[WILL NOT BE PRESENT] Commander Upgrades

Discussion in 'Planetary Annihilation General Discussion' started by kalherine, September 23, 2013.

  1. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    You wish. That stuff is only precise while on the software level. Actuators always have a significant spread in reaction speed.
  2. nanolathe

    nanolathe Post Master General

    Messages:
    3,839
    Likes Received:
    1,887
    The lore says they're functioning at a speed that is at least in the measurement of nanoseconds... able to simultaneously receive, analyse and advise on strategic deployments from multiple sources in the space of just a single billionth of a second.

    Their leaders are also terminally unimaginative amnesia sufferers.
    Last edited: September 25, 2013
  3. bobucles

    bobucles Post Master General

    Messages:
    3,388
    Likes Received:
    558
    And that's why they miss. But electronic projectiles fire with SCARILY PRECISE ability.

    But there's a deeper issue happening here. You seem to think that there's some kind of Rock paper scissors paradigm where bombers are supposed to beat tanks. That's not possible. Bombers, by design, are artillery platforms. Instead of lobbing a large, slow loading shell over long distance, they fly straight up to the target and drop it. That feature allows bombers to pick and choose their victims at will(such as say, a Commander), making them the most snipe-y of artillery type units.

    You don't take a unit built for hit+run sniping, and tell it to flatten armies. It just doesn't work.
  4. GoogleFrog

    GoogleFrog Active Member

    Messages:
    676
    Likes Received:
    235
    exterminanus are you proposing weapon parameters such as these? http://springrts.com/wiki/Units-WeaponDefs#Targeting_.26_Accuracy
    That page could be useful for some terminology along with short explanations. Specifically you seem to be proposing properties similar to targetMoveError, leadLimit and predictBoost.

    I dislike these properties because it can look really strange when some artillery piece can hit structures with perfect accuracy but files very wide for units moving along a straight path.
  5. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    targetMoveError has a similar effect in terms of numbers, but works a bit different and can run into hard caps since it causes the projectiles to fail to keep up, it will never lead to far ahead and is also only applied to automatic targeting.

    It looks strange because Spring makes projectiles fall behind all the time. If you only enforce a wrong lead by introducing jitter into the timing, then the projectiles will still align with the path so it looks like the artillery tried to guess the movement, but partially failed at the task, with projectiles hitting before and behind the target equally. You actually have a calculable chance to hit the target, despite the movement.

    However, the projectiles will still only hit the unit path in direct proximity to the unit, nothing else. Even with manual targeting, they would still have the very same issue since you couldn't control the timing of the launch precisely.


    Besides: The Spring engine featuring such parameters and actually using them (otherwise you wouldn't have complied about the aesthetics) is a good sign, because that means that it works just the way I thought.
    Last edited: September 25, 2013
  6. krakanu

    krakanu Well-Known Member

    Messages:
    540
    Likes Received:
    526
    These are called missiles and we have them right now.

    I'm all for not having arbitrary rules, but if a unit has almost no hope of hitting another unit, why should it even try? If tanks (or mobile artillery) had a long enough range, would you argue they should try to shoot at orbital units too?
  7. GoogleFrog

    GoogleFrog Active Member

    Messages:
    676
    Likes Received:
    235
    Spring projectiles don't fall behind all the time. They lead by assuming no acceleration on the unit and solving the equation.

    How is your proposal different?
  8. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    It's not counter-able by manual targeting, which is still possible with Spring, and it only requires a single variable. Apart from that, it has very similar effects in terms of balancing and the visual pattern of shots might slightly vary compared to what Spring does.
  9. GoogleFrog

    GoogleFrog Active Member

    Messages:
    676
    Likes Received:
    235
    So how do you make your system not counterable by manually targeting the terrain?
  10. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    Remember what you would need to do in order to hit a moving target reliably.

    You would need to lead the target and fire at the precise moment. You already had considered the duration of every pre-firing animation and you would still have the same hit chance as if you had aimed at an stationary target.

    That does no longer work, if the duration of the pre-firing animation is affected by the jitter, you have the very same issue the automatic targeting AI has, you can no longer time the shot reliably if the unit is moving to fast. You still have a chance of hitting, but that chance decreases as the targets unit speed decreases.

    A perfect weapon (which is supposed to hit running targets) has no such jitter, if fires the moment you (or the target AI) order the launch, or at least only with a constant delay.

    A bad weapon (e.g. high damage weapons like tank canons and artillery) has a huge jitter. The targeting AI considers the median of the jitter as a constant value which is then used to determine the time and arc to be used. So in average, projectiles will hit behind and in front of the unit with the same chance. This can be slightly biased to enhance the chance of hitting units when they stop.


    So in the end, the targeting AI does the best job possible to hit the target. It is not possible to surpass the automatic targeting by manual terrain targeting.
  11. occusoj

    occusoj Active Member

    Messages:
    187
    Likes Received:
    34
    Do guided shells already hit planes? I honestly dont know if this technology has advanced that much.
    I mean the stuff that isnt self propelled and gets fired from gun barrels, usually not called a rocket.
    Such experiments with guiding non-selfpropelled projectiles date back as far as the seventies AFAIK.

    Because a chance of, lets say 2%, to survive an attack or shoot down an enemy plane is still better than a straight out 0% by not even trying.
    Id rather see my units almost beeing killed instead of totally ;).

    If there is a possibility to hit, they should at least be able to be ordered to shoot at it.
    Of course the targeting system has to choose higher priority targets first, no point in aiming at the stars when theres an army of tanks rushing in. Obviously this will be constrained by the maximum possible ammount an unit can tilt its weapon, so no firing straight up for tanks in their current shape.

    Would you forbid people in BF3 to shoot down choppers and jets with AP shells from an ordinary tank?
    Possibility to hit is not that high but still huge fun when you get one down.
  12. krakanu

    krakanu Well-Known Member

    Messages:
    540
    Likes Received:
    526
    What if the unit in question uses energy to fire? Should it be wasting resources in a vain attempt to shoot down something it has only a rare chance to hit?

    Edit: Guided munitions: http://en.wikipedia.org/wiki/M982_Excalibur
    exterminans likes this.
  13. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    Requiring energy to fire is a good idea, but a different one. (Although I am in complete favor of making all weapons require energy for the operation, and if it was only to deal with the steam roll effect where you can crush the enemy at once once you have overcome the line of defenses in the first and only assault.)

    I have marked the important point in your post, before everything else, the chance needs to be adjustable. And not only a choice between the two extremes 0% and 100%.
  14. krakanu

    krakanu Well-Known Member

    Messages:
    540
    Likes Received:
    526
    If the projectiles are simulated, it is impossible to give a unit some fixed % chance to hit or miss.
  15. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    Try reading the last two pages of this thread again. It is possible, in a very natural way.

    There are actually two factors which can be used to determine a miss chance against ALL targets (accuracy) and against moving targets (jitter in projectile prelaunch time).

    You can't waste shots against dense groups with simulated projectiles, but you can do calculable miss chances against single units or sparse groups.
  16. GoogleFrog

    GoogleFrog Active Member

    Messages:
    676
    Likes Received:
    235
    No it's not impossible. Even if you can't calculate the chance to hit (which I find unlikely) then just simulate the situation a few hundred times for an approximate chance to hit. This does not have to be done on the fly, chance to hit could be statistically determined during balance testing.
  17. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    It is calculable. You can calculate average damage, hit chance, and derivation on the damage as a function of unit radius, movement speed, accuracy, base damage, AoE radius and prelaunch time jitter.

    A unit can say exactly "I have a chance of XX% of hitting the target under the current conditions and I'm going to do XXX damage on average. I have a chance of XX% to destroy the target with a single shot.".
  18. occusoj

    occusoj Active Member

    Messages:
    187
    Likes Received:
    34
    Give the player an option to decide if it should fire at targets that its unlikely to hit and to define "unlikely". May default it to something that feels like "off", firing only at stuff its very likely to hit, so that beginners dont get into huge energy stalling by units firing at seemingly hopeless stuff.

    A perfect example of what I meant.

    I too would like units to consume energy when firing. Have to build enought energy storages to rush someone to death in one go then. And such storages make up pretty decent targets for counter attacks. Opens new possibilities of gameplay, very nice ;).
  19. bobucles

    bobucles Post Master General

    Messages:
    3,388
    Likes Received:
    558
    Because the odds of hitting another unit are not static. A "silly" shot against one aircraft may do nothing, but that same shot against a thousand aircraft is guaranteed to hit something. If the shot is never taken, then no kill is possible.

    One of the huge mitigating factors for TA air was the massive increase in damage air units took as they grouped up. It was not just AoE damage, but the result of many many more tank shells successfully hitting anything in the air.
    Last edited: September 25, 2013
  20. exterminans

    exterminans Post Master General

    Messages:
    1,881
    Likes Received:
    986
    Your point is invalid.
    We already have flocking behavior in PA which ensures that even air formations are sparse by default. That means if you are not causing AoE damage, a missed shot will not hit anything else unless firing in plane. Sure, if 30% of the plane is occupied, you are also guaranteed a 30% hit chance at least. But that's up to the attacking player to choose a wider formation.

Share This Page