Tacticular Cancer: We'll have your balls

  1. Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.
    Dismiss Notice

Indie Strategy Games - post 'em if you find 'em

Discussion in 'Strategy Gaming' started by Jed, Dec 9, 2005.

  1. LCJr.gender: ⚧ Erudite

    LCJr.
    Joined:
    Jan 16, 2003
    Parrots:
    2,469
    Click here and disable ads!
    Rogue-likes really aren't my thing but I did play Falcon's Eye about a year or so ago. Graphics are fairly simple but acceptable. Definately an improvement over ASCII. Other than the eye candy it's the same old game. Actually I'm not sure if Falcon's Eye is being updated anymore so may be for an older version of Nethack.



    Previously mentioned Shrapnel has the Horse and Musket series which focuses on Frederick the Great. Can't remember who but someone took the Sid Meier Gettysburg/Antietam engine and did another Waterloo game.
     
    ^ Top  
  2. Balorgender: ⚧ Arcane

    Balor
    Joined:
    Dec 29, 2004
    Parrots:
    4,632
    Location:
    Russia
    Hmm... I heard of Cossacs 2 - Napoleonic Wars (or something like that).
    Try to google it :).
    ____
    Doh, I forgot that it's not indy. Oh well.
     
    ^ Top  
  3. Mefigender: ⚧ Erudite

    Mefi
    Joined:
    Apr 7, 2005
    Parrots:
    1,118
    Location:
    waiting for a train at Perdido Street Station
    Aye. Cossacks 2 smells though. ;)

    I know there is a Waterloo mod for Civi War Generals 2 (it's not particularly great though). Will have poke about for something based on another Civil War game.

    Tried that indie game for Horse and Musket. Was quite good fun for a couple of hours.

    Battles of Napoleon is probably still the leader in the field (and that is ancient). Talonsoft have done some (but the AI is very poor) and there is new-ish game out done by a small indie company (Campaign Napoleon) but again, the AI is rather bad.

    Will keep hunting through the Indie sites I guess. See if someone has managed to convert a decent tabletop ruleset to something other than ASCII.
     
    ^ Top  
  4. Jedgender: ⚧ Cipher

    Jed
    Joined:
    Nov 3, 2002
    Parrots:
    3,287
    Location:
    Tech Bro Hell
    http://www.freeciv.org/: Venerable Civilization clone. Looks really sharp, but I haven't had a chance to spend much time with it.
     
    ^ Top  
  5. chaedwardsgender: ⚧ Liturgist

    chaedwards
    Joined:
    Jun 10, 2004
    Parrots:
    352
    Location:
    London
    On the same note - Civ Evo www.c-evo.org

    Like Civ2, but with a different design philosophy ie
    - AI doesn't cheat, on a level playing field with the player
    - Downloadable alternate ais, and ai creator
    - Design units a la SMAC
    - Wonders do different things
    - Need to build a town hall in new cities in order to collect tax from them.

    Unfortunately, it's too easy on easy and too hard for me on medium, so I'm playing Chariots of War instead at the moment
     
    ^ Top  
  6. Section8gender: ⚧ Erudite

    Section8
    Joined:
    Oct 23, 2002
    Parrots:
    4,321
    Location:
    Wardenclyffe
    That's always a plus. I've never been able to play any of the Civs on high difficulties, just because I don't like playing on such a tilted playing field. Likewise for the easier difficulties. It's not much fun if one party is always going to be ahead of the rest.
     
    ^ Top  
  7. Atrokkusgender: ⚧ Erudite

    Atrokkus
    Joined:
    Feb 6, 2005
    Parrots:
    3,089
    Location:
    Borat's Fantasy Land
    In most strategy games (esp. RTS) it is practically impossible to create a non-cheating and very hard AI. Only when the real AI is invented will it be possible to truly challenge the human mind in such games. Strategies are just way too complicated, compared to something simplstic like chess.
     
    ^ Top  
  8. LCJr.gender: ⚧ Erudite

    LCJr.
    Joined:
    Jan 16, 2003
    Parrots:
    2,469
    ^ Top  
  9. chaedwardsgender: ⚧ Liturgist

    chaedwards
    Joined:
    Jun 10, 2004
    Parrots:
    352
    Location:
    London
    I wouldn't say chess was simplistic - otherwise I wouldn't suck at it, when in fact I do. But what is good is that it has a small core set of rules, meaning that the computer doesn't have to think about too much. The same is supposedly true of C-Evo:
    In fact, the principles of the design behind C-Evo are pretty interesting reading on the whole. http://www.c-evo.org/text.html#design
     
    ^ Top  
  10. Section8gender: ⚧ Erudite

    Section8
    Joined:
    Oct 23, 2002
    Parrots:
    4,321
    Location:
    Wardenclyffe
    Yeah, it has a very simple set of rules, and a huge degree of elegant complexity that can be built on that. For instance, coding up a chess game is quite simple if you only intend it to be played PvP. (hah) Creating a decent AI for chess I would imagine as being insanely difficult. However, even so, it's greatly aided by the limited rule set.

    I agree, except for the "Only when..." bit. From my experience with multiplayer RTS gaming, it's generally the simplistic strategies that win. From Red Alert tank rushing, to Protos Carriers or Huntress rushing, none of my "intelligent" efforts managed to succeed against the idiot singlemind.

    Plus, you have to realise that a great deal of complexity in games is added purely for the sake of it, or in the name of "realism." Chess works because it's simple to learn, but very challenging in depth. Chess would be fucking retarded if someone made the following changes:

    * Get rid of squares. Realistically, army units can move in any direction
    * Real-Time combat, (with optional pause, because we're naive enough to believe that's an acceptable substitute for turn-based) because realisitically, everything moves at once. Plus, our attention span can't stand to wait for another player.
    * All Pieces should resemble what they're supposed to be What the hell are those nobbly little lathed pieces supposed to look like? They're stupid, change them.
    * One-on-One is too limited. The game should be able to support at least 32 players, and preferably somewhere around 10,000 in a persistent world that we can charge people to log into.

    And that, is why the games industry sucks balls.* Developers need to recognise when abstraction of realism is actually more benefit than encumbrance, particularly with relation to AI.

    * Incidentally, when I recently passed my 1000th post mark, I got an impartial observer to analyse the typical content. Here it is for your viewing displeasure.
     
    ^ Top  
  11. Jasongender: ⚧ chasing a bee

    Jason
    Joined:
    Jun 30, 2005
    Parrots:
    10,673
    Location:
    baby arm fantasy island
    Yes, according to recent statistics, the "sucks balls" content has seen significant gains in most analytical pie charts. Or so the Magic Hobo in my closet tells me.
     
    ^ Top  
  12. galsiahgender: ⚧ Erudite

    galsiah
    Joined:
    Dec 12, 2005
    Parrots:
    1,598
    Location:
    UK
    Certainly there are difficulties, but I wouldn't say it's impossible. Also, there is as far as I know, no good definition of *real* AI. What we have now is as *real* as it gets - it just isn't as good as it gets. In the end, all systems - including humans - are formed from deterministic sets of rules on some level. We're very different from computers, but there's no clear separation to distinguish our AI from computer AI.
    Also, I'd say RTS AI is much easier to code than TBS AI. In RTS the computer is on a level playing field in terms of computation time. In TBS it might have thirty seconds to respond to a thirty minute player turn (most TBS AIs won't "think" on your turn, since this wouldn't be liked by players - you don't want the AI to come up with the most cunning strategy ever while you pause for lunch).

    In one sense that is right - in any system which involves chance, it is much harder to come up with a near perfect AI. As soon as chance is involved, calculating all posibilities even two or three turns ahead becomes impossible. Equally, the added complexity of many units / possible moves makes a near perfect AI much harder in something like Civ than it is for chess.

    However, because chess is a comparitively easy problem (in computational terms), it is very well understood. A "good" AI for chess needs to be near perfect, or any good player will beat it. Constructing a world class chess AI is extremely complex, usually involving huge databases of best possible opening moves, known winning endgame positions, strategic rules etc. Chess tactics are relatively easy to code - just look at every possibility a few moves ahead. Chess strategies are much harder - e.g. can you come up with an algorithm for a good first move without using chess "common knowledge"?

    A "good" AI for playing Civ doesn't need to be anywhere near perfect. It just needs to be pretty good. If it makes the tenth best moves every turn, it'll beat you without any trouble. It is difficult to get anywhere near the best moves for such a complex problem though.

    One hard problem is coping with chance and risk. Most players will occasionally take large risks - whether knowingly or not. If these come off, they'll get a large advantage. If they don't come off they might well lose. When most players have a very large setback, they reload. Pretty much every player will reload after actually losing (or before that), rather than starting a new game.
    What all this means is that most players will occasionally adopt high-risk strategies, getting the benefit if all goes well, and reloading if it all goes horribly wrong. If an AI adopts a similar approach - e.g. say by concentrating on non-military development to start with so as to maximize production - and it all goes wrong, what would a player deduce? Does the player think - "The AI must have been taking a calculated risk for long term gain."? Unlikely. More likely he thinks "That AI sucks - it left itself wide open.".

    As a consequence, most AIs will need to avoid any high-risk strategies to avoid appearing dumb if they fail. Also, you need to consider that the goal of creating a hard AI is not to get it to win as much as possible. It is to make the game an interesting challenge for as long as possible. Therefore it is no good to make an AI which adopts high risk strategies - even very good ones -, if these result in either the player losing quickly (not much fun), or winning easily (also not much fun).

    The player will want occasional surprises, but he won't want the AI to take extreme risks which either mean he'll lose if they come off, or win if they don't. Even in Civ type games, the player wants his decisions to be the largest factor - not the whim of a wacky AI. This puts the player at an advantage, since only he will be able to adopt high-risk strategies.

    Also, this lack of extreme risk taking by the computer AI has implications for the player's decisions. Generally, it is safe for the player to assume that the AI is making reasonably consrvative decisions - nothing high-risk. Therefore he can narrow down the AI's actions quite accurately and make decisions accordingly.
    The computer AI, on the other hand, can make no such assumption - it is quite possibly, even likely, that the player is taking many risks, and behaving quite strangely (from the AI's point of view). The AI has to account for the possibility that the player might be doing anything in areas it can't see. Then it has to respond in the safest way possible - so as to avoid appearing stupid.

    The odds are therefore stacked against the AI in ways they are not in Chess. Chess is a game of complete knowledge - no move is hidden from the opponent. Therefore a player can't complain when the AI does something odd - he can see it happening. Chess also doesn't involve chance, and the object isn't to make the game "challenging but fun" for your oponent (not usually anyway). Creating an AI for a Civ style game is a different kind of problem from creating a chess AI, with a different goal.
    The hardest aspect of creating a good game AI is that you're not aiming to win, you're aiming to make the game as challenging and fun as possible for as long as possible.
     
    ^ Top  
  13. bryce777gender: ⚧ Erudite

    bryce777
    Joined:
    Feb 4, 2005
    Parrots:
    4,225
    Location:
    In my country the system operates YOU
    It's all about strategy. You just have to choke out your opposition - literally. By fortifying in key positions, you can keep them from ever getting to your cities in many cases, or away from prime land youw ant to colonize. You also need to go for lots of cities instead of perfectly placed cities. Especially early on.

    Then, once you get to the railroad era, you can often make enough artillary to take out entire nations in one round before they can even respond.
     
    ^ Top  
  14. bryce777gender: ⚧ Erudite

    bryce777
    Joined:
    Feb 4, 2005
    Parrots:
    4,225
    Location:
    In my country the system operates YOU
    Actuaslly, a lot of the reason that they make RTS games is that it is much, much easier to do the AI.

    When you can't even get tolerable pathfinding in a game these days, you can't expect much from AI.

    The biggest reason it sucks is programmers today suck, though. There are far more games made than ever before, with much more high level tools, by generally much less able programmers. When they have to take on something that requires actual computer science knowledge and thinking, they often fail miserably.

    A lot of making a decent AI is breaking down something into the core components and defining them well. If you have a game system with tons of exceptions special cases it becomes very difficult to plug all the loopholes.

    Even if you do have a killer AI, the sad thing is you will likely have to tone it down because otherwise it can play nearly perfectly each time in many cases. For example, in a jetfighter game, it would be extremely easy (on a relative scale) to mimic the use of wingman tactics perfectly. That is not too realistic either, though, because in real life pilots have to communicate, etc. - they can't just instantly always know the heading and speed of their buddy and of you and go right to your six and blast you away.
     
    ^ Top  
  15. Atrokkusgender: ⚧ Erudite

    Atrokkus
    Joined:
    Feb 6, 2005
    Parrots:
    3,089
    Location:
    Borat's Fantasy Land
    2Galsiah: Thanks for a well thought-out essay.

    What I mean by "real" AI is exactly that which so many scientists today try to create. An equivalent of human brain, but with or without the ability to submit to external commands. And when I say "equivalent" I mean a perfect match for human's thinking ability, not just a bot with lots of predesigned behaviors and no sense of humor.

    I disagree. In a game like HoMM3 it is much easier to make a decent AI than, say, in StarCraft or Warcraft. Here's why:
    1. Tactics: AIs just cannot employ clever unit positioning, as well as "unit dance" and other micromanagement techniques. Yes, they do pull a hit unit away for a while, but that's about it. They don't take advantage of terrain, they don't set up ambushes, they don't surround (both with a single company and with several regiments), and the list goes on...
    In turn-based games, it's somewhat easier to make AI look smart, there's not much room for showing his internal cretinism. However, a decent player always notices large and small mistakes that AI is doing all the time.

    2. Building. While in games like HoMM3 there is no factor of building positions (patterns), in a game like Warcraft3 it is very important.

    3. Analyzing the short- and long-term threats and opportunities. In a pretty fast game like StarCraft it is important to quickly recognize the possible strategies that the enemy might come up with, in order to counter yours. AI just cannot do that, he does not predict the possiblities, he just makes what units he has presets for in his script-file, without either analyzing player's current forces, as well as predicting what he might come up with later.
     
    ^ Top  
  16. bryce777gender: ⚧ Erudite

    bryce777
    Joined:
    Feb 4, 2005
    Parrots:
    4,225
    Location:
    In my country the system operates YOU
    All of these things are doable, though. A big reason why it should be easier is that the computer can do any number of things at once. They could easily do unit fades and fighting retreats against any attack you made and sweep around with a counterattack at the same time. I think the reason the AI is not better is simply because if it were better, most people would be too frustrated to play the game and/or the programmers just suck and are unimaginative.

    As for real AI like that, don't hold your breath. The computing power in one brain cell alone is astronomical. In fact, memory is holographic and supposedly a single cell can contain your entire lifetime's memory.
     
    ^ Top  
  17. galsiahgender: ⚧ Erudite

    galsiah
    Joined:
    Dec 12, 2005
    Parrots:
    1,598
    Location:
    UK
    metallix:
    First, note that a "better" game AI is not one that wins more often, it's one that's fun to play against. RTS AIs might make stupid mistakes, but players usually enjoy taking advantage of them. An extremely effective AI might be less fun (and thus worse) than a half decent one.

    Also, that RTS AIs are currently not as good as TBS AIs does not mean it's a harder problem. RTS players are either casual players - probably quite thankful that the AI makes quite a few mistakes -, or hardcore players. Hardcore RTS players spend most of their time playing other humans. There is no demand for a much improved AI from the casual RTS player, and no demand from the hardcore player, unless it were an absolutely amazing AI.

    RTS AI programmers would have to put in a disproportionate amount of effort into the AI to do any better than they do currently.

    TBS games are marketed at a different audience. The casual TBS player is not the casual RTS player - he wants strategy, not to watch soldiers run around and blow stuff up. The hardcore TBS player might well spend a lot of time playing single player, although multiplayer TBS is becoming more popular.
    A very good AI for a TBS is a requirement, and an expectation of the player, so a lot more work usually goes into AI for TBS than RTS.

    Also, bear in mind that TBS AIs are free to do whatever they like within the rules. RTS AIs are not: an RTS AI has to stick within boundaries that the player will consider fair. If an RTS AI were to be able to control all units to any degree of accuracy at any time, the player would call it unfair, since he cannot possibly do the same with his user interface.

    It would be quite possible to make an RTS AI where each individual unit automatically responded appropriately to a situation, e.g. by dancing back and forth, or circling a player unit etc. If this weren't practical for the player to do too, however, it would be considered cheating. Therefore a per-unit RTS AI cannot be too complex. Either the RTS AI needs to opperate through a simulated player user interface, giving commands after "selecting" units perhaps once per second, or - much more simply - the per unit AI can be made less responsive so that all units can act autonomously without the AI appearing to cheat.

    A near perfect RTS AI would not look natural at all, and not seem fair to the player. The need to cripple the AI to make it seem natural means that it ends up performing less well. Making an efficient RTS AI isn't too hard. Making one that's efficient and looks convincing is hard.
    There aren't similar restrictions on TBS AI - the AI is free to aim for perfection, but it's hard to achieve.


    I don't think many scientists today are aiming for that. The view of AI as a search for "true" / "human" intelligence has largely disappeared in practical AI research: most AI scientists don't even like the vague and misleading term "Artificial intelligence".
    Most scientists in the field today are concerned with designing AIs to solve useful problems or perform useful functions. Philosophers are still concerned about what "real" intelligence is, what mind is, what consciousness is etc. etc. but most scientists aren't too bothered - what matters is what can be done with AI, not whether it's "real" or not.

    In most AI fields the goal isn't to emulate human behaviour / thought, but to get an AI to act rationally. Game AI differs in this respect, since the goal usually is some form of "human" behaviour - emulation of a human player.

    What you seem to mean by "real" AI is a computer that thinks in the same way as a human out of the box. That's never going to happen. Human thinking doesn't just happen - to get an AI to be just like humans you'd need to:
    Put it into a physical "body" capable of sensing its environment.
    Constantly give it a load of inputs.
    Allow it to move around in the world and experiment.
    Introduce it to a community.
    Encourage it to have some goals.
    Educate it.
    Wait about fifteen years.

    Human thinking evolves over time after many outside inputs, a load of experimentation, and involvement in a community of other minds. Give a computer this, and maybe it'll have your idea of "real" AI. Otherwise the best that can be hoped for is rational action and complex problem solving - i.e. human thinking without all the mistakes and junk. The mistakes and junk are what makes us human, but they're usually not something to aim for when designing an AI (though for games they are).
     
    ^ Top  
  18. obediahgender: ⚧ Erudite

    obediah
    Joined:
    Jan 31, 2005
    Parrots:
    5,051
    What experiences/references are you drawing this from? The ideal opponent of any game is someone playing to win, maybe not equivalent skill level - but close enough that the outcome is up in the air at the beginning of the game. So the "best" game AI would provide a stand-in at a wide range of skill levels, and plays to win. In a game like civ, you'd hope to see a wide range of AI difficulty levels, strategic or moral codes, and levels of acceptable risk - and when you turn down the difficulty you should be turning down the AI's capabilties, not it's desire to win.

    You're mixing up the concepts of AI and AI-game interface. The problems you describe should be solved in the interface the AI has to control it's forces. If you don't want an AI to micromanage all units simultaneously, don't make that ability available in the first place. Speaking of a per-unit AI, if these games have it, why not have my units default to it to level the playing field - have all units on both sides have a basic ai, and then just implement a strategic ai for the computer side.

    It wouldn't surprise me if this is how RTS AI is handled, tieing relatively mindless automation into the mechanics at the lowest level that can maul humans because of it's processing speed and total awareness , then crippling it until it performs poorly enough to be confused for a person. Anyone with even the slightest interest in AI would puke at this and then study how people play the game and go about making an interface that an AI system could use to play the game. In an ideal world, every game would have an interface API that humans and AI's would share.

    This seems pretty true. But there will always be rich loonies to fund this research. And like most fields, todays work on the seemingly futile, "big thought" problems, will be the groundwork for more practical work in 5-50 years.

    This isn't true at all. There is lots of activity in the engineering, scientific, and philosophical partitions of AI. Sure there are people out there to whom ANNs and GAs are just tools in the toolbox, but there are also lots of people studing learning, robotics, complexity, linguistics, .... and they want to learn big things, not just turn out tools.

    A bold, but baseless assertion. You give some good reasons why it is difficult, and our grand children may have to exhume your corpse to tell it you were wrong - but it's no FTL travel.


    And everyone remember, the "Artificial" in AI doesn't mean "fake", it means "man-made", and it should probably be Artifactual Intelligence. The goal of AI isn't to pass the Turing test, it's to explore the huge space of pontential engineered (or artifactually evolved) intelligences.
     
    ^ Top  
  19. galsiahgender: ⚧ Erudite

    galsiah
    Joined:
    Dec 12, 2005
    Parrots:
    1,598
    Location:
    UK
    I agree, but you're outlining a strategy to make an entertaining (entertaining is probably a better word than fun I guess) AI for as many people as possible.

    My point is that the goal of the AI is to make the game entertaining. Everything else is subordinate to this and comes afterwards. Forgetting that and (for instance) focusing totally on making an AI to win the game nearly all the time is dangerous - you might find that the AI ends up being boring to play against. Even making AI at a variety of levels might end up not being enjoyable to play against if the AI shows little tactical / strategic variety.

    I mostly agree with your above analysis - it's usually right. I'm not sure it always is though: the AI doesn't necessarily need to play to win in order to provide a fun game for the player. This is usually the case, but it should not be a starting point unless you've thought things through from the initial goal of making an entertaining game.

    For instance a player might very much enjoy playing a tennis game against an opponent who is immensely skilled, doesn't try to win, but keeps the player on his toes and pulls off insane trick shots for the fun of it.

    Parallels in strategy gaming are less obvious, but how about:
    An brilliant but insane enemy general who likes to demonstrate his superiority by fighting you using only half his units, while forming the others up into giant letter shapes to spell out the message YOU SUCK and demoralize your fighter pilots.
    Perhaps there are gamers in the world who would find this amusing (it's a frightening world, so you never know :)). It's conceivable they might enjoy it more than a less capable AI trying its best to win.

    If you start further on than "A good AI makes a game entertaining", then you're cutting out options prematurely. Depending on your target audience, they might be good options.

    Yes - that's how things should ideally be done. However, I have my doubts that they always are. I'd guess it's quite often an omnipotent but crippled AI, rather than an AI using a interface at a player's speed.

    You could do that - I'm sure that is done a lot. It's very unlikely to be ideal from the both the player's viewpoint and the AI's viewpoint though. The player won't want the per unit AI to do anything too drastic - he won't want the per unit AI to create a situation he has to manage unnecessarily, even if that action were theoretically the best thing for the unit to do.
    The computer AI on the other hand would want probably want the per unit AI to do the best thing regardless, even if it required attention, or higher level plans to be changed. The AI won't get annoyed at interruptions or fast changes of "thinking" or focus, whereas the player will.
    The per unit AI will almost certainly be designed to suit the player's needs - falling short of the ideal for the AI opponent.


    Probably, yes - at least unless there's a very good reason to do things differently. One problem with this might be that the user interface isn't finalized before the AI is created. It would be difficult to know how fast various AI operations should be to parallel human play without knowing details of the UI.
    For example how long should it take the AI to select all units of a given type on the screen? Does it require a only a double-click or hotkey press? Does it require a drag? Does it require an individual click on each unit?
    To get an AI to appear to use the same UI as the player, you'd need to consider such timings. Changing these late on might have significant effects.

    Oh I know - I didn't mean that people are only concerned with using AI. What I meant is that the goal isn't to think "what is" AI, but rather "how does it work?" and "what can we do with it?".
    I guess I left out the "how does it work?" above. I meant to include solving useful problems in AI rather than just using AI to solve useful problems. Theoretical results in learning / complexity theory... are mostly studied because of their eventual usefullness (on some level) rather than to further our understanding of what AI is.

    Baseless perhaps, but still true ;). It's a pretty vague statement in any case.
    You'd need accurate definitions of "computer", "thinks in the same way as a human" and "out of the box" before you could answer the question. Depending on the definitions it'd probably be trivially true, or trivially false. That's pretty much why all this "What is AI?" stuff rarely leads anywhere.
     
    ^ Top  
  20. bryce777gender: ⚧ Erudite

    bryce777
    Joined:
    Feb 4, 2005
    Parrots:
    4,225
    Location:
    In my country the system operates YOU
    They could still do some actual strategy, though. Such as the occasional attack that falls back in a retreat and then a sneak attack with a bigger force comes from the other direction. The reason they don't is just because the programmers/designers just suck.

    I agree the 'strategy' in real time games is lucky if you even see the units working together in clumps let alone anything even basic like a reserve or a multiple prong attack.
     
    ^ Top  
  21. Atrokkusgender: ⚧ Erudite

    Atrokkus
    Joined:
    Feb 6, 2005
    Parrots:
    3,089
    Location:
    Borat's Fantasy Land
    And though I myself am a hardcore player, I have a friend who has never played SC with other people, and naively expects a challenge when playing with AI. He's very pissed off about the fact that not only SC, but virtually every single RTS has an insane AI that has the upper hand not because of advanced scripts, but because of banal cheat (twice as much resources, fast-building etc). And my firend is not the only one, I'm sure many people are too shy or dont have time or friends to play online, so they just want to a little brain-cringing, hands-twisting in the cold winter evenings after work.

    So the interest rate is equal: both RTS and TBS players want an artificial adversary that could pose a challenge without resorting to stupid cheats.

    Why do you think that TBS players are in any way inclined to play in singlepalyer mode? Take HoMM series for instance: these games were designed to be played in multiplayer mode (you can even play it without CD in).
    However, I do agree that HoMM's AI is no way better than that of StarCraft's in terms of challenge, but at least the former does not resort to cheats, and that's a clear indicator of a somewhat lesser difficulty in creating TBS AI, due to some of the reaseon I've listed above.


    I wouldn't divide the players in casuals and otherwise. Some RTSs have a pretty bad multiplayer mode, they focus on campaigns and SP skirmishes, and yet their AIs are in no way better. First Warcrafts didn't even have battle.net support (cuz it did not exist at that time).


    And yes, AI behaviour in RTS is much harder to code, and I've already given the reasons.


    And please don't try to say that making an extremely hard AI without giving him cheats is possible right now. If it was possible with current means, someone would have done it. So many great companies out there. Take Blizzard for instance: they made an almsot perfect RTS with perfect balance and diversity. However, even they failed utterly in creating a decent AI.

    And no: there's no need to cut their microcontrol capabilities, it is not a cheat. Why? because it doesn't make AIs good microers.
    Get any semi-professional player against an AI in micro-wars match and you'll see the blatant, screaming, gaping flaws in AI's micro capabilities, NO MATTER how many actions he can do in one second.
    And that's micro only. Now think macro: computers can't analyze the situation properly, they can't predict what the player might do, they don't have a fluid strategy choices: they are stuck with one algorithm. They might build counter-units if they see that player makes a certain unit, but they won't be able to predict that.
    And the list goes on...
     
    ^ Top  
  22. Mefigender: ⚧ Erudite

    Mefi
    Joined:
    Apr 7, 2005
    Parrots:
    1,118
    Location:
    waiting for a train at Perdido Street Station
    Gracias :)
     
    ^ Top  
  23. Mnemongender: ⚧ Educated

    Mnemon
    Joined:
    Jul 20, 2004
    Parrots:
    63
    Well not really a Strategy game, nor a Tactics one either given it works with dice ... but - this just fits in here more then in the RPG section.

    Any one know Bloodbowl? Here you go: http://fumbbl.com/ Turnbased and free (though you have to own the original board game for legal reasons), multiplayer only ... requires Java.

    Wanted to post this since ages, given I used to be part of the staff for some time, until I just couldn't afford investing that much time anymore.

    -Mnemon
     
    ^ Top  
  24. TycoonBriangender: ⚧ Novice

    TycoonBrian
    Joined:
    Feb 1, 2006
    Parrots:
    4
    ^ Top  
  25. LCJr.gender: ⚧ Erudite

    LCJr.
    Joined:
    Jan 16, 2003
    Parrots:
    2,469
    Matrix Games http://www.matrixgames.com They have a lot of good indie developers and some old hands like Gary Grigsby. IIRC this is where Tom Proudfoot is working now.

    You can get Steel Panthers World at War for free. Has something to do with SSI and the license. They can't sell the actual game but they've improved it and sell add on "Mega campaigns".

    http://www.matrixgames.com/games/downloads.asp?gid=297


    Shrapnel Games and the Camo Workshop are doing the same thing with Steel Panthers:Main Battle Tank.

    http://www.shrapnelgames.com/SPCamo/wSPMBT/1.htm



    Master of Magic Clone Still chugging along. No AI yet but has multiplayer.

    http://www.masterofmagic.co.uk/
     
    ^ Top  

(buying stuff via the above buttons helps us pay the hosting bills, thanks!)