[frommallamadingdong]

In trying to make a bot appear more human like, maybe someone could create a program to record time intervals for certain actions, with a goal of modifying random number generators by introducing patterns.

For example: when chopping ivy in varrock, many people AFK (and this is still considered legitimate). Someone could make a program that records the time intervals between an ivy section's disappearance and the player's action to chop a new section. I imagine my distribution would take on a bimodal shape, due to "AFKing" in another browser and AFKing when (for example) cleaning the kitchen; maybe the "different browser" and "true AFK' peaks would be centered about 2s and 8s, respectively.
—this data could then be used to create a weighted random number generator, slowing efficiency in training but increasing human-like properties of the bot.

 

[a savage]

#iamverysmart

 

[frommallamadingdong]

Thanks
#IAmGrouchyBasementVampire

 

[american express]

In trying to make a bot appear more human like, maybe someone could create a program to record time intervals for certain actions, with a goal of modifying random number generators by introducing patterns.

For example: when chopping ivy in varrock, many people AFK (and this is still considered legitimate). Someone could make a program that records the time intervals between an ivy section's disappearance and the player's action to chop a new section. I imagine my distribution would take on a bimodal shape, due to "AFKing" in another browser and AFKing when (for example) cleaning the kitchen; maybe the "different browser" and "true AFK' peaks would be centered about 2s and 8s, respectively.
—this data could then be used to create a weighted random number generator, slowing efficiency in training but increasing human-like properties of the bot.

Pretty much everything you've stated there is already a part of RuneMate's PlayerSense API.

See:
Tutorial - How, and why, to use the PlayerSense API
Tutorial - Use custom PlayerSense keys in your bots to simulate different playstyles across your userbase

 

[savior]

Pretty much everything you've stated there is already a part of RuneMate's PlayerSense API.

See:
Tutorial - How, and why, to use the PlayerSense API
Tutorial - Use custom PlayerSense keys in your bots to simulate different playstyles across your userbase

The playersense api has like nothing to do with what he suggested.

Anyway @FrommaLlamaDingDong i think that your suggestions are on the right track. Instead of blindly implementing these sorts of patterns based on nothing else but their own intuitions like most authors with „antiban“ bots do, you suggest to collect data in order to come up with a legitimate candidate. It‘s a proper way of approaching this topic.

Obviously you‘re not the first one to come up with this, and i believe the reason no (popular) bot has ever done this is because even if you have all the data, it‘s still a moderate issue to work on a system that ends up reproducing that data approximately. You could maybe even use AI and train it with that data too, but it‘s getting increasingly more complex at this point.
And all while it‘s not 100% known what types of behavior even leads to a ban.

On the other hand clients like runemate took these routes to crowdsource human-like mouse patterns for example, so on a greater scale where its worth enough this is already being done.