galciv 2 gave extremely massive bonuses to the AI on higher difficulty levels, but I think they were constant throughout the whole game i think they'll add something like that if they haven't already
The_Biz
preodered. if it ends up working even half as good as it sounds, it'll easily be the best strategy game since stuff released 10+ years ago and I'm not just taking that at face value. the design of the game (more info here ) really does have the potential to successfully avoid the problems which always plague traditional strategy games <
[quote who="OsirisDawn" reply="48" id="3463117"] Quoting The_Biz, reply 42so many people are just dismissing AI as some impossibly hard problem the AI in 4X games is terrible because developers are a combination of lazy, poor, and incompetent ... That is quite insulting. Or are you in any position to objectivly rate the effort of AI programmers?[/quote] it's not insulting or derogatory. it's jus
so many people are just dismissing AI as some impossibly hard problem the AI in 4X games is terrible because developers are a combination of lazy, poor, and incompetent if you put skill levels on a scale between 0 and 3000 (eg. some type of elo rating), then the AI you see in 4X games today is way less than 1000 getting it to 3000 might be hard when you want a game to run on a laptop, but that doesn't mean it can't play at 2000 or 2500 which migh
what differentiates "powerful magic" from the type of magic in all the civilization+magic games because people make those with adversarial multiplayer, so there must be some hope for balance usually creativity + ingenuity (hard AI problems) are only required to create strategies, not to execute them. eg. an AI doesn't need to understand how to create a strong custom sovereign in order to play as one eg. an AI can follow build orders without i
after ~100 hours of play I still don't understand how to set up the game so that it stays fun for more than a few hours. there are just too many variables: enemy difficulty monster difficulty victory conditions how I approach the game: many small armies? focus on 1 strong army of heroes? spam cities? conquer AI opponents? ignore the other empires? how good I am at the game what faction/heroes I pick</l
seems like it's a civ 5 mod rather than a new game kind of like colonization was for civ 4
can't blame nvidia for this they made it for supercomputing (i.e. researchers/scientists/engineers) it just happened to be good for gaming as well. the gtx 7## cards are the more mainstream ones
this is more about difficulty levels imo these games are really designed to be played where you and your rivals are equally skilled. if you are pulling ahead easily, then the AI is just too weak
I don't know why none of your games are getting past the openings... this is nothing like my SC2 experience when I played casually. some games would end when one player got too greedy with expanding, but that was uncommon and certainly not frustrating (you know it's a risk, so it's fine to lose because of it...) I don't know the current state of SC2 matchmaking, but I think ranked mode is best for casual play. it's counter-intuitive but that&
don't compare billiards or chess to 4X games... they are leagues apart in terms of complexity and thought-requirements. strategy game AI has to be more than a simple search algorithm... that's what makes the games actually strategic instead of just brute force computation. 4X game developers always make excuses to justify their incompetence when it comes to AI. at first the excuses actually had some validity. up until 2010 people could at least be tricke
it can grow as long as developers keep transitioning away from strategy and towards more casual / accessible concepts (like "look at the pretty spaceship" or "shoot lasers and blow up that planet") in terms of the actual 4X genre and its original target audience, it has been downhill for quite a while, mainly because of there being negative progress since civ4/galciv2 the money (eg. civ5's customers) comes from people who want to run an empire or look at cool things,
the parkour / vertical gameplay / insane speeds really set it apart from CoD it really is very different when you can move fast because you can actually do something besides pop in and out of cover
the suspense and decision-making involved with venturing into the land and not knowing for certain whether or not my heroes / cities will survive against the dangerous monsters originally, the quests/narrative added some level of interest but that went away once the quests were just repeats of what I had seen in earlier playthroughs
hope there's more asymmetry than the current elemental games have. the 4X model of starting with similar stuff and then racing towards victory just doesn't work when the AI needs massive cheats just to stay in the race more emphasis on magic / rpg-things / monsters would be better than some empire-building which the AI can't do
sonic racing, hearthstone, and scrolls really stand out devil may cry was great fun for a few hours shadow warrior, legendary heroes, heart of the swarm, and xcom were okay, but are either not polished enough or not much better than predecessors
I don't think I want the ability to micromanage individual planets' focus, because I will end up doing it all the time to make things finish 1 turn quicker (either local production or global research). Because everything will be useless until it's 100% completed, it will always be optimal to avoid spillover (even if there is carry-over). This micromanagement moves the focus away from strategy and towards time-consuming tasks. I th
galciv actually has decision-making... you can't number-crunch a decision-tree (exponentially-hard problem) I fully expect a non-cheating AI to lose simply because it won't figure out how to balance economy, military, and technology that doesn't change until the AI is designed by an expert player (or actually learns how to play the game over time, which needs very advanced learning algorithms which probably haven't been invented yet)
the AI doesn't need to cheat anywhere near as much as it does it only does that because the AI is bad and is so far away from competent that it uses stuff like 100% bonuses on everything just to compete with decent humans who are beginning the game
with meld, there's finally a reason not to just spend 1 hour on each mission by advancing slowly they didn't fix anything else though :( the balance is still horrible. satellites pay for themselves way too soon. research requires 0 adaptation it seems like no matter what you do, there's just a simple recipe to playing optimally that you follow every game the game just needs a more realistic choice between
sounds like another EA/Bethesda who are using Frostbite/idtech in many of their upcoming games It's nice to see attempts to do something similar for strategy games. We'll see how it ends up the key difference is that id software perfected action gameplay long ago. beyond graphics, it's just a matter of popularizing some combination of maps + weapons + movement ( respawning + matchmaking for multiplayer) meanwhile, strategy game
has any Steamworks multiplayer ever been good? they sell it to all these devs by namedropping stuff like CS, L4D, TF2, etc. but all those games use servers then when the games that actually use steamworks come out, the networking is complete garbage (eg. civ 5, aoe 2 HD)
it'll actually create more work as everyone will expect worlds to be bigger and units to be more detailed... developers will run out of budget for important things but more memory is still a good thing :)
[quote who="Frogboy" reply="12" id="3414860"] 2 of the 3 founders are professional AI developers and someone in here thinks we won't put effort into that? [/quote] this sounds unlikely: 1) there aren't enough resources to work on AI when there are lots of people on a project 2) reduce team size to 4 3) prioritize AI but I'd be happy if I'm wrong :)
let's hope the multiplayer is good reading between the lines, there will be 0 effort put into AI