Hah, I also had tons of those notebooks full of game designs and variant rules (and code, pixel art in colored pencil on quarter inch graph paper, because we didn't have a computer in the house).
Not for chess though. I never liked the game at all. It didn't take me long to realize that what seemed to be a vast field of possible states was constrained severely by optimal opening and end states, and that the real game was played within a tiny subset of midgame states. Other people had already figured all of this out decades and centuries ago, and so middling players would memorize books full of moves and positions and work through those until they became intuition or else they gave up... And even though that still yielded a huge number of possible games much of it was rote and this offended me somehow.
In a similar sort of way blackjack offends me as a card game: the best way to play is to dispassionately calculate probabilities, or if you're not sharp enough for that, remember some optimal strategies and rules of thumb. But the house has a slight inherent advantage so over a long enough timeline you can only lose slowly by being very good, and the house only has to deploy a standard predictable strategy to whittle away your money. Boooring.
I really found my passion in MtG for a while. Unlike these fixed piece games, you never knew what your opponent was bringing to the table. You'd get a feel for other players and how they liked to play, and you'd learn their decks and adapt, but they'd come back next week with new decks and new strategies tweaked in response to what they'd learned about you. That game was crack and I had to quit because it was destroying me.
Anyway, I do think I might know what you're doing here. I may be wrong. But it sounds like crack.
"It didn't take me long to realize that what seemed to be a vast field of possible states was constrained severely by optimal opening and end states, and that the real game was played within a tiny subset of midgame states."
Yep. That’s kind of what I meant by saying it eats little boys (and a rare few girls), for a game that's only slightly played in the middle, by all but the top GMs. Chess as a sport is a little more intriguing, since it taxes multiple aspects of mind, body, and spirit. OTOH, you could say that about almost any sport played at the top level.
I also find blackjack offensive. It's the ultimate degenerate gambler's pastime, in a way, offering more of an illusion of control than something like slots or craps, while simultaneously eating more of your time.
I found MtG interesting from a design perspective, but never got into playing it. I think I intuited that it might also threaten to eat too much of my time and money, and its capacity for endless expansion threatened eventual incoherence due to market pressures. "Learn these new cards, or you'll fall behind the curve."
> its capacity for endless expansion threatened eventual incoherence due to market pressures
It did, and that's exactly what happened.
Unfortunately it was more due to dumbing down to make the game more accessible to newcomers in the face of all that had come before. Lots of win condition in a box things: "get these three cards in play and you win automatically!" What we used to spend hours and days trying to figure out, they just printed into the rules. Really took the fun out of it.
But we solved that for a while by just refusing to play the newer sets. Which got expensive, because old cards are expensive.
Per your request I hearted this post. I did skim read it, but being so far out of my area of understanding I could do no more. I'm sure anyone, usually males, I'd guess who are proficient at the game "got it" but what can I say? I am a woman - I don't get chess, or even very many games of strategy. True for me but not to say that's true of most women.
I just typed a very long and detailed response to this, then accidentally lost it by fat-fingering a button. I'll try to recreate it when I get home. Until then: thanks.
My sympathies! I've done that more times than I can count! I usually can't make myself re-create the lost masterpiece I'd just composed, but I hope you will try to re-create yours.
I've often speculated that the two modes ( nowadays) of search tend to fail where there are nonrandom hidden elements. Inference works well on subtle patterns ( this is how AlphaZero self trained) but fades to noise in absence of sufficient law. Graph searching shows potential for perfect optimization in the form of massive switch statements, but cannot succeed if there are unprunable cycles.
Yarv in his essay "a reservationist epistemology" states that no Bayesian reasoner can get the right answer if for example when you are pulling balls out of an urn, you keep pulling the ball until you get the one you want. You could in theory convince the opponent that the balls were all red even when they were all but one, blue. A theory of mind is required to detect deception.
Chess and Go are interesting in that they present a potentially infinite search space that can be mapped and therefore grasped in terms of only the game itself, without reference to the player. They are in practice noninfinite, but offer many interesting traps for the unwary, a good practice for strategy if you also already have a theory of mind.
Two things.
One, I suspect your project here is to understand a generalized theory of how to beat strategizing automatons, whether they be graph based, inference based or a mixture of both (as AZ is).
Two, while I cannot know what your Lover's chess is, I am reminded of Hussie's idea in Homestuck, that of allowing the disguise of kings and queens, provided the real piece does not violate it's real move set.
Chess is strange in that it is only interesting to beginners & grandmasters. I think mass printing might be the culprit behind its' demise, without the books allowing players to memorize all the possible moves the game remains more interesting.
I think many of us went through a similar experience. I had the same experience with chess. I'm don't like rote memorization and was not going to spend all of my time on a game that required a significant amount of it. I preferred games which followed the cycle of strategy-execute-randomized outcome-reajust. The type of games that require quick intuitive decisions applied to a changing situation. I became a Squad Leader and then an Advanced Squad Leader fanatic. I don't know if there have been any attempts to train an AI on one of these games, but it would be interesting to know how well they perform.
I want my A/i master buried with me when I pass on. When the Heavens ask me a question in the next life, my A/i master will always have the right answer. Maybe the Heavens might be persuaded not to send me back to earth as a computer chip.
Hah, I also had tons of those notebooks full of game designs and variant rules (and code, pixel art in colored pencil on quarter inch graph paper, because we didn't have a computer in the house).
Not for chess though. I never liked the game at all. It didn't take me long to realize that what seemed to be a vast field of possible states was constrained severely by optimal opening and end states, and that the real game was played within a tiny subset of midgame states. Other people had already figured all of this out decades and centuries ago, and so middling players would memorize books full of moves and positions and work through those until they became intuition or else they gave up... And even though that still yielded a huge number of possible games much of it was rote and this offended me somehow.
In a similar sort of way blackjack offends me as a card game: the best way to play is to dispassionately calculate probabilities, or if you're not sharp enough for that, remember some optimal strategies and rules of thumb. But the house has a slight inherent advantage so over a long enough timeline you can only lose slowly by being very good, and the house only has to deploy a standard predictable strategy to whittle away your money. Boooring.
I really found my passion in MtG for a while. Unlike these fixed piece games, you never knew what your opponent was bringing to the table. You'd get a feel for other players and how they liked to play, and you'd learn their decks and adapt, but they'd come back next week with new decks and new strategies tweaked in response to what they'd learned about you. That game was crack and I had to quit because it was destroying me.
Anyway, I do think I might know what you're doing here. I may be wrong. But it sounds like crack.
"It didn't take me long to realize that what seemed to be a vast field of possible states was constrained severely by optimal opening and end states, and that the real game was played within a tiny subset of midgame states."
Yep. That’s kind of what I meant by saying it eats little boys (and a rare few girls), for a game that's only slightly played in the middle, by all but the top GMs. Chess as a sport is a little more intriguing, since it taxes multiple aspects of mind, body, and spirit. OTOH, you could say that about almost any sport played at the top level.
I also find blackjack offensive. It's the ultimate degenerate gambler's pastime, in a way, offering more of an illusion of control than something like slots or craps, while simultaneously eating more of your time.
I found MtG interesting from a design perspective, but never got into playing it. I think I intuited that it might also threaten to eat too much of my time and money, and its capacity for endless expansion threatened eventual incoherence due to market pressures. "Learn these new cards, or you'll fall behind the curve."
> its capacity for endless expansion threatened eventual incoherence due to market pressures
It did, and that's exactly what happened.
Unfortunately it was more due to dumbing down to make the game more accessible to newcomers in the face of all that had come before. Lots of win condition in a box things: "get these three cards in play and you win automatically!" What we used to spend hours and days trying to figure out, they just printed into the rules. Really took the fun out of it.
But we solved that for a while by just refusing to play the newer sets. Which got expensive, because old cards are expensive.
Per your request I hearted this post. I did skim read it, but being so far out of my area of understanding I could do no more. I'm sure anyone, usually males, I'd guess who are proficient at the game "got it" but what can I say? I am a woman - I don't get chess, or even very many games of strategy. True for me but not to say that's true of most women.
I just typed a very long and detailed response to this, then accidentally lost it by fat-fingering a button. I'll try to recreate it when I get home. Until then: thanks.
My sympathies! I've done that more times than I can count! I usually can't make myself re-create the lost masterpiece I'd just composed, but I hope you will try to re-create yours.
I've often speculated that the two modes ( nowadays) of search tend to fail where there are nonrandom hidden elements. Inference works well on subtle patterns ( this is how AlphaZero self trained) but fades to noise in absence of sufficient law. Graph searching shows potential for perfect optimization in the form of massive switch statements, but cannot succeed if there are unprunable cycles.
Yarv in his essay "a reservationist epistemology" states that no Bayesian reasoner can get the right answer if for example when you are pulling balls out of an urn, you keep pulling the ball until you get the one you want. You could in theory convince the opponent that the balls were all red even when they were all but one, blue. A theory of mind is required to detect deception.
Chess and Go are interesting in that they present a potentially infinite search space that can be mapped and therefore grasped in terms of only the game itself, without reference to the player. They are in practice noninfinite, but offer many interesting traps for the unwary, a good practice for strategy if you also already have a theory of mind.
Two things.
One, I suspect your project here is to understand a generalized theory of how to beat strategizing automatons, whether they be graph based, inference based or a mixture of both (as AZ is).
Two, while I cannot know what your Lover's chess is, I am reminded of Hussie's idea in Homestuck, that of allowing the disguise of kings and queens, provided the real piece does not violate it's real move set.
Good essay!
Chess is strange in that it is only interesting to beginners & grandmasters. I think mass printing might be the culprit behind its' demise, without the books allowing players to memorize all the possible moves the game remains more interesting.
Have you ever actually attempted to test this revised game of yours?
I think many of us went through a similar experience. I had the same experience with chess. I'm don't like rote memorization and was not going to spend all of my time on a game that required a significant amount of it. I preferred games which followed the cycle of strategy-execute-randomized outcome-reajust. The type of games that require quick intuitive decisions applied to a changing situation. I became a Squad Leader and then an Advanced Squad Leader fanatic. I don't know if there have been any attempts to train an AI on one of these games, but it would be interesting to know how well they perform.
I can't figure out the thesis of this essay. I might be a robot?
"camouflage for a great bovine war".... BOOM! pink mist of my blown mind. thank you!
I want my A/i master buried with me when I pass on. When the Heavens ask me a question in the next life, my A/i master will always have the right answer. Maybe the Heavens might be persuaded not to send me back to earth as a computer chip.
C̵̨̖̮̟̣̘̻̗͈̰̈å̷̧̟̹̯̚̕͝ͅn̵̛͉͉̹̜̺̮̾̌́̋̕̚ ̶̡͚͚̯̩̥̻̐̂̽͜y̶͙̓͋o̷̱̯̫͕̣͊̉ư̸̘͗͌͗̒͆̅̈́̽ ̵̤̣̼̩͉̈́̃̓͠͠r̵̛̙̰̤̐̓̀̎̂̋̚ę̷̠̝̪̟͙̞͚̗̓͒͗̿̏̋̕͜͠ẩ̴̰͈͖͈͚̈́̀̃̍̊̅͘d̷̠̝̣̞͎͋̃͋̉̂̕̚͘͝ ̵̤͍̞͍̞͇̺̻͙͂͊̚͜t̴̰̦̣͓͙̙̘̄̒h̷̡̧̛̙̯͓̙̟̲̩̎̌̐͗̌̒͒̃̕i̶̛̘̠̦̣̟̦̳̅̑̊͒͘͠š̵͇̠̚͜?̶̢̬̓̀̌̈̉̅̌͜
Yes I can.