• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

156 Neutral

About thelurch

  • Rank
  1. ANN

    If all you want to do is creating an interesting sample space in which to observe the behaviour of a neural network then this sounds a bit like overkill. Afterall, it means you first have to create the environment in which you want to test it. i.e. an environment in which your choices of {ATTACK,HIDE,FLEE,WEAPON_SEARCH,MEDECIN_SEARCH,BLOCK,IDLE,SEARCH,FOLLOW,DEAD} all make sense. For instance, how clear is the difference between HIDE,FLEE and BLOCK? and in what situations will one be suitable instead of the other. I would suggest you start with a smaller sample space first e.g.{ATTACK, FLEE, HEAL}. If, however, you are working in a pre-existing environment, e.g. modding a game, then your outputs could be mapped to states the way xEricx described. From what I know of fuzzy logic it's practically the same thing. As for your inputs you need to figure out what information is important in determining the best state. For instance, distance from opponent will probably be important, availability of cover, current health status, current weapon status, weapon oppenent is carrying, opponents health status, ... basically, think of everything that would help you in making a decision of how to act in that environment, and give that information to it.
  2. Selecting Multiple Targets - RTS

    Well, depending on how deep you choose to make the flagging system you could introduce a bit of the strategy you've described. If for instance there is a preparatory stage before battle actually begin then you could use something similar to the second system, where for each type of enemy unit you define what attitude each of the flags should take. Then assign groups of your own unit to each of the flag attack patterns as they are created. Of course, this would require you to know which unit types your opponent is using beforehand. And it will be difficult to change once battle starts simply becuase it would take too long to make the necessary adjustments. Or you could combine the above with a simplified version once the game starts which allows you to assign short term tactical objectives. The idea you've described of marking friendly will be slightly simpler than the first sample I gave. In that case what you basically have is a squads who can be given complex behaviours by marking different enemies in thier squad color, with instructions on how to behave to each. The next level will allow you to make a squad of different unit types, and set the behaviour of each of the unit types in that squad seperately. i.e. by giving each member of the squad a different flag. In the end it's basically a matter of where you want to put the slider between complexity and speed of control.
  3. I think there is an oversight in your original question. Your question assumes that the brain has no idea whether it's got a 'right' or 'wrong' answer while learning. There are two ways the brain judges right and wrong, the first as PERECil mentioned is feelings/emotions. This allows a single brain to judge the rightness of each individual action it performs. and these tend to be largely hard wired from birth (with variations in wiring between individuals). The second is survival and procreation. This allows a community of brains to highlight the brains (and bodies) which are closer to the 'right' answer and design more little brains based on those. (In case there's any doubt, the right answer involves being alive long enough and being attractive enough to convince a partner to help you create little brains) As such humans have become very good at identifying things that are dangerous due to the evolved pain reflex. And can use that to figure out which actions are 'right' and which are 'wrong'. This sounds rather similar to the project Sagar_Indurkhya described. The 'teacher' network is the emotions and is evolved via genetic algorithm. (If a person is wired to enjoy being burnt alive then that would be a wrong answer and they are unlikely to propagate). Then the more complex 'self-teaching' network which may actually be described as a prediction network which attempts to predict which action it's teacher (i.e. emotions) will consider 'correct'. People try to make plans that will make them feel good. Judging the success of this one is trickier, and as failure here is not does not necessarily remove a person from the gene pool there is a much wider variation in this sample space.
  4. Selecting Multiple Targets - RTS

    Sounds like a good idea. The main reason I loved Total Annihilation was becuase I rarely had to micro-manage battles. How about if you have different tag groups. (Sort of like grouping squads in TA or Warcraft). Like squad groups you can assign your units to different Attack tags and your enemies can be tagged for each group. Ok, let me try to explain coherently. Let's say you have three tag colors. Red, blue ang green. You could select an enemy buildingsor group of units) and tag them: Green - Concentrate Fire. Then for fast moving enemy infantry: Green - Ignore. You could then select your own bombers and select attack pattern tag - Green! Which would mean they will concentrate on bombing the buildings you selected. This way you could quickly tell different types of units how you expect them to behave towards different enemy units quickly. They are a few variations of this as well. In the first every enemy has only one tag flag. Say for instance you have defensive enplacements with default set to "Fire at will" and attack pattern Blue. If a group of enemies come by and none are tagged blue then it will simply fire into them randomly. If in that group one is tagged: Blue - Overwhelm! then the enplacements will ignore the others until they have assigned enough units to obey the overwhelm command while the others continue to firing into the crowd at will. This will give really good control in mass battles without needing too much time for assigning commands. ----- Another method could be to give each enemy a full set of flags (by defualt set to null) and assign the attack pattern of each flag. For instance you could have an enemy dragon and assign it's Blue tag to 'Concentrate' and it's Red flag to 'Ignore'. Then assign your archers to Attack pattern: Blue and your infantry men to Attack Pattern: Red. (The archers will concentrate on the dragon while the infrantry ignore it). This method will give even more control for battle tactics but can become very complicated very quickly. I can imagine a dragon flying over my base and, when I finally succeed in clicking on it, I'm frantically trying to remember which color my wizards are before it burns my poor villagers to ashes?. IMHO I think this system will need a slow or frequently pausable gamestyle for it to work well. Other possible implementations can be used but I'm sure you get the general idea.
  5. Ad-Supported free MMO Action RPG

    Hmmm..., IMHO, you actually have taken a valid question i.e. why are all MMORPG based more on dice roles than player strategy/twitch skill? But some of yur conclusions are just waaaayyyyy out there!!! First, about the whole advertising thing. The rate you've given in your calvulations is the market rate for cost per click. However the scenario you've painted is actually cost per view. The price advertising are willing to pay for a single view is usually fractions of a penny. In fact on yahoo banner rates are between $35 and $65 per thousand views!(source:http://www.travelemarketing.com/banner-ads.htm). If you insert those numbers into your equations, and compare them to monthly registrations, the rest should become quite clear. As for the fighting systems. The main issue is always lag! If you can come up with a technique of reducing lag to LAN levels (I believe single digits will be barely sufficient) then the problem is solved!. Even when playing quake on a LAN with 12 people you need a rather powerful system as server and respectable clients systems. (Otherwise every now and then a player may teleport a few inches, and in a game like that a few inches is quite a lot!!!) Alternatively of course, you could design a slower fighting system. (Which, from the previous poster's description, is what Matrix online tried to do, but in thier case IIUC,they still didn't remove the dice only weighted it based on choice, or had both of you choose your moves at the same time, turning it into rock, paper, scissors). In Street fighter for instance, some moves had to be seen and reacted to in a 60th of a second (about 20ms)! Something I wonder though. In most FPS multiplayer stages, the map and all the models are held locally. thereby allowing the bandwidth to be used purely for positional and status data. In some MMOROGs, you actually see the world loading polygon by polygon. which probably means they streaming the info to me. In that case won't storing stuff locally improve much of the lag??
  6. Using the avatar metaphor for the "soul"

    Hmmm..., IMO one important thing for whatever representation system you use is that it should be able to morph easily from one form to the other. E.g. in Black and White. If you were a good or evil God, your hand and castle change so slowly that by the time you realise you have changed, every one else can see it very clearly (sort of like in reality). Using your garden rings example for instance. You could start out with a ring made of brown wooden braid with a few short stalks branching of it. As the game continues the branches grow more numerous and more etangled. If you approach good (or love or purity etc) it begins to tint green -> then the stalks turn into buds -> then you get leaves -> flowers. Until you have a rich circular hedge speckled with blossoms (complete with falling cherry blossoms). If you approach evil instead the branches instead start to turn sickly gray, the stalks become long thorns, from gray it moves to a dull metallic silver, traces of blood begin to appear on the edges and the thorns begin to contain a few full fledged blades. The picture in my mind will also allow you to go from one end of the spectrum to another. E.g. the blossoms start to shrivel and fall off, the leaves begin going that sickly gray, the thorns start growing, the leaves shrink and fall away etc. This will prolly work best for a bi-polar system but it can be made to work for a more complex system as well. As long all the metaphors have a common root. (Although it may require some fancy artwork for non-jarring transitions). Slightly O.T., I don't see why an accomplished evil villain should not be able to perform the odd good deed. I mean perhaps he's a megalomanic psychopath with a soft spot for the dads of twelve year old boys (becuase that is how old he was when his dad died). I think it should be sufficient to have a representation system and to have events in-game react to his reputation.
  7. unconventional multi-user game design

    Well currently, apart from a few technical details, your world sounds a lot like Second Life. Now this is not necessarily a bad or good thing, but you might want to look into it just to get an idea of some of the implications of a build-whatever-you-want and do-whatever-you-want (within limits) world. Also, Like Kaz said, you haven't really given much information. Currently all you've described is a world that will have a wide variety of easily creatable scenery, (and possibly NPCs). If what you want feedback on is specifically your ratings system (which is where you've given most detail) I think you might want to have the users specify why they are giving thumb ups or downs (or allow them to rate different areas, graphics, speed, topic etc.) As it would seem a bit harsh for some one with slow bandwidth to get his account suspended becuase his users didn't like the lag.
  8. Is luck a strategy?

    I've been following this thread for a while and at first I misunderstood your initial post (I think the wording of the second paragraph was the culprit). Anyway, it seems most people who replied made the same mistake. So..., If I understand your second post properly, what you actually mean is that you want to remove all the dice rolling in the game. So for instance if a peon hits a dragon you can be sure it will do exactly 10 pts of damage not 1*d20 pts of damage. So when you say "entirely deterministic", you don't mean totally determined by whatever original random seed was originally generated but totally determined by the actions of the player! And in Quote: the main one being that a player with knowledge of what obstacles are between them and a goal could determine if they will succeed or not given their current resources before they even attempt the challenge. What you mean is that when the player sees an obstacle, he can make a plan based on his resources and be 100% sure whether he will succeed or fail (Unlike most RPG's where dicerolls mean you can do exactly the same thing two times and get very different results) I still not too sure what Quote:Essentially each time you start a new game a new seed would be generated which would be used to determine all the procedurally generated content as well as how all events resolve without player intervention. means. My original impression, and apparently that of most of the posters, was that you meant once the random seed was generated, there would be nothing the player could do to affect his 'destiny'. However, from your second post it appears this was wrong. So to use the wildly popular maze analogy. A.) A maze where there is only one possible path you can walk. All doors that deviate from this path are locked but any doors on this path are unlocked B.) A maze where there are many possible paths and each time you meet a door a dice throw determines whether it opens or closes. C.) A maze where there are one or many possible paths and each door requires a key which you might have or might be able to get. Your question is actually B vs C and not A vs B So, I'll base my reply on the rephrased question (If I'm wrong please feel free to shoot me out of the water). I.e. would I like my actions in a game to always produce dependable outcomes as opposed to being determined by a toss of a biased dice. Personally, my answer is totally Yes!!! I always assumed that the reason games used dice was becuase that was the closest way they could model the interaction of dozens of pieces without getting bogged down in calculation. e.g I'm a level 50 paladin wearing heavy armour and wielding a razor sharp broadsword, my opponent a high level ninja (monk)fighting, for one reason or another, with bare hands. We are fighting at night, on my castle grounds. Who wins. In reality it will be a complex mixture of strategy, technique, location and equipment that will resolve this. On a game board it is so much easier to assign probability bonuses to each of the elements and toss the dice. The problem with this is that dice throws tend to cripple strategy. True this is dependent on implementation but you could have what would actually be a perfect strategy in real life (e.g. lock yourself in a steel box with a small hole and shoot the swordsman with your SMG as soon as he enters the room) but in a dice based game one swordsman will eventually survive enough bullets to reach the box and, with his old rusty sword, will still be able to cut your heart out. Now while games are not necessarily meant to be a simulation of life, what actually happens with dice is that, sometimes, failure is not your fault. Personally, I don't like that. I will always prefer chess to yahtzee. (Not sure about poker though, it seems to be more about how good a liar you are and not how good your hand is. Still don't like it anyway) P.S. I was going to apologise for such a long post but then I realised an apology would only make the post longer [smile] [Edited by - thelurch on March 18, 2005 10:39:19 AM]
  9. I recently fininshed playing Prince of Persia:Sands of time. Not an RPG true but I'm getting to the point. I really loved the game. Both the action AND the story. After finishing I heard about the secret level (The first level from the original game) and was happy at a chance to continue playing. Afterall I had just perfected my sword skills. My combos were finally coming purely by thought without having to negotiate with the keyboard first. I was greased lightning with a bad attitude. And I really wanted someone to kill :P I ran through the whole level (which was deviod of any sword fodder) and ended up in a big empty room with a picture of the design team on the wall. Nice touch true, but what I really wished for was that the room would be full of enemies (preferrably respawning ad-infinitum) maybe with a score counter that would show me how well I did before I inevitably got killed. Now, THAT would have been great fun. Plus it would have provided ample opportunity for bragging rights. Now POP:SOT was definitely great fun, but like you said that just seems to make it sadder when it ends [sad]
  10. Just an idea to add to the chime thing In Rome:Total War (turn based) you get little notification icons that stack up on one side of the screen. Glancing at the image gives you an idea what the message refers to, mousing over it gives a summary of what the message is, clicking on it gives you the full details. So you can very quickly determine if you're interested in the message or not. You could have something similar, so for instance you would get a "High Priority" message (with menacing audio cues), glancing at it (or the equivalent of a mouse over) will show "Stand and Deliver!!!". Then selecting it will give the detailed threat. Such a message shouldn't take you more than five second to glance through and maybe ten - twenty seconds at most to respond to. And the truth is if whoever was threatening you could not keep up with you for a mere 20 seconds then odds are you weren't going to stop anyway. In most real life scenarios (and by real life I mean action movies, cartoons and comics :P ) whoever is demanding you stop and listen has to at least be able to chase you for a while, while they try to convince you, before they decide War is the best option. Such a system also allows you to have icons like 'Request for trade', 'SOS' etc. and if it blends well with the game can allow you to deal with what would othewise be considered a ton of junk mail! The issue with such a system will be the interface. If you have FPS controls (which sound like what you've described) interfacing with on-screen icons or messages becomes difficult. I guess you could have something like 'hold down middle mouse button to switch to mouse pointer mode' at whch time your movement goes into a sort of cruise control.
  11. The big problem with Abrahamic religions.

    Quote:Original post by Etnu Uhm, Judaism is a very small religion, and nowhere near the 3rd largest in the world. Yeah, sorry bout that. I corrected it in my next post. Ratio is more like 1:3 not 1:10
  12. The big problem with Abrahamic religions.

    Research: According to http://www.adherents.com/Religions_By_Adherents.html Christianity plus Islam = 55% Hinduism + Budhism = 21% roughly 3:1 Most of the others are neglible except for a huge chunk of Un-believers :P Adjust for the fact that the media is much more interested in the midle east than in east asia and you probably have roughly similar tendecies.
  13. The big problem with Abrahamic religions.

    Quote:Original post by Etnu ...The question I was asking is why it's so much more prevelant in the Abrahamic religions than it is in others. IMO this is actually simply a matter of visibility. Using the statistics provided by zerowolf. Nearly a third of world population is christian. if you add Islam and Judaism to that (the next two biggest religions in the world) you are probably talking of roughly half the world following Abrahamic religions. Which would make it only natural to expect half of the world's violence to come from people who claim these religions. In contrast Hinduism, Budhism, Taoism etc probably cover less than 5% (just guessing here, will try to find actual statistics and report back) of the world's population. So naturally you would expect ten times less violence from these people. Not becuase thier religion is inherently less violent, but becuase they are less people in that religion to perpetrate the violence. I mean there was even a cult that worshipped a death god/goddess which was on the news a couple of years ago. Now I'm sure they were very violent but luckily for us they were only a few dozen of them to 'spread thier beliefs'. Oh and Falkone Quote: And I do not believe that violence begets violence becuase there is no proof or logical succession. Violence is more of a constant than a cause-effect. All the proof you need of the procreational tendencies of violence can be found in your nearest original Kungfu movie(complete with bad lipsyncing and fake effects). funny as it seems the truth often runs in a very similar vein. =============== Excerpt from upcoming Kungfu movie "The Violent Fist!!!" :-P "Aha! Huang Lo, I have finally found you! You killed my father and now, my Tiger Crane technique will kill you!!" "Haa!! That was becuase your father used his Tiger Pigeon to kill my father! but Snaking Monkey technique returned the favour" "Humph! My father was only taking his revenge becuase your father used his Monkeying Worm to kill my uncle Yun Tang! He could never let your father get away for that." "Hehehehe! That was only becuase your uncle used his Pussycat Budgie technique to..." *a couple of hours later* "Hai!!! But that was only becuase your great great great grandfather refused to return the favourite chopsticks which he borrowed from my great great great grand uncle!!!!!!! Now let's fight Hai!" -->Cut to trampoline segments, cue weird punching sound effects , blood drips from side of lip, ...aaannnnnddd it's a wrap!!
  14. The big problem with Abrahamic religions.

    I believe the idea that any religion promotes violence is fundamentally flawed, becuase it is focusing on the excuse not the reason. People are fundamentally violent and always seem to gravitate towards an Us vs Them scenario. It doesn't matter whether the dividning line is defined by race, sex, nationality or religion. People are attracted towards conflict. They need to believe the 'Them' are evil so that they feel that 'Us' are good. You will notice that whether or not most coflicts may have a good excuse at the beginning in the end it always seems to degenerate into a "We are good and most destroy the evil them no matter what!" and it always helps the people in power (be it religious or political) to whip thier followers into this kind of rabid frenzy while they use it as an opportunity to extend thier own power base. This pattern can be seen from the old testament days, right up through the crusades, to the cold war and the current war on terrorism. Unfortunately, no matter how much I think about this problem I can't really see a way out ofit.
  • Advertisement