Walking to a Destination (DumbMan's Webwalker)

[Qubit]

            So I was wondering does using the path[] make detection easier. I am assuming it does because of a player or players going to the exact position hundreds of times?

 

So I was thinking what would be a better way to deal with this.. I multiple paths or Area[], you randomly start off by one random position in the Area[0]. Then you follow that path in the Area[].randomPosition()... But you have a small chance of moving over one lane and even a much small chance of moving over 2 lanes up one position.... Since I am not educated enough to write a web walker, but i sure will next year at my first year of college for computer science, I thought this would be better temporary replacement than the regular position[]. What do you think?

 

Ex(s). of the algorithm being run

 

 

 

Edited June 2, 2015 by javant

[Twin]

A lot of people on here say that no matter what, a computer can't be random and a pattern can always be found. I don't know how true this is, but I think webwalking is made more for getting around quicker than it is detection.

[ProjectPact]

You aren't considering obstacle handling though ;)

 

that's one of the biggest aspects of a webWalker

[FrostBug]

            So I was wondering does using the path[] make detection easier. I am assuming it does because of a player or players going to the exact position hundreds of times?

 

So I was thinking what would be a better way to deal with this.. I multiple paths or Area[], you randomly start off by one random position in the Area[0]. Then you follow that path in the Area[].randomPosition()... But you have a small chance of moving over one lane and even a much small chance of moving over 2 lanes up one position.... Since I am not educated enough to write a web walker, but i sure will next year at my first year of college for computer science, I thought this would be better temporary replacement than the regular position[]. What do you think?

 

Ex(s). of the algorithm being run

 

 

 

Could work well. But taking extremely inefficient or silly paths to a destination could probably look at least as suspicious as walking directly to it. I reckon it's more humanlike to walk the shortest distance (or something close to it) than walking zigzag

[Precise]

Those paths are far from human-like which would attract more attention which you don't want.

 

have multiple paths, randomly select one of them, walk that.

 

and for every position you walk in the path, vary the x and y coords by 1 or 2.

 

i like the diagrams ^_^

[fixthissite]

Rather than creating the path all at once, why not generate your next "random" movement after reaching/finding the location you are currently walking to.

 

You generate (asynchronously) a bunch of "next possible step" nodes and choose one. That way, you can generate billions upon billions of different paths. For those saying "the time itll take to process all those nodes" should pick up a multithreading book at some point. For those saying "too much memory", read up on how HotSpot manages garbage collecting.

 

Trying it won't hurt. Creating paths before-hand is pretty damn insane if you're looking for realism.