R4tt3xx
I want to believe
- Joined
- Apr 10, 2005
- Posts
- 2,175
- Location
- South Africa
- Society
- Freelancer
- Avatar Name
- Alexis Sky Greenstar
Please note that the following post is framed in terms of mining not hunting and concerns mining only.
So this is a simple concept. Does "system" exhibit neural network (https://en.wikipedia.org/wiki/Neural_network) / reinforcement learning (https://en.wikipedia.org/wiki/Reinforcement_learning) characteristics ?
How do we test for this, if "system" is essentially an "AI" (https://en.wikipedia.org/wiki/Artificial_intelligence) what would it's goals be ? What would it do in order to make sure that it's goals are met etc ?
So this has been bugging me for a while, after various interactions with "system" and monitoring the game's response, I think that reinforcement learning can be used against "system"
Allow me to explain. I can think of two obvious goals for system if it was an AI / reinforcement learning system.
1) Entertain the player in such a way that they are willing to spend peds inside Entropia Universe.
2) Ensure that the players remain engaged with the system for as long as possible.
My goals in this "game" so not match either of these proposed "system" goals as my goal it is simply to make as much ped as possible... So I have been testing "system" the following way.
1) Probe a unique set coordinate pair.
3) IF I receive no monetary reward, ie if I miss, I terminate my interaction with system and the game (ALT-F4). Returning to the game after a semi set period of time.
4) Head to the next coordinate pair that matches the unique set that it at least 55m away from the current coordinate pair.
5) Goto step 1 until probes have been depleted.
6) Recharge probes with profits of mining run and logoff.
This little procedure netted me a size 15 claim last night, and I will continue to test until I am so bored that I loose interest or peds, whichever one comes first.
I am theorizing that ideally the game factors goal 2 over goal 1 if the loot in the loot pool will allow it. I am also suspecting that I will have to spend more time in a "cool down" state before I will receive any sort of resource reward.
Can anyone think of any additions to add if my hypothesis is correct ?
So this is a simple concept. Does "system" exhibit neural network (https://en.wikipedia.org/wiki/Neural_network) / reinforcement learning (https://en.wikipedia.org/wiki/Reinforcement_learning) characteristics ?
How do we test for this, if "system" is essentially an "AI" (https://en.wikipedia.org/wiki/Artificial_intelligence) what would it's goals be ? What would it do in order to make sure that it's goals are met etc ?
So this has been bugging me for a while, after various interactions with "system" and monitoring the game's response, I think that reinforcement learning can be used against "system"
Allow me to explain. I can think of two obvious goals for system if it was an AI / reinforcement learning system.
1) Entertain the player in such a way that they are willing to spend peds inside Entropia Universe.
2) Ensure that the players remain engaged with the system for as long as possible.
My goals in this "game" so not match either of these proposed "system" goals as my goal it is simply to make as much ped as possible... So I have been testing "system" the following way.
1) Probe a unique set coordinate pair.
3) IF I receive no monetary reward, ie if I miss, I terminate my interaction with system and the game (ALT-F4). Returning to the game after a semi set period of time.
4) Head to the next coordinate pair that matches the unique set that it at least 55m away from the current coordinate pair.
5) Goto step 1 until probes have been depleted.
6) Recharge probes with profits of mining run and logoff.
This little procedure netted me a size 15 claim last night, and I will continue to test until I am so bored that I loose interest or peds, whichever one comes first.
I am theorizing that ideally the game factors goal 2 over goal 1 if the loot in the loot pool will allow it. I am also suspecting that I will have to spend more time in a "cool down" state before I will receive any sort of resource reward.
Can anyone think of any additions to add if my hypothesis is correct ?