AaronEthridge
Chieftain
- Joined
- Mar 11, 2013
- Messages
- 2
First off, hello everyone! I'm Aaron Ethridge the lead developer for Worlds of Magic. You'll have to forgive my first post being a response to a post about the game I'm working on. (Most of my forum time ends up spent on our forums for obvious reasons, lol.)
LdiCesare, allow me to respond:
First, I opened that post on our forums warning the readers that I was speaking in the broadest possible philosophical terms. The post was about our approach to developing the AI, not HOW we were going to develop it. In light of the fact that I was exploring the philosophy and not the method I feel your calling it hollow is unjustified. The idea has to proceed that action, the concept before the code.
I didn't mean to offend anyone with my statement , but it explains why a lot of AIs fail. The coder who is writing the AI doesn't know how to play the game. That's a problem. I readily admit that it was presumptuous, as I haven't personally played against all the coders of every game I've ever played. However it's something that seems safe to presume, at least in some cases. I wasn't talking about games like Master of Magic (which I actually mentioned later in my post) or Civilization or MoO or any of the other greats. I was referring to games cranked out in mass that leave you with the feeling that The programmers must have never even PLAYED this game!!! If you've never felt that way then you're lucky. Most of us have. In any event it's not place to judge their play abilities (especially having never played against them) and I stand corrected. My real point was that we CAN play and that will help us in our crafting the AI. As that was my point I should have stuck to that. I misspoke and would appreciate it if you would give that the go by and look at what else I had to say.
Codeing an AI is harder than playing a game. No argument there. Making a car is considerably more difficult than driving one. (And yes, I admit that, a bad car design doesn't necessarily mean that the engineers can't drive. Lol.)
Now, I completely disagree that the AI will always react the same way to the same stimulus. That's up to the coder to decide. A player certainly won't react the same way to the same stimulus every time. Still, I believe I understand your point. That being that the TRICK is to make the AI react in a natural way to the information it's collected. If that is your point I agree entirely.
And no, I didn't say anything about that programming of the AI. There are two reasons for that. First, it's too simple to be worth discussion. Second, it's too complicated to discuss in a single post on a forum. Especially one that is opening the topic. Let me give you an example:
How do we plan to program the AI? Well, each turn the AI is going to collect all the available information. That information will then be processed by what is essentially an extremely complex series of IF statements. Based on the results of those IFs it will then set city building orders, move units, explore the map, engage in combat, etc. The entire system will be influenced by a number of factors, such as the AIs personality which will cause different AIs to react differently to the same stimulation. Some will expand as quickly as possible, others will build super cities only expanding when prime opportunities present themselves.
Now, the above is true. That is exactly how it works. Does it really tell you anything? Do you want to see the code we're actually using? ( I can't actually show you that for obvious reasons. I'm sure you understand my point.) One of the main reasons I did the post was to answer questions. You didn't ask any. I'm sorry if it strikes you as hollow and I'll be glad to fill in parts for you. What do you want to know?
Again guys, sorry my first post is a response. I think I need hardly say that I am VERY busy right now. LdiCesare, I honestly appreciate your post. It gave me an opportunity to jump in here and say a few words about what we're working on. (It's something I'm very excited about.) I invite all of you to come by and let us know if you have a question or suggestion. If you've looked at our forums at all then you know that we are very involved with our own community and take feedback very seriously. Thank you all for the opportunity to respond
LdiCesare, allow me to respond:
First, I opened that post on our forums warning the readers that I was speaking in the broadest possible philosophical terms. The post was about our approach to developing the AI, not HOW we were going to develop it. In light of the fact that I was exploring the philosophy and not the method I feel your calling it hollow is unjustified. The idea has to proceed that action, the concept before the code.
I didn't mean to offend anyone with my statement , but it explains why a lot of AIs fail. The coder who is writing the AI doesn't know how to play the game. That's a problem. I readily admit that it was presumptuous, as I haven't personally played against all the coders of every game I've ever played. However it's something that seems safe to presume, at least in some cases. I wasn't talking about games like Master of Magic (which I actually mentioned later in my post) or Civilization or MoO or any of the other greats. I was referring to games cranked out in mass that leave you with the feeling that The programmers must have never even PLAYED this game!!! If you've never felt that way then you're lucky. Most of us have. In any event it's not place to judge their play abilities (especially having never played against them) and I stand corrected. My real point was that we CAN play and that will help us in our crafting the AI. As that was my point I should have stuck to that. I misspoke and would appreciate it if you would give that the go by and look at what else I had to say.
Codeing an AI is harder than playing a game. No argument there. Making a car is considerably more difficult than driving one. (And yes, I admit that, a bad car design doesn't necessarily mean that the engineers can't drive. Lol.)
Now, I completely disagree that the AI will always react the same way to the same stimulus. That's up to the coder to decide. A player certainly won't react the same way to the same stimulus every time. Still, I believe I understand your point. That being that the TRICK is to make the AI react in a natural way to the information it's collected. If that is your point I agree entirely.
And no, I didn't say anything about that programming of the AI. There are two reasons for that. First, it's too simple to be worth discussion. Second, it's too complicated to discuss in a single post on a forum. Especially one that is opening the topic. Let me give you an example:
How do we plan to program the AI? Well, each turn the AI is going to collect all the available information. That information will then be processed by what is essentially an extremely complex series of IF statements. Based on the results of those IFs it will then set city building orders, move units, explore the map, engage in combat, etc. The entire system will be influenced by a number of factors, such as the AIs personality which will cause different AIs to react differently to the same stimulation. Some will expand as quickly as possible, others will build super cities only expanding when prime opportunities present themselves.
Now, the above is true. That is exactly how it works. Does it really tell you anything? Do you want to see the code we're actually using? ( I can't actually show you that for obvious reasons. I'm sure you understand my point.) One of the main reasons I did the post was to answer questions. You didn't ask any. I'm sorry if it strikes you as hollow and I'll be glad to fill in parts for you. What do you want to know?
Again guys, sorry my first post is a response. I think I need hardly say that I am VERY busy right now. LdiCesare, I honestly appreciate your post. It gave me an opportunity to jump in here and say a few words about what we're working on. (It's something I'm very excited about.) I invite all of you to come by and let us know if you have a question or suggestion. If you've looked at our forums at all then you know that we are very involved with our own community and take feedback very seriously. Thank you all for the opportunity to respond
