Its Just Too Massive!

Discussion in 'Planetary Annihilation General Discussion' started by stevenrs11, January 17, 2014.

  1. iron420

    iron420 Well-Known Member

    Messages:
    807
    Likes Received:
    321
    I'd rather have build templates and configurable policy settings
  2. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    But that is only a 15% increase. That isn't much.
    Even without this system, if you have 8 metal spots and I have 8 metal spots and take one from you will lose 12.5% and I will gain 12.5% which totals to a net difference of 25%.
    Metal spots already promote territorial control.

    Edit:Actually, it is a net difference of 28.6%. 9/7 = 1.2857...
    Last edited: March 6, 2014
  3. kingbarber

    kingbarber New Member

    Messages:
    7
    Likes Received:
    1
    I think you did your math wrong.

    In the current system and let’s just simplify by saying each mass point produces 2 mass per second. If we both have 10, we both make 20MPS. If I take one from you, I will have 22MPS and you will have 18MSP. That is a difference of 18.2%

    In my proposed system, if we both have ten, no one gets the bonus! So we again bother earn 20MPS. Now if I take one from you, I will now earn 25.3MSP and you will only earn 18MPS

    That is a difference of 28.9% or 10.7% the increase to the player with the most control over resources in the region from only a difference on one mass point.
  4. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    The thing is, you are always striving to deny the enemy income and increase your own income. Unless it is FFA, you aren't that likely to initiate a fight just because your are close to controlling a region.
    Even if you control a region, you want to push on so that you can deny the enemy all the metal spots in that region. A 15% bonus from controlling more metal spots in 1 region doesn't really change that.
  5. kingbarber

    kingbarber New Member

    Messages:
    7
    Likes Received:
    1
    True, it doesn't really change that. The real change is how much metal would be available to everyone as it s based on the entire regional amount only and not individual resource points per say. In this way, different regions of the world and even different planets could produce more or less resources. That is the real big change.
  6. arsene

    arsene Active Member

    Messages:
    166
    Likes Received:
    114
    I haven't read all of the thread, but I wanted to plug an article I found:

    http://en.chessbase.com/post/computer-resistant-chess-variants

    I thought this was interesting and surprisingly relevant to this thread. Having a very high number of metal points could stress board awareness. If you have too many points of interest, you could say this would add too much complexity and eventually create a game that does not reward the clever strategist, but someone with a brute force, machine-like approach to strategy (attack many different targets, constant unit movement, constant micro). If with human intelligence you could pierce through the chaos and find all sorts of patterns not apparent to someone mindlessly managing his units and economy then there is hope for the game.

    To quote a part of the article:

    Last edited: March 6, 2014
  7. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    This can already be achieved by changing the number of metal spots on the planets. There will come a slider for that so you can adjust how many metal spots there should be on the planets that you are playing on.
  8. arsene

    arsene Active Member

    Messages:
    166
    Likes Received:
    114
    To add to my previous post. Some simple math: if you have 5 items (points of interest on the map) and you want to pick (do something with) 2 of them, there are 10 ways to go about this. But if you have 10 items and you want to pick 4 of them, you have 210 options. i.e. there is exponential growth. No human can possibly weigh 210 alternatives for their particular merits, so you might find yourself lost as the complexity of the game grows.

    To ensure that a mindless player who excels in brute-force can not thrive in this environment, a clever player should be able to find ways to turn those 210 alternatives into a vastly smaller number. So before you can have interesting decisions in the game, first you need to enable players to think about the game in terms of higher level strategic themes.

    The article mentions "local spatial reasoning", which is what humans are good at, apparently. A simple example for Planetary Annihilation would be the fact that metal clusters that are close by are easily seen as a single cluster of metal points, greatly reducing the complexity of the position.

    It's a bit theoretical, I don't know if it's applicable to this discussion, (especially since brute force in chess is very different from brute force in an RTS), but I thought it was interesting.

    --
    In Starcraft II it used to be the case that a typical game involved two base economies on both sides. There was a lot of effort put into changing that number to around four bases. This made a massive difference in terms of gameplay, actually. I think you'll find that as long as there are around twenty total metal points on the map in PA, that any more will be strictly unnecessary. It might still be nice just to have variation in clusters of metal points, but I think that generally speaking anything above twenty is equivalent.

    Someone said that more metal points makes expanding less binary, but this is only true when you go from ten to fifteen metal points, not when you go from fifty to sixty metal points.

    So having established this, if you want to reduce the mechanical difficulty of the game, you really want the lowest possible number of metal points while still keeping the non-binary aspect of expanding, and still keeping the same feel of being spread out over the map. Or alternatively you can have UI to more efficiently manage multiple bases etc.
    Last edited: March 6, 2014
    cdrkf and thelordofthenoobs like this.
  9. thelordofthenoobs

    thelordofthenoobs Well-Known Member

    Messages:
    368
    Likes Received:
    356
    I agree with most of what you say but wanted to add a little detail: If you have more than 2 players having more metal spots is possible, as players generally divide the map up into chunks amongst them so there have to be more metal clusters to give each player interesting options.
  10. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    Interesting read but I don't think it is that relevant to AI in a game like PA because it is so much more complex in computability than a board game.


    I strongly disagree. Every single unit and their location ingame could be seen as a point of interest. It is not just metal patches that should be seen as "points of interest".


    Since every unit can be seen as a point of interest we really do have to weigh the ways in which we use the units.

    This is exactly why humans are better at RTS than computers because they start of with an approach from higher strategic thinking. Brute-forcing the endless possibilities in PA is simply not possible.

    Similarly mexes can be seen as territory control rather than discrete units. How many points of interest does area control amount to?

    We already have area commands to make any amount of mexes. More metal extractors doesn't necessarily raise the mechanical difficulty of the game.
    Last edited: March 6, 2014
  11. goldshekelberg

    goldshekelberg New Member

    Messages:
    20
    Likes Received:
    8
    My numero uno idea for a fix for poor attention spam in a micro-intensive strategy game: git gud.
    360 degree shitstorm, 2 frames per second, hundreds of units getting stuck in your poorly designed factory layout, your friend shouting "NOOOOKES" in your ear over VOIP and world's worst user menu is what I absolutely love in PA and wouldn't change for a thing in the world.

    No, but seriously, the problem isn't within the game, it's within you. My only complaint concerning micro is game's performance. It bottlenecks my abilities, the game later on starts lagging so badly even on powerful systems.
  12. mushroomars

    mushroomars Well-Known Member

    Messages:
    1,655
    Likes Received:
    319
    I figure I'll weigh in, I think that a lot of PA's micro will evaporate once Uber finishes their GUI overhaul that they talked about for Gamma. Right now, there are a lot of little things that make me want to pry my fingernails out with rusty Spackle knives. Double click to select all of a type on screen, the GUI delay as the game drags past the hour mark, view rotation not working if you activate it while the cursor isn't on the planet, somewhat finicky and sometimes buggy strategic view, multiple issues with the drag area select, the list goes on.

    I find that around 25% of my effort, if not time, is spent wrestling with the interface, and it's kinda draining for longer or more competitive games.
    cdrkf and tatsujb like this.
  13. Slamz

    Slamz Well-Known Member

    Messages:
    602
    Likes Received:
    520
    Not sure if anyone has mentioned this, but what's up with metal density.

    I feel what the OP is saying about too much metal is way more true on tiny planets, where the game seems to think it needs to pack the metal spots in. On a scale 4 world it's harder to defend a bunch of metal spots because they're all spread out. On a scale 2 it's pretty easy to defend a horde of metal because they're tightly spaced. In fact, going from scale 4 to scale 3 doesn't reduce the metal count as much as it just packs almost the same number closer together.

    I do feel that tiny planets get way too much metal. I guess it helps keep it fair but it's almost like playing on an old TA "metal planet" where metal was unlimited.
  14. neutrino

    neutrino low mass particle Uber Employee

    Messages:
    3,123
    Likes Received:
    2,687
    Once we get build restrictions in just set a limit of extractors. Next.
    eroticburrito likes this.
  15. thelordofthenoobs

    thelordofthenoobs Well-Known Member

    Messages:
    368
    Likes Received:
    356
    Our lord has spoken :D
  16. tilen

    tilen Member

    Messages:
    92
    Likes Received:
    58
    I've said this before and I'll say it again. This game is great, but there's a whole lot of things that SupCom already did very good and there's little reason not to include them in this game.

    Take previously mentioned build templates for example. Building snap-to-grid (optional?)- in most of the impressive screenshot in the Steam store page, the buildings are neatly placed. Build order cancelling and moving. The list could go on.

    Something to think about, yeah? xoxo
  17. thelordofthenoobs

    thelordofthenoobs Well-Known Member

    Messages:
    368
    Likes Received:
    356
    That a feature is not in the game right now does not mean that it is not planned (I am 90 % sure that the build order stuff will arrive as that part is obviously not finished..and I hope that we will be able to have factories assist each other to "copy" their build order and so on).
  18. tilen

    tilen Member

    Messages:
    92
    Likes Received:
    58
    That is a possibility I am aware of. :) Still, I have concerns I simply had to voice.
  19. Dementiurge

    Dementiurge Post Master General

    Messages:
    1,094
    Likes Received:
    693
    I protest. AIs are not limited to playing memorized possibilities, nor are humans better at RTS because of their strategic thinking. Humans are better because there hasn't been nearly as much time put into making AIs that win at Starcraft, as has been put into Chess. It was continually hypothesized that computers would never beat humans at Chess until they did.

    An even more apt comparison might be Poker. It's about at the tipping point as far as I can tell, and it has something in common with most RTSs: It's a game of incomplete and misinformation, and anticipating your opponent. If a computer can weigh the probabilities in PA as it does in poker and then guess what you're going to do, it will have outsmarted you.
  20. stevenrs11

    stevenrs11 Active Member

    Messages:
    240
    Likes Received:
    218
    I think the difference between AIs for RTS games and board games primarily lies in the difficulty of recognizing state. In board games, there are discrete positions and formations that combine to form a given state, but in a strategy game, it is far more amorphous. Units and buildings can occupy any position, and the AI needs to parse the relative groupings into some computationally pertinent 'threat value'.

    So not only do we have to figure out which moves to make next given the current state (chess), we also have to an immense amount of work figuring out which state we are actually in (PA).

    Anyway, I cannot wait for uber to unlock the mass sliders and provide build templates. With build templates, I can plop down by turret/wall balls EVERYWHERE. Seriously, I dont think I have lost a game to actual units that has gotten past the 20 minute mark in ages. A combat fabber behind walls+turrets is nearly impossible to dislodge.

    Something else I would really like to see with build templates is persistence. Lets say I give a fabber orders to build a template, and he finishes. A bomber destroys part, and that fabber should rebuild the destroyed part. That, imo, would help the micro burden immensely.

    Is there a thread dedicated to build templates yet? If not, I may make one.

    That would definitely solve the issue as well, heh. Though maybe at the cost of feeling a bit contrived, and discouraging expansion once you have maxed out your extractor count.
    corteks and godde like this.

Share This Page