Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Users of my iOS Game Teach Me a Lesson MIT Didn’t (2011)

blog.aaroniba.net

Post reply on HN
Print view View original

1–10 of 60 posts

Re: Users of my iOS Game Teach Me a Lesson MIT Didn’t (2011)

#2

Post by stormbrew »

I think it's interesting that people who make games often don't think of their players as basically a giant parallel computer. You're essentially sending a puzzle off to hundreds, thousands, or millions of 'cores' that will through a combination of brute force, intelligence, and a complicated but large network of communication, come to an optimal solution in a time span you as an individual probably can't even comprehend. They will not only solve your puzzle, but they will solve for how to solve it and then spread that knowledge.

Thus I found it hilarious when Blizzard thought it would take people months to completely finish Diablo 3. Real time? A week or two.

Re: Users of my iOS Game Teach Me a Lesson MIT Didn’t (2011)

#3

Post by sherm8n »

stormbrew wrote: post #2

I think it's interesting that people who make games often don't think of their players as basically a giant parallel computer. You're essentially sending a puzzle off to hundreds, thousands, or millions of 'cores' that will through a combination of brute force, intelligence, and a complicated but large network of communication, come to an optimal solution in a time span you as an individual probably can't even compre…

It's human nature. When your goal is to challenge or stump someone you think there's no way they can be smarter than you. You designed it so you should know it best.

Re: Users of my iOS Game Teach Me a Lesson MIT Didn’t (2011)

#7

Post by tikhonj »

This reminds me a lot of the section on constraint-satisfaction problems from an AI class I took a while back. You can solve constraint satisfaction problems using backtracking, but this is very difficult--it takes exponential time. However, for certain problems (like the well-known n-queens problem), the optimal way to solve it is to start with a random configuration and jiggle it around.

This is the same thing but with people instead of CSP-solving algorithms.

Re: Users of my iOS Game Teach Me a Lesson MIT Didn’t (2011)

#8

Post by sherm8n »

tikhonj wrote: post #7

This reminds me a lot of the section on constraint-satisfaction problems from an AI class I took a while back. You can solve constraint satisfaction problems using backtracking, but this is very difficult--it takes exponential time. However, for certain problems (like the well-known n-queens problem), the optimal way to solve it is to start with a random configuration and jiggle it around. This is the same thing but…

Omg, do you remember solving shortest path graphs by hand?
Post reply on HN