I would be interested in a program that allowed playing poker with fake money against bots, in a way that collects statistics and grades the quality of your play. Playing online poker for real money against real humans is not a precedent I want to set, even if it is a valuable training exercise.
I wish we could do this but playing without real money wouldn’t work for improving rationality and I don’t think it would even work for testing rationality in a meaningful way.
Also, I think this sets a good precedent. People who will bet on cards are far more likely to bet on their beliefs in general, which is a sign that they actually have beliefs as opposed to just “things they would like to be true”. Not taking +EV bets that represent negligible portions of your net-worth is epistemically irrational so it seems like a rationalist community should be encouraging betting when it is done at reasonable stakes as a way to train or test beliefs.
Not taking +EV bets that represent negligible portions of your net-worth is epistemically irrational so it seems like a rationalist community should be encouraging betting when it is done at reasonable stakes as a way to train or test beliefs.
After some amount of practice, playing poker should be +EV. Betting money while still a poker noob, however, is certainly -EV. I’d be more willing to bet money if I had some indication, preferrably in the form of an Elo rating, that I can expect to come out ahead.
I’d be more willing to bet money if I had some indication, preferrably in the form of an Elo rating, that I can expect to come out ahead.
After we cover some introductory material on our site, I plan to recommend people test out play money games after signing up to learn how the software works and get a feel for the game, then take these twotests. I’ll try and get some scores from winning, marginal, and losing players so I can make a recommendation for a cut-off on whether to play or keep studying first.
The a priori argument that using money is important doesn’t stand up under closer examination. If you are incapable of generalizing from in-game currency to dollars, you won’t be capable of generalizing from poker to other activities. And player behavior does not seem to be grossly different—take for example the fact that prediction markets work the same with real money or fake money.
I’ve played poker for play money. There’s not as much pain when losing so there is correspondingly less motivation to exercise discipline, come to correct beliefs, or deal seriously with the difficult psychological situations. Learning to take those things seriously and perform under “survival instinct”-level pressure is the point of this training. You want the actions you take to imprint your habits and reshape them to be more rational over time and play money can’t do that. See the post I linked to above for a longer explanation.
The football study you link to shows me that in cases where there’s already a rich information market, real and pretend markets both mirror it well. That doesn’t indicate to me that people wagering in the play money markets had a similar visceral experience or would learn the same lessons from failure or success. I strongly suspect they didn’t—and would bet money on it if I could figure out how.
Playing poker for play chips online quickly teaches that there’s a huge difference between $0.01 and $0.00 when it comes to quality of play. Games that don’t require something be risked end up being jokes.
I think there’s a place for it—A few months ago I didn’t know anything about poker except the relative values of the hands. Then I got a free iPhone Texas Hold ’em app that has a few nice features, like a colored bar showing the current strength of your hand (although it doesn’t seem to adjust for the number of opponents). I’ve picked up a lot more intuition, at the very least.
The a priori argument that using money is important doesn’t stand up under closer >examination. If you are incapable of generalizing from in-game currency to dollars, you >won’t be capable of generalizing from poker to other activities. And player behavior >does not seem to be grossly different—take for example the fact that prediction markets >work the same with real money or fake money.
It may not be much different for prediction markets, but it is VERY different for online poker. Even if you play exactly the same with or without money, your opponents will not, and therefore you will be “training” on different data than you think. This applies especially to NL games; risking 1,000 points on one bet is a lot different than risking 1,000 dollars.
This is an excellent idea. The ability to create a simple measure (profit/loss) for a complex subject obviously loses something in the process, but makes up for it by allowing reproducible experimentation with a robust and unforgiving metric. If we claim something is valuable, then we should be prepared to find simple metrics capable of measuring it. I think you have done so.
I would be interested in a program that allowed playing poker with fake money against bots, in a way that collects statistics and grades the quality of your play. Playing online poker for real money against real humans is not a precedent I want to set, even if it is a valuable training exercise.
I wish we could do this but playing without real money wouldn’t work for improving rationality and I don’t think it would even work for testing rationality in a meaningful way.
Also, I think this sets a good precedent. People who will bet on cards are far more likely to bet on their beliefs in general, which is a sign that they actually have beliefs as opposed to just “things they would like to be true”. Not taking +EV bets that represent negligible portions of your net-worth is epistemically irrational so it seems like a rationalist community should be encouraging betting when it is done at reasonable stakes as a way to train or test beliefs.
After some amount of practice, playing poker should be +EV. Betting money while still a poker noob, however, is certainly -EV. I’d be more willing to bet money if I had some indication, preferrably in the form of an Elo rating, that I can expect to come out ahead.
After we cover some introductory material on our site, I plan to recommend people test out play money games after signing up to learn how the software works and get a feel for the game, then take these two tests. I’ll try and get some scores from winning, marginal, and losing players so I can make a recommendation for a cut-off on whether to play or keep studying first.
The a priori argument that using money is important doesn’t stand up under closer examination. If you are incapable of generalizing from in-game currency to dollars, you won’t be capable of generalizing from poker to other activities. And player behavior does not seem to be grossly different—take for example the fact that prediction markets work the same with real money or fake money.
I’ve played poker for play money. There’s not as much pain when losing so there is correspondingly less motivation to exercise discipline, come to correct beliefs, or deal seriously with the difficult psychological situations. Learning to take those things seriously and perform under “survival instinct”-level pressure is the point of this training. You want the actions you take to imprint your habits and reshape them to be more rational over time and play money can’t do that. See the post I linked to above for a longer explanation.
The football study you link to shows me that in cases where there’s already a rich information market, real and pretend markets both mirror it well. That doesn’t indicate to me that people wagering in the play money markets had a similar visceral experience or would learn the same lessons from failure or success. I strongly suspect they didn’t—and would bet money on it if I could figure out how.
Playing poker for play chips online quickly teaches that there’s a huge difference between $0.01 and $0.00 when it comes to quality of play. Games that don’t require something be risked end up being jokes.
I think there’s a place for it—A few months ago I didn’t know anything about poker except the relative values of the hands. Then I got a free iPhone Texas Hold ’em app that has a few nice features, like a colored bar showing the current strength of your hand (although it doesn’t seem to adjust for the number of opponents). I’ve picked up a lot more intuition, at the very least.
It may not be much different for prediction markets, but it is VERY different for online poker. Even if you play exactly the same with or without money, your opponents will not, and therefore you will be “training” on different data than you think. This applies especially to NL games; risking 1,000 points on one bet is a lot different than risking 1,000 dollars.
I have no idea whether you could set up a feasible system for this, but perhaps you could have people bet karma points instead of money.
This is an excellent idea. The ability to create a simple measure (profit/loss) for a complex subject obviously loses something in the process, but makes up for it by allowing reproducible experimentation with a robust and unforgiving metric. If we claim something is valuable, then we should be prepared to find simple metrics capable of measuring it. I think you have done so.
It might be worthwhile to write a post on this for RationalPoker.com.
I thought I elaborated on my point in the link to the RationalPoker.com post I gave in my comment.
D’oh! Nevermind, The Stupid got me.