipedia.org/wiki/Vaisheshika https://en.wikipedia.org/wiki/Vaisheshika) and Charvaka (https://en.wikipedia.org/wiki/Charvaka https://en.wikipedia.org/wiki/Charvaka). Bayesian epistemology with priors have always been problematic since if there were no constraints on prior beliefs than anything goes. But in order to constrain the priors you need to know the distribution to calculate probabilities.
rramadass 9mo ago
See also The Book of Risk by Dan Borge.
ericdykstra 9mo ago
I studied poker a lot in high school, and it still influences how I think today. Probabilistic thinking about the EV of your decisions is a good framework, but "Just keep making good decisions" is the hard part. Same as in poker, though, the hard part about making a decision in life isn't the middle-school level probability math, but about making the right estimates about payoff potential and success rate based on incomplete information. Analyzing things as they are rather than how you wish they were, being able to separate useful information from noise, and taking a step back to look at second-order effects are all useful skills that will help you make better decisions the more you develop them.
neltnerb 9mo ago
It gets messy when the other poker players are your mentors and family whom you love and respect, when the money is real, when the amounts to you feel enormous (whether they matter to the investor is a different story), and when losing carries unpredictable emotional and social consequences. And when mentors and family do it you are probably their only bet. It feels like betting other people's money and I totally get how that's a psychological hangup to feel that it is wrong. I am not being sarcastic or flippant in this, it limits me to not be willing to take advantage of the risk/reward system, but it's somehow deeply tied into my ethics in ways I didn't even really notice until my second startup. And I'm okay with that limitation, but I'll work through it sooner or later and regain my appetite for risk. Aside from that, many people despise thinking this way about their work. Like me, even though I'm willing to do it. I've met founders who loved negotiation. Like I swear they must have been nightmare children. I hate negotiation, I want people to treat each other fairly and will stick with a good human that does so forever. I suspect those people don't mind ripping off family and friends, much less strangers. It is harsh, and I've seen things, but anecdotes aren't statistics, and that's a huge obvious red flag to me. If the leaders like to do internal negotiation games I just want out. It's just painful to have to deal with being around when you just want to make stuff work. It's important to talk about ahead of time, and get in writing, but that's also something easily overlooked among family and friends who do not do this kind of thing much. And the downsides need to be honest and clear to avoid mistrust. Now, after working at two startups and starting three, my third is a consulting company that exclusively does things hourly except in very special circumstances. It's never going to grow fast because it scales as my hours of labor. But I am way more able to pick a price for my labor than decide how to price something that I haven't done yet. I could figure it out, I just kinda don't want to. For now?

teiferer 9mo ago

> You risk $X money and time for a payoff of $Y that comes Z% Unlike in poker where you know those 3 numbers if you are paying attention, in real life you know only X, and sometimes not even that. The rest is a guessing game, and that estimation is the hard part.

foltik 9mo ago
In poker you learn to strategically size $X based on your estimates of $Y, Z%, and current bankroll. Even if your estimates have large error bars, it can still be a good bet if you believe it’s +EV and have a large enough bankroll.

altmanaltman 9mo ago

I read about mathematical expectation in a poker book for the first time a long time ago and it's really an interesting way to think about the world. In real life, it's a bit different though since there are other factors than just raw percentages like poker hands. For example, you could do everything right and still fail while someone else (like a nepo hire) can do everything wrong and still succeed. Life's odds are rigged.

diab0lic 9mo ago
I think the metaphor is pretty extensible to handle the things you point out. > you could do everything right and still fail Sometimes even a high probability draw doesn't get there on the river. > while someone else (like a nepo hire) can do everything wrong and still succeed Some people always start with pairs, Axs, or suited connectors.
- altmanaltman 9mo ago
  No, that's not what I meant. I meant more like you could flop a high probability draw that does get there on the river, but the nepo kid still wins even though they just had one pair. In poker, there are rules everyone has to follow. In real life, the rules can differ from person to person. Hence, the odds are rigged.

mhogers 9mo ago

One of my favorite star trek scenes: https://www.youtube.com/watch?v=t4A-Ml8YHyM https://www.youtube.com/watch?v=t4A-Ml8YHyM

PlatoIsADisease 9mo ago

If you want me to ruin this for you, look up Bertrand's Paradox and The Problem Of Priors. I recently had to deal with this, ugh, I just pick a prior or two and see what the outcome looks like. I'm not sure we can calculate probability of success, but rather use probability to find losing ideas that should never be done.

rramadass 9mo ago
Nice. What are some good books to build this sort of perspective on Probability/Randomness/Chance?
- PlatoIsADisease 9mo ago
  These are ~1 layer removed, but it shook me from an epistemological perspective. Basically it destroys the idea of metaphysical truth. I'll start with the easy one that was written in native English, William James's Pragmatism. Its only 4 hours long, and teaches 1 of the 3 branches of metaphilosophy(Pragmatic). For ROI in philosophy, it doesn't get better than this. It teaches one of the 3 theories of Truth. Next is significantly more difficult, but everyone only understands 10% of Wittgenstein. Read both early and late Wittgenstein. If I remember correctly, its ~12 hours of pain as you will barely understand him. However, it shakes foundations of math and science as you will find these are mere constructs of language. This will also teach you 1 of the 3 branches of metaphilosophy(Analytical), and teach you how language works. You really need to read both early and late. But I imagine reading early Wittgenstein will convince you to read late. Its super painful, but you should deal with the pain. (the remaining branch of philosophy is Continental... that takes a long time to learn, but its mostly nonsensical rebuttals to Plato, take it from someone who read this for 8 years...) Regarding that initial thought on the Problem of Priors, here is a quick youtube video: https://www.youtube.com/watch?v=xy6xXEhbGa0 https://www.youtube.com/watch?v=xy6xXEhbGa0
  - rramadass 9mo ago
    William James i have read, both "Pragmatism"(not completely though) and "Talks to Teachers" (my favourite) but not much of Wittgenstein. There is a lot here which jives with ancient Hindu Schools of Philosophy, specifically; Nyaya (https://en.wikipedia.org/wiki/Nyaya https://en.wikipedia.org/wiki/Nyaya), Vaisheshika (https://en.wikipedia.org/wiki/Vaisheshika https://en.wikipedia.org/wiki/Vaisheshika) and Charvaka (https://en.wikipedia.org/wiki/Charvaka https://en.wikipedia.org/wiki/Charvaka). Bayesian epistemology with priors have always been problematic since if there were no constraints on prior beliefs than anything goes. But in order to constrain the priors you need to know the distribution to calculate probabilities.

rramadass 9mo ago