General distribution and colour detection.

[ElmoNZ]

I'm very new to autoit and very excited to learn how to use it.

I'm making a macro/bot for a browser game I play and I made one successfully with random click delays between left and right.

I was thinking that my practicing autoit would be creating better scripts to bot different things on my game but to do this I was wantings some of your thoughts on creating random numbers relative to general distribution (Like the Bell curve.)

Also, I'd like to know about any colour sampling and colour detection functions AutoIt might have.

Thanks,

-Alex.

[BrettF]

Pixel functions for colour.

What is the game?

I can't see any problems with random numbers but why?

[ElmoNZ]

Runescape. And thanks I'll look into these pixel functions.

Creating random numbers that adhere to probabilities in general distribution would make a much more human like macro. If I could use random distribution in clicking on a certain part of the screen It'd be much more convincing.

General distribution is like how the average IQ of the world is a certain number and the most common IQ would be the top of the Bell curve on the general distribution charts. If you want to find how many people have an IQ of higher than average the higher you go the rarer they get, and the lower you go than the average the rarer the amount is.

For example if you were to have a random number generator using probabilities with GD, say with a median of 100 and a period of +/- 50 and ran it for thousands of clicks you'd get the majority of random numbers being 100 and the ones close to 100 and as you drift further away from 100 the fewer there are.

So if general distribution was used instead of plain old random number generation you'd get a less erratic but still human like behaviour if it were applied to mouse positioning and clicking. As opposed to complete random mouse clicking within a certain area, where there is no element of sense, general distribution is still random but with more predictability and realism.

So if you were to make the most human like macroing bots for games you could use general distribution to generate random positions for licking in an area, and combine it with random shifts in the median ALSO using random distribution and also random chances of there being completely unusual misclicks and the timings between the clicks also using the same principle with the whole general distribution within itself to generate random but human like timings you'd have a far, FAR superior bot to any other that could be made.

That's my idea anyway. I'm a very mathematical person.

-Alex.

Edited May 25, 2009 by ElmoNZ

[BrettF]

Search for the runescape bot then?

Cheers,

Brett