Flynn

Members
  • Content count

    234
  • Joined

  • Last visited

Posts posted by Flynn


  1. I'm not sure how exactly enemy shield strength scales with sectors... it feels like every ship gets a fixed number of points randomly distributed, which when applied to crewless ships, results in slightly higher caps for weapons and shields since there is no o2 or medbays. But even so running into 3 shields in sector 2 still seems like an a outlier -- did that ship have just one weapon point? Why not jump away from that battle, use the glaive on the rest of the easier prey in sectors 2 and 3?

    125 for the teleporter is a better investment... if you don't have another way to deal 3 shields, because otherwise later sectors will be a brick wall. I suppose you could say choices like are dependent on whether you want to maximize how far you get in an average playthrough vs your odds of getting a victory that game. Shields seems better if you luck out and find a way to deal with shields later. Though boarding pays for itself soon enough (and more crew as loot) so teleporter may be better in either case.

    TWO defense drones and a med bay, and I only have missiles and boarding party? JUMP. A rare combo even in the later sectors.


  2. Really? Tell me how you propose to win a battle against an automated scout with 3 shields when all you've got is a glaive beam.

    You know you'll certainly be running into 3 shield scouts long before you actually run into them. If all you have is a glaive at that point in the game then that was a long term mistake. At some point in the earlier sectors you could have purchased an ion/bomb/missile/burst/boarding drone/antiship drone/etc. I'm not sure I've ever not found at least a teleporter, and anyone with level 2 tele can survive taking out the shields, or Rock/Crystal even at level 1 can do it.

    The closest thing to unwinnable situations I run into are the ion storms, especially if you beam into one already damaged fleeing a fight. And they've really taught me how drastically you can cut power and still make it by constantly shuffling, and there's more to the skill ceiling than you might realize at first. When two weapons are coming your way, one laser, one missile, you can unpower your weapons/shields, power sheilds, absorb the lasers, then immediately depower shields to maximize dodge chances for the missile attack, power weapon just long enough for one flurry, back to shields, and so on. Pausing at some point to recharge to air.

    I think teleporters smooth out the item availability a bit too, simply because you find so many more items when capturing ships intact. I think I've won every game where I got an early teleporter + the aug that lets you go back to any previously visited jump.

    I do agree the game is not 100% winnable every time you play though. Especially with something weird like the stealth ship you can get pretty boned by not finding the right thing. But there is quite a lot of room for skilled players to up their odds dramatically. This is probably obvious to anyone but when you choose your route through the system, you want to plan a route that touches as many adjacent nodes as you can to maximize your chances of finding a store. I can't imagine getting all the way to level 3 shields without at least running into a few, even with the worst possible luck.


  3. You guys touched a bit on whether general learning AIs could ever make AI in general more effective (machine learning, genetic algorithms, etc.) without requiring the designer to intimately know every best strategy about the game and construct the perfect decision trees and whatnot, which is never feasible.

    One approach that has been more successful and easier to integrate is Monte Carlo based systems. That's what the latest Go AIs that are finally giving humans a run for their money use.

    It's also what powers Mu and Tichu AIs, here's the developer talking about it: http://hfog.blogspot...01_archive.html

    These approaches really are limited by current CPU cycles -- if there was 10,000 more cycles available, you can really go crazy.

    I believe this absolutely brutal Race for the Galaxy AI was generated using machine learning (trained over a period of months of computing time) http://www.keldon.net/rftg/

    He talks about it in this thread: http://www.boardgamegeek.com/thread/438698/race-for-the-galaxy-ai/page/37

    Some of the AI neural nets for the Brink of War are only partially-trained; the AI performance in big games is way too slow and takes weeks to train. I'll need to fix that both so that players won't have to wait and to improve training speed.

  4. A few scattered thoughts:

    I think hidden-role games like Battlestar Galactica do the "spies, espionage, and covert action" theme right, if you broaden the category just a bit. There is even a sub genre of Starcraft Custom maps that work a bit like this: where group of players uses both in game mechanics, as well as regular text based diplomacy, to ferret out the traitor(s) (who usually have huge in game advantages, but they can't use them without being careful to give away that they have them...)

    Similary, in WW2, the allies had broken the German codes, but they had to figure out how to act on that knowledge without revealing this knowledge. I think there's room for some great game mechanics along those lines.

    Another game that had a really interesting take on espionage is the ill-regarded Game of Thrones strategy game. The single player in that game in crap, but in multiplayer, it has some pretty great stuff. Just a couple of examples:

    Spies can take over cities and castles without the opponent realizing they have been taken over. They give you half-income, but if War is declared, the opponent realizes they have been taken over and you get full control. You could have a unit producing structure infiltrated, so that it produces double agents that you actually control but your opponent thinks he does (until it's too late!). All kinds of crazy stuff.