3 ms·

Disagree with basically everything written here. The neural net is a well-researched cnn variant - the activations are quite interpretable these days. The met

Disagree with basically everything written here. The neural net is a well-researched cnn variant - the activations are quite interpretable these days. The metric’s interpretability is independent from the neural net used for gameplay, so the black box comment doesn’t make sense (unless I’m misunderstanding the comment...). The metric is also very human friendly: think of it as 30% of players in the given position, at the given game’s stage and state, successfully continued on to a victory.

ZephyrBlu 6y ago
> The metric is also very human friendly: think of it as 30% of players in the given position, at the given game’s stage and state, successfully continued on to a victory What I am trying to say is that this is a meaningless piece of information, because that's not how games work. You can't freeze the game state and say "I have a xx% chance to win here", because the state of the game continues evolving over time. If you were actually estimating your chance to win, it would be a forecast. Except if it's anything like this kind of NN [0], then it is literally just estimating based on a snapshot of game state and not the context of the game. At best, this "win probability" is snapshot estimate of the current game state and has nothing to do with the actual outcome of the game. This value is also impossible to falsify, so it could be spitting out any random number and you can't say that it's wrong. [0] https://medium.com/analytics-vidhya/a-simple-neural-network-for-in-game-win-probability-modeling-of-ncaa-basketball-games-58ab4e3ca0f9 https://medium.com/analytics-vidhya/a-simple-neural-network-...
- anonymoushn 6y ago
  What's "the context of the game" and why can't it be included in the encoding of the game state?
  - ZephyrBlu 6y ago
    Context of the game = previous events that occurred. Don't you think it's a bit odd to be forecasting a "win probability" value without taking the events of the game into account? > why can't it be included in the encoding of the game state? How do you encode time series or event data into singular values?
    - anonymoushn 6y ago
      > Don't you think it's a bit odd to be forecasting a "win probability" value without taking the events of the game into account? No. To the extent that these events have no impact on the rules or the outcome of the game beyond the current game state, they don't influence how much equity a player has in the game. You may be interested in "The probability that I win this game given my beliefs about the tendencies of my opponent and my own tendencies, not given optimal play." That's a fine thing to be interested in, but there are other things that can reasonably be called "win probability." > How do you encode time series or event data into singular values? I don't know, how do you encode a sequence of 1-byte values into a 1kb string of text? You can read the alphazero paper if you want to learn how a sequence of go game states were encoded.
      - ZephyrBlu 6y ago
        > To the extent that these events have no impact on the rules or the outcome of the game beyond the current game state, they don't influence how much equity a player has in the game This is a really odd statement. Past events can most definitely affect events beyond the current game state. I think the AlphaStar paper [0] has a lot of good examples of my overall points. Here are a few excerpts: "Central to AlphaStar is a policy [formula], represented by a neural network with parameters θ that receives all observations [formula] from the start of the game as inputs, and selects actions as outputs" "To manage the structured, combinatorial action space, the agent uses an auto-regressive policy7,10,11 and recurrent pointer network" I.e. the policy does account for past actions, and these are not encoded in any way. "The agent parameters were subsequently trained by a reinforcement learning algorithm that is designed to maximize the win rate (that is, compute a best response) against a mixture of opponents" I.e. the winrate value is _not_ a "win probability", it's a maximization like I said in my original comment. Do these hold true for discrete action games and/or perfect information games? I don't know, but I find it likely they hold for at least discrete, imperfect information games. [0] https://www.nature.com/articles/s41586-019-1724-z.epdf?author_access_token=lZH3nqPYtWJXfDA10W0CNNRgN0jAjWel9jnR3ZoTv0PSZcPzJFGNAZhOlk4deBCKzKm70KfinloafEF1bCCXL6IIHHgKaDkaTkBcTEv7aT-wqDoG1VeO9-wO3GEoAMF9bAOt7mJ0RWQnRVMbyfgH9A%3D%3D https://www.nature.com/articles/s41586-019-1724-z.epdf?autho...

ZephyrBlu 6y ago

> The metric is also very human friendly: think of it as 30% of players in the given position, at the given game’s stage and state, successfully continued on to a victory What I am trying to say is that this is a meaningless piece of information, because that's not how games work. You can't freeze the game state and say "I have a xx% chance to win here", because the state of the game continues evolving over time. If you were actually estimating your chance to win, it would be a forecast. Except if it's anything like this kind of NN [0], then it is literally just estimating based on a snapshot of game state and not the context of the game. At best, this "win probability" is snapshot estimate of the current game state and has nothing to do with the actual outcome of the game. This value is also impossible to falsify, so it could be spitting out any random number and you can't say that it's wrong. [0] https://medium.com/analytics-vidhya/a-simple-neural-network-for-in-game-win-probability-modeling-of-ncaa-basketball-games-58ab4e3ca0f9 https://medium.com/analytics-vidhya/a-simple-neural-network-...

anonymoushn 6y ago
What's "the context of the game" and why can't it be included in the encoding of the game state?
- ZephyrBlu 6y ago
  Context of the game = previous events that occurred. Don't you think it's a bit odd to be forecasting a "win probability" value without taking the events of the game into account? > why can't it be included in the encoding of the game state? How do you encode time series or event data into singular values?
  - anonymoushn 6y ago
    > Don't you think it's a bit odd to be forecasting a "win probability" value without taking the events of the game into account? No. To the extent that these events have no impact on the rules or the outcome of the game beyond the current game state, they don't influence how much equity a player has in the game. You may be interested in "The probability that I win this game given my beliefs about the tendencies of my opponent and my own tendencies, not given optimal play." That's a fine thing to be interested in, but there are other things that can reasonably be called "win probability." > How do you encode time series or event data into singular values? I don't know, how do you encode a sequence of 1-byte values into a 1kb string of text? You can read the alphazero paper if you want to learn how a sequence of go game states were encoded.
    - ZephyrBlu 6y ago
      > To the extent that these events have no impact on the rules or the outcome of the game beyond the current game state, they don't influence how much equity a player has in the game This is a really odd statement. Past events can most definitely affect events beyond the current game state. I think the AlphaStar paper [0] has a lot of good examples of my overall points. Here are a few excerpts: "Central to AlphaStar is a policy [formula], represented by a neural network with parameters θ that receives all observations [formula] from the start of the game as inputs, and selects actions as outputs" "To manage the structured, combinatorial action space, the agent uses an auto-regressive policy7,10,11 and recurrent pointer network" I.e. the policy does account for past actions, and these are not encoded in any way. "The agent parameters were subsequently trained by a reinforcement learning algorithm that is designed to maximize the win rate (that is, compute a best response) against a mixture of opponents" I.e. the winrate value is _not_ a "win probability", it's a maximization like I said in my original comment. Do these hold true for discrete action games and/or perfect information games? I don't know, but I find it likely they hold for at least discrete, imperfect information games. [0] https://www.nature.com/articles/s41586-019-1724-z.epdf?author_access_token=lZH3nqPYtWJXfDA10W0CNNRgN0jAjWel9jnR3ZoTv0PSZcPzJFGNAZhOlk4deBCKzKm70KfinloafEF1bCCXL6IIHHgKaDkaTkBcTEv7aT-wqDoG1VeO9-wO3GEoAMF9bAOt7mJ0RWQnRVMbyfgH9A%3D%3D https://www.nature.com/articles/s41586-019-1724-z.epdf?autho...