Metaculus Help: Spread the word
If you like Metaculus, tell your friends! Share this question via Facebook, Twitter, or Reddit.
When will a robot poker player win a significant live (not online) competition?
Poker is a challenging game of bets, raises, and re-raises, calculation of odds and expected payoffs, game-theoretic mixed strategies, and tradeoffs between unexploitable and exploitative play.
Poker is also a challenging game of visually identifying and then gripping small plastic discs and rectangles, and discriminating them between them based on their colors and symbols printed on their surfaces. These discs and rectangles must be moved around on a table in accordance with a game ruleset that is partially given in advance, and partly determined by verbal instruction from dealer and floor coordinator.
An example of the complexity of object manipulation that is required: the rectangles belonging to the player must be picked up in such a way that the symbols on the front can be inspected by the player, but kept oriented so that no other player can inspect those symbols, until such time as the ruleset dictates that the cards and symbols be revealed. Violating this constraint can lead not only to substandard play that is exploited by opponents, but also to penalties and even potential disqualification.
And so we ask: when will a computer or robotic player become skilled enough at both the theoretical game and the physical game of poker to win a significant Texas HoldEm competition against humans in live play?
Definition of “significant competition”:
A Texas HoldEm tournament or cash game in which:
- more than 300 hands are played
- robotic and human players obey the same rules, without special accommodations for the perceptual or motor skills of the robotic player (including complying with verbal direction from the dealer and staff)
- competitors include at least one human member of the top 200 of the Hendon Mob All Time Money List (at the time of play)
- the winner (best-outcome player) is awarded more than $10,000 US (whether that money is awarded to the robot itself or to its owners or developers), with the same payouts made to robotic and human players.
There are a wide variety of ways in which a competition like this could come about, and the intent is to cast a wide net. Two contrasting examples, which would both count for positive resolution:
- A “Deep Blue vs. Kasparov”-style challenge, in which the whole point of the competition is to pit human against machine.
- Harold takes a week off from his middle-management desk job, walks into the Rio All-Suites Hotel in Las Vegas, and plunks down $10,000 in cash for a seat in the World Series of Poker Main Event, because (since the recent passage of Robotic Personhood legislation) no one can tell him that he can’t, and after all it’s his money.
Clarifications and fine print:
- The robot player must finish #1 in the event. In the case of a cash game, this means netting the most money by the end. In the case of a tournament, this means playing until a robot is the only one left.
- The win must be for the entire event as scheduled. For example, if a televised cash-game event is scheduled to span multiple days, a win over the course of a single day does not suffice. If the event is a single-elimination heads-up tournament, it does not suffice to win the semi-final game, even if that guarantees a purse greater than $10,000 US.
- The robot need not literally sit in a chair at the poker table, but needs to be at the table taking up approximately the same portion of the perimeter that a human player would be allowed.
The rules must allow for the kinds of sensory modalities that would normally be legal for human players, but not for perceptual abilities that would correspond to cheating for humans. Examples:
- Scanning cards in a way that allows reading through the back to see symbols on the front – disallowed.
- Scanning the backs of cards to record their imperfections and distinguish between them in the future (i.e. “Every card is marked.”) - disallowed.
- Scanning the faces of opponents to determine their emotional state – fair game.
Metaculus help: Predicting
Predictions are the heart of Metaculus. Predicting is how you contribute to the wisdom of the crowd, and how you earn points and build up your personal Metaculus track record.
The basics of predicting are very simple: move the slider to best match the likelihood of the outcome, and click predict. You can predict as often as you want, and you're encouraged to change your mind when new information becomes available. With tachyons you'll even be able to go back in time and backdate your prediction to maximize your points.
The displayed score is split into current points and total points. Current points show how much your prediction is worth now, whereas total points show the combined worth of all of your predictions over the lifetime of the question. The scoring details are available on the FAQ.
Note: this question resolved before its original close time. All of your predictions came after the resolution, so you did not gain (or lose) any points for it.
Note: this question resolved before its original close time. You earned points up until the question resolution, but not afterwards.
This question is not yet open for predictions.
Metaculus help: Community Stats
Use the community stats to get a better sense of the community consensus (or lack thereof) for this question. Sometimes people have wildly different ideas about the likely outcomes, and sometimes people are in close agreement. There are even times when the community seems very certain of uncertainty, like when everyone agrees that event is only 50% likely to happen.
When you make a prediction, check the community stats to see where you land. If your prediction is an outlier, might there be something you're overlooking that others have seen? Or do you have special insight that others are lacking? Either way, it might be a good idea to join the discussion in the comments.