There Be Dragons: Slaying the Deathball

Discussion in 'Planetary Annihilation General Discussion' started by ledarsi, April 1, 2013.

  1. ledarsi

    ledarsi Post Master General

    Messages:
    1,381
    Likes Received:
    935
    Deathballs have been coming up in a lot of threads lately and I think the concept deserves some explanation. Deathball gameplay is arguably the single largest critical fatal flaw with most current RTS gameplay. I think it is important to have a PSA about what deathballs are, what causes them, why they are bad, and how to avoid having deathballs in PA. Modern RTS games are plagued by deathballs. They are everywhere like cover-based shooters. It has gotten so bad that much better RTS games are devolving, such as Starcraft II and SupCom 2. Hopefully PA can reverse this trend and show a new generation of RTS gamers what is really possible for a genuine game of strategy.

    I want PA to return to classic strategy form. Less emphasis on composition and deathballs, and more emphasis on large-scale strategy and force positioning, distribution, and movement. In my opinion if PA can manage to be a deep strategy game of grand strategy, territory control, position, tactics, and so on and NOT revolve around deathballs, then PA will be the greatest RTS game of its generation. Even ignoring its potential from creating a much larger scale, and having multiple planets, PA has the tech and philosophy to be the greatest RTS in terms of basic strategy fundamentals.

    It's simple. We kill the deathball.


    Background

    What Is a Deathball?

    A deathball is a very large group of units that is extremely strong together, and vastly weaker apart. In some games a player's entire mobile military will function together as a single deathball. In RTS games that encourage deathballs they typically dominate the field with wildly disproportionate strength compared to the same units spread across the map.


    RTS HP DPS Acronym Soup Primer

    Making this point clear calls for a primer on how RTS players conceptualize damage. Units in RTS games have HP values representing their current vital status, and their HP drops when they are damaged. However in almost all RTS games, a unit's HP status does not directly affect its performance; the unit remains effective until its HP reaches zero and it disappears from the field. This is a subtle but critically important point.

    Units in RTS games will also have weapons which deal varying amounts of damage. However not all weapons function similarly- some fire very rapidly for low individual damage, and others fire infrequently for high amounts of damage. To capture a general, abstract "damage output" value, RTS players talk about DPS, or "damage per second." Any type of weapon can be measured in its damage output per second, regardless of its damage and cooldown. The important point here is that DPS and the target's HP together create a "time to kill" factor. If the target has 100 HP, and your unit (or group of units) has 20 DPS, it will take 5 seconds for that target to be destroyed.

    In general, a canonical deathball results when a group of units together have the DPS to destroy an enemy force before the enemy has enough time to get through the HP of even one of the deathball's members, or at least can only inflict token casualties.


    Deathball Game Theory

    There is a second aspect to the nature of deathballs which is the reason they are so pathological in terms of gameplay. In terms of game theory, a sufficiently strong deathball being possible will tend to make it a dominant strategy.

    By dominant strategy, I mean that a deathball strategy will be strictly superior to other, non-deathball strategies due to explosive absolute strength of the deathball. If a "normal" army is a quadratic strength curve, a deathball is an exponential strength curve.

    Suppose we have a game which encourages deathballing. A force that is twice as large as another is so much more powerful that it can defeat the half-size force with minimal casualties. The other player must then also deathball, because splitting their forces makes them combat ineffective against a deathball. Trying to cover two different places each with half your forces will simply result in the enemy deathball steamrolling one of your forces, leaving the other half completely unable to defeat the intact deathball.

    This isn't to say larger forces shouldn't outperform smaller ones, or that they shouldn't win efficiently against smaller forces and take fewer losses than a smaller force. However they should sustain casualties in battle which require replacements, and which weaken the force incrementally with each fight. A deathball's hallmark feature is that it does not become significantly weaker after fighting unless it engages another deathball-like force.


    On Deathball Gameplay

    What Causes Deathball Gameplay?

    Deathballs emerge from players being intelligent about the types of assets they have, and how best to maximize their use. Players are smart; they will use their units to maximize their damage output and odds of survival. If a deathball is the best approach, players will rely upon them.

    So, what causes a deathball to be the best approach? Simply put, a deathball results when a force can stack DPS sufficient to prevent itself from sustaining casualties. Long range is the largest contributing factor to deathballs. Long range units in sufficient numbers can stack together and kill enemies forces of arbitrary size before they close to range to deal any damage. Splash damage also frequently leads to deathballs because increasing numbers of splash sources start to overlap in higher density, and allow increasingly tougher targets to be wiped out in large numbers.

    Nevertheless, otherwise conventional direct fire units can also be conducive to deathballs. High HP and low DPS units stack well into deathballs because a large number of such units together have a high DPS, and it takes the enemy a long absolute amount of time to destroy any particular member which does not change regardless of the size of the group.


    Why Are Deathballs Bad?

    Deathballs are absolutely terrible for gameplay. They constrict both players' deployment options down to where they place their single army. Deathballs are numerically large stacks of army strength that always must fight together. This means dividing your forces for any reason makes it impossible to engage the enemy army.

    Worse, in the likely event that one side simply has more units than the other, that player's victory is almost assured. The game is akin to a chess board with only one piece per player, with a number attached to that piece indicating its strength. The player with the stronger piece can capture the other's, but not vice-versa. It is nothing short of ridiculous.

    And worse yet, once a decisive victory is won, the deathball is likely to be largely intact. The other player's army is completely destroyed, but the deathball is still combat effective to a significant extent. It is virtually impossible for the player who lost that single decisive battle to rebuild a force large enough to ever win a battle, especially since the deathball is likely growing also.


    Deathballs and Composition Rock-Paper-Scissors

    Many RTS games are not even aware that having a single blob of units is an issue. Many RTS games take it as a matter of course that you keep your entire army together. These games rely on composition to create "strategy."

    A complex web of unit counters is implemented, with certain units designed to be effective against other units. This creates the pseudo-strategy situation where a smaller force can defeat a larger force, provided that the smaller force "counters" the enemy army. The theory is that you can send 5 Paper to counter 10 Rock at a location, leaving you with 5X to use elsewhere. In practice, it is just more effective to always make a large mixed-composition deathball. The ideal composition of this deathball is indeed a complicated calculus depending on the relative strength of many options and strategies, but "standard play" deathballs will emerge in only a few days or weeks which will then be widely copied.

    The relationship between deathballs and composition-centric RPS gameplay is another involved subject. Long story short the system is made slightly more opaque, but remains a deterministic numerical contest between two armies.


    Slaying the Dragon

    How To Avoid Deathballs

    There are a few good guidelines to avoid deathballs. The first is to have large numbers of units which die easily. Very durable units take either more units or more time to kill. And by having a large group of these durable units, it is possible to deprive the enemy of time using their own firepower. This leaves the enemy with only one option; use an equally large group of units to engage. It should be possible to engage a large force with a small force and inflict casualties before being destroyed, and the best way to create this dynamic is to have units that die quickly.

    Conceptually related to the concept of having squishy units is having very numerous units that are easily constructed, and which don't fight together efficiently. They get in each other's way, they block each others' attacks, their range is short relative to unit size or force size. The entire group consumes a lot of space relative to their movement speed, and so on.

    The general rule is to avoid any mechanics that run away with escalating or explosive efficiency or power. Mechanics that run away with scale, coupled with an arbitrarily scalable economy, will prompt players to adopt strategies that have the runaway mechanic as their centerpiece. Most types of units or assets should be designed to become less efficient with scale. For example, by having short range relative to unit size so units in the back of a blob are not participating in a fight, or making units unable to fire through each other so they can fire together perfectly. Even as the larger force's absolute strength increases with each addition, each extra unit should add less marginal utility or strength than the one before it, not more.

    A common misunderstanding about discouraging deathballs is that splash damage per se disincentivizes deathballs. While splash damage does punish tight clusterings of units, it does not necessarily discourage having a deathball which engages together. The player with the blob might be able to split their units against small splash damage (micro), or might be able to neutralize the splash damage source specifically by some method. Small splash damage doesn't necessarily discourage the enemy from deploying lots of units together. Large splash damage such as readily accessible nukes does discourage an enemy from engaging with a large number of units in the same local region. However if large-scale splash damage (such as counterable nukes) is used, then that could actually encourage deathballs by mandating including the countermeasure with the ball, which may be expensive, and require an army be a certain size or larger before it can warrant having such an expensive defensive countermeasure.


    Removing Deathball Gameplay

    The only way to actually slay the dragon and remove the deathball gameplay pathology entirely is to have all units in the game be less efficient to use in very large groups. Which raises the question of what constitutes a "very large group."

    The meaning of a deathball within a specific game will depend on the significance of a single unit. If an RTS game's units are small, cheap, squishy, and expendable, then a "deathball" will consist of an exceedingly large horde of such units. If an RTS game's units are relatively significant, large, and expensive, then a deathball might consist of relatively few units, such as a Starcraft II deathball composed of as few as 30 or 40 discrete units.

    If PA is designed expecting players to make thousands of units in the course of a regular game, then a deathball would consist of thousands of units engaging together. And as a result it is much easier to design such units to avoid high-efficiency engagements than if there are only a few dozen units. A blob of a thousand units is much easier to design to create inefficiency than a small group of only a few dozen units. It is also much easier to cause such a large group to sustain casualties against small groups. In fact, such an army might be so large and occupy so much space that it would actually take a considerable amount of time for members from one side of the army to move to support the opposite side. In such circumstances, it makes far more sense to split such an inefficient army into pieces to perform separate tasks in different locations, where each smaller group will be more efficient.


    Good Design

    Good strategy game design is about positioning and moving pieces and their interactions with your opponent's choices for how to position and move their pieces. Naturally this type of strategy game is more interesting when there are more "pieces" on the board, when we consider a force of units as a single piece.

    A deathball is a single extremely strong piece, which is necessary if your opponent also has only a single extremely strong piece. However if having so many units together in one place does not produce an extremely efficient piece, such that the opponent can gain more total utility by creating a large number of weaker pieces, then neither player will deathball.

    When neither player is using a deathball, then instead of scouting your opponent's deathball to determine their composition, you are scouting to find where their forces are, and how strong each is. Both players want to engage the enemy at specific locations (which likely differ) with barely enough forces to win locally. Using too many is inefficient because it likely may lead to a defeat elsewhere. Using too few may lead to a defeat unless a player maintained some reserve forces. Strategic concepts and practices like reserves, reinforcement, limited engagement, retreat, and others will immediately appear once a player is encouraged to exercise more thought than simply taking every single soldier they can muster and push all their chips onto the table to straight up defeat the enemy deathball.


    Takeaways

    A deathball is a very large group of units which engages together, frequently being the only military presence a player has on the board. For a too many reasons to mention, deathballs are boring, deterministic, pathological gameplay.

    It is my assertion that PA should consider every game mechanic, unit, and structure from the perspective of avoiding deathballs. Everything from the economy to unit properties should encourage territory control and judiciously spreading your pieces across the map. Especially considering PA wants to have large maps, even multiple battlefields and planets, it is critical that players must have a presence through space, and avoid just consolidating their entire economy into a single deathball and marching into a single enemy-held main base to win the game.
    Last edited: April 1, 2013
  2. Polynomial

    Polynomial Moderator Alumni

    Messages:
    1,680
    Likes Received:
    53
  3. joe4324

    joe4324 New Member

    Messages:
    36
    Likes Received:
    0
    ..................
  4. ayceeem

    ayceeem New Member

    Messages:
    473
    Likes Received:
    1
    Include full friendly fire as is present in Balanced Annihilation and Zero-K. Preventing blobs from shooting through each other would solve much of the cause of power increase from force concentraion.

    Don't make every map feature into wide open fields, which allow massive forces to freely maneuver.
  5. meltedcandles

    meltedcandles Member

    Messages:
    101
    Likes Received:
    1
    What is the point of replying? that speech has everything, the problem ways the problem is formed, and a solution to said problem! :D [​IMG]
  6. RCIX

    RCIX Member

    Messages:
    664
    Likes Received:
    16
    It's still important to make sure a lack of replies is not mistaken for a lack of interest.
  7. Raevn

    Raevn Moderator Alumni

    Messages:
    4,226
    Likes Received:
    4,324
    I agree with ledarsi's premise, however if the goal of this game is to make a strategic, rather than tactical, game I believe the Low HP "large numbers of units which die easily" method should not be used. Having the smallest advantage in a field with low HP units means a massive shift in favour, as there is no opportunity to respond, retreat, reinforce or use attrition, because battles are over as fast as they begin. Battles become so decisive as to make a single mistake in a matter of seconds cost you the match. These are all the very opposite of a strategic game, and they raise the difficulty enormously for new players.

    In addition to preventing a single blob being far more effective than it's numbers would indicate, I believe the ideal solution should also taken into account:

    a) the favouring of greater strategic use of forces, both before (army positioning) and during (manouvering, retreating, reinforcing) combat.
    b) Mistakes should punish a player, but not destroy them immediately; skill should be given a chance to play a part in recovering from a setback.

    My view is that staying power of an army is important in making the game strategic rather than tactical (a). This is obviously mutually exclusive with purely low HP units (relative to DPS). Several of the methods ledarsi spoke of would still be able to be applied however (lower ranges, larger overall quantity of units necessitating splitting). (b) is necessary to prevent people quitting after a single battle - there should remain as late as possible the opportunity for a losing player to succeed.
  8. GoogleFrog

    GoogleFrog Active Member

    Messages:
    676
    Likes Received:
    235
    I agree with ledarsi and like his points on what would make a large-scale-manuever-type game with just the basic mechanics from TA style games. As in he isn't adding mechanics such as supply lines.

    raevn I don't think you and ledarsi mean the same thing when you say strategic and tactical. To me army positioning and maneuvering with the intent to fight in a single battle sounds tactical, at least in this context. Many armies fighting in a single battle isn't quite a deathball because you can involve flanking. But a single battle with flanking is not what ledarsi was talking about. His post is about many armies positioned far away from each other taking part in many different battles.

    Low unit HP for anti-deathball seems unintuitive but I think I can make sense of it with some extreme examples.

    Say your units have such low health that they one shot each other as soon as they enter range. If a large army with X units and a small army with Y units. They attack each other and one by one pairs of units come into range and annihilate. The large army will be left with X-Y units. The extra units in the large army didn't make it fight more efficiently.

    To take the other extreme let each unit take a long time to kill itself. In this case most of the damage dealt throughout a battle is dealt after all the units in the battle are in position and applying their damage to an enemy unit.
  9. veta

    veta Active Member

    Messages:
    1,256
    Likes Received:
    11
    you lost me here:

    That doesn't really make sense, especially when you're going with the StarCraft example. StarCraft battles last all of what 10 seconds and once you engage there is no retreating. High HP to relatively low DPS leads to map control being important, you see a lot more position play with High EHP (effective hp) Low EDPS (effective DPS). TA was a high EHP low EDPS game once wrecks were on the field because of how much they obstructed unit fire. Now, if you have High DPS and low EHP units like the protoss Colossus then you will want to mix in other units that have relatively good hp even if their damage is bad to sponge enemy fire. If all units have high EHP (effective hit points) and relatively low damage then this is a non-issue.

    LABs are an example of a low hp high dps unit, they generally don't last long and once they engage they usually aren't long for the world. If you want examples of very high HP and relatively low DPS look at the commander unit, a command takes a long time to kill another commander.

    The reality is low hp to dps ratios force players to ball up their armies in order to secure a region because a few units alone will quickly be destroyed by a slightly larger contingency. This is the case because as units quickly die off in low HP to DPS ratio engagements their collective DPS also decreases. Units that have higher HP and lower dps can use delay tactics and wait for the cavalry to arrive. Therefore a player with less units can effectively control as much or more territory with clever tactics. This is NOT true in games like SupCom, and it was true in TA because of the EHP increase that occurred on regularly fought over regions, which is also why TA developed frontage as the game continued. You rarely saw players exchanged armies over the same ground for very long in SupCom because there was no major EHP increase as the game went on.

    What drives armies apart should be map control, ergo you need your troops in different places to maintain map control. Without map control TA would just be 1 giant army vs another giant army constantly which is what StarCraft 2 devolves. In StarCraft 2 if your army is at home it will probably win due to shorter supply lines, if it isn't at home and you're being assaulted there ARE NO DELAY TACTICS you can use for your army to arrive home, your only play becomes assaulting their now undefended main base, e.g. a base trade. What do I mean there's no delay tactics? I mean there's nothing you can do to prevent the enemy from crushing your workers and any contingent military forces in the vicinity before your main army arrives. If units had high hp and relatively low damage you would have a lot more opportunity to keep your base in tact, and base trading with death balls wouldn't be your only play.

    The key to discouraging death balls is emphasizing map control, which starcraft does not do. Luckily MEX points do emphasize map control and that's why you don't see scenarios play out on TA and FA as they do on SC2.

    I'm not sure what you mean about composition affecting strategy. Of course composition should affect strategy, different compositions should be used through out the game to throw off your opponent - that's tactical production. Oh they have no AA in their powerful land army? Better throw in some gunships. That is the cornerstone of dynamic gameplay.

    Yes in practice this leads to players using mixed armies, but guess what - if you're careful and tactically clever with your armies you can do damage with surgical unmixed attacks. Spot some undefended MEXs in the rear of their base? Send in the luftwaffe. See a base with few units and lots of pd? Send in the kaiser's artillery corp.

    "Large numbers of units" together are deathballs no? I'm not understanding here, what you're describing is what occurs when units have range that is relatively long to how much space they take up, such that they can all engage at the same time - which is the problem in StarCraft 2. When units in a ball take very little time to all get in position to fire upon each other AND they have a low HP to DPS ratio the most intelligent play is to bunch your units up together. What you are advocating is units to take up space such that you need formations to effectively use unit range and DPS. That's not a bad thing, but it has little to do with what you describe as the problem.

    Essentially what you are describing as a solution, which I might add is to a problem that doesn't exist, is that units do so much damage in their initial volleys that they will decimate opposing forces before the entire enemy force is in position to engage. The reason this is bad is because it leads to relatively equal trading in early exchanges and ultimately resolves to battle to whomever has the most troops. Battles are also over in a matter of seconds in such a system which is in my opinion another problem in StarCraft 2 that should not be emulated in PA. There is no battlefield management in StarCraft 2, you engage and you either win or lose - you have maybe a few seconds to disengage. If you employ a similar low hp to high dps system in PA the battles will function the same way, you either engage and win or lose or you don't engage and just stare angrily at your opponent across the map. Battles should last long enough for you to actively affect the end result through military manuevers, feints, delay tactics, reinforcement, etc. And that is only possible if the EHP (effective HP) of units is much higher than their relative DPS. When units can 1-2-3 shot each other you get death ball gameplay - players keeping all their units together such that any enemy engagement will meet your full force. Low HP to DPS ratios make the game very tactical, how do I know this? Play a game where units can 1 shot each other and see how strategy that feels.

    This doesn't even make sense. StarCraft has supply caps and deathballs are usually the entirety of a players supply cap. If players could have 1000 supply StarCraft wouldn't play any differently, you'd just see 1000 supply deathballs. You're also confused on the reality here, StarCraft units ARE low HP to high DPS units. A marine kills another marine in 6 game seconds and in StarCraft units can't miss so it's even less than the equivalent in a simulation style RTS like PA. In Forged Alliance even the lowest HP to DPS ratio units Light Assault Bots can fight for longer than 6 seconds against each other. That's because LABs speed makes up for their low HP by increasing their EHP.

    Let me reiterate: StarCraft II is low HP to high dps, especially on fastest speed which is what everyone plays. TA is high EHP (even if HP numbers can be low) to relatively low EDPS (effective DPS). Nobody remembers how epic all their units killing all their opponents units in a few seconds was in TA, they remember how epic long fronts and battles were. We don't want short battles in PA, we want long battles like what TA gave us.

    100% agreed which is why we shouldn't mimic the lower EHP to EDPS model that Forged Alliance gave us as it results in exponential management over battle management, skirmishes over pitched battles, and shorter battles hyphenated by standoffs instead of frontage like TA. Battles need to last longer than the 10 seconds they do in StarCraft 2 and that is accomplished by increasing EHP to EDPS ratios. This is even more so important in PA where it is likely you will be dividing your attention across multiple battlefields and will need to order your units, look elsewhere and then return to the ensuing battle to manage it. In your model your units would be dead before you returned to the battle.


    Not only are deathballs not going to be a problem in PA because of the emphasis on Metal Points and map control in general, I'm not sure if you realize how deadly effective such a system would make kiting units and kiting micro in general as it already is on Forged Alliance. The problem of kiting because magnified magnitudes greater when you are not able to pay full attention to every single battlefield and so if UberEnt wants to emphasize large games and huge playfields they probably shouldn't use design heuristics that make microing and kiting demolish hordes of other types of units alla StarCraft 2.

    Have you played TA or Forged Alliance? The deathballs you're describing aren't really possible on most maps. And yes it is map dependent, because it is MAP CONTROL not the ratio between hp and dps that affects the natural tendency of keeping your army in tact. If anything units dying in less than 5 shots from each other would ENCOURAGE balling up armies such that you can deliver as much damage as possible to any approaching force before they deliver recourse.

    I somewhat suspect this entire thread was in response to my own thread as well which covers a similar topic and where your points were met individually already. To anyone who actually read this post it's pretty clear you're using the revulsion "deathballs" instill to persuade people that doing things your way is the only way to avoid deathballs. To me it appears as though you're inventing a problem and then providing a snake oil solution.

    viewtopic.php?f=61&t=44859&start=40

    Or maybe you legitimately believe that StarCraft 2's Stalker unit is the problem in StarCraft 2 -- and not the insanely high dps to cost, low hp, long range Protoss Colossus unit, or the equivalently silly Zerg Infestor. Or maybe you would cede that the Terran Marine, by all counts a very high DPS, low hp for cost unit is the problem? StarCraft 2's problem is that it tries to add high dps units in the late game and balances their impact with a high cost and low hp. This results in HP padding with units that have higher HP for cost in order to ultimately get a balled force of the highest possible HP and DPS. The reason the lower HP and higher costs of these, if not imbalanced, poorly designed units does not offset their DPS is that map control and terrain in general are not very important in StarCraft 2. The difference is very clear if you played even StarCraft 1 competitively which like TA did not suffer from death balls either.


    googlefrog:
    This is a x-post from the thread I linked and discussed a very similar topic:

    I think it's important to understand that strategy DOES occur on the battlefield and it is MORE than building an army and then clicking the enemy. Let me give an example:

    1) A player scouts his opponent and notices asevere lack in anti air
    2) The player then proceeds to make bombers to wreck vulnerble targets
    3) The player sends out his bombers
    4) The player spots an enemy interceptor and brings his bombers home

    Where was the strategy here? Was there strategy here? Obviously 1) is just intelligence gathering. 2) is clearly a plan but it seems pretty obvious - is that really strategy? 3) is just issuing orders, that's more like a task and 4) is just the player retreating which is a tactic.

    The answer is all four make up strategic play, the player formed a plan based on intelligence and then tasked his units to execute until a tactical retreat. The whole thing is strategic play. Yes in StarCraft you can impact the game with a properly placed spell or you can win the game with a properly placed Tactical Missile in TA. Both are at a tactical level but it is the recognition of weaknesses and advantages that makes game play strategic. In short strategy is the evaluation of the battlefield and execution is what separates the pretenders from the contenders. Building a base and managing your economy is strategic in a literal sense, in that your economy directly affects your military capacity but beyond that economy is not interesting or strategic. And economy and base management should not be the focus of a game trying to live up to TA.

    Strategic play is about adapting to the situation as it becomes apparent to you and that's exactly what battlefield management and that's what a successor to TA should be about, not massing a bunch of units and watching them blow up a few seconds later.
    Last edited: April 1, 2013
  10. bmb

    bmb Well-Known Member

    Messages:
    1,497
    Likes Received:
    219
    The thing about deathballs is that having more stuff in one place will always be more powerful than spreading it out. It's simply the nature of the beast. You can't "fix" it.

    What you could do is mandate some requirement to actually have units go in different locations. But this is entirely up to the player and what the player would usually do is build a base and defend that. So what you end up with naturally is base defences vs death balls. It's simply unavoidable.

    It should be noted that a natural weakness of a death ball is the things that it is currently not doing. Since it can only be in one place. You can also naturally split units up with good level design making some positions have several paths, important locations and vulnerabilities that all have to be accounted for. But at the end of the day in 1v1 it is commander vs commander, base vs base, planet vs planet. And the most effective tactic here is deathballing.
  11. sylvesterink

    sylvesterink Active Member

    Messages:
    907
    Likes Received:
    41
    One thing I've noticed is that all the more strategic games I've played tend to work around the concept of units in line formations. This makes sense conceptually, because you're bringing the greatest amount of firepower to bear on the enemy, while providing the greatest amount of protection to your units. (Assuming this is a combat situation.) The problem is that there's a certain point where a line of units loses its effectiveness, as they spread so wide that they can't all fire on the target, so at that point, it's best to add multiple rows. Of course, this reaches a point where the row can't get any thicker without again losing effectiveness.

    It's at this point that it's more advantageous to create a second group to augment your first. This gives you the ability to attack two points instead of one. So now, with this consolidation of power into these lined formations, the goal of attack is to attack the weak point in the lines, which is the side, or flank. Such flank attacks are especially effective because the line formation can't bring its full firepower to bear as it should. Such games end up bringing grand-scale tactics to the floor.

    Zero-K does an admirable job of this, and I suspect that even without the friendly-fire aspect, it would still be most effective to play the game like this. In Supcom, it wasn't as evident, as mobility tended to be the preferred method of engagement. As a result scattered units that should have been wiped out by enemy line formations generally performed better. Some of the Total War series, and even The Battle for Wesnoth follow this concept, and it does a great job of making the game less about deathballs and more about tactical unit play.

    Starcraft, and similar games, can't achieve this type of gameplay because their battles are focused on such a small scale that getting units into firing range is quick and simple, meaning such formations wouldn't be as effective.

    So really, the key to mitigating deathballs is to make them less effective, and the concept of line formations (and formations in general) is one example of a potential solution. The key is to ensure that such formations DO actually provide a benefit, much like they do in real life.
  12. yogurt312

    yogurt312 New Member

    Messages:
    565
    Likes Received:
    2
    Wreckages are basically self forming trenches, units at the front die and create a movement block and a projectile block.

    As for the deathball, I do not see how it could effectively be in this game. You cannot mass enough units in a certain area so that they can effectively apply their dps to smaller targets, the extra units become reinforcements for the front line. The thing about the infamous starcraft deathball is that is takes up such a small amount of space that allows you to instantly put all your dps out instantly. With 200 tanks, the front lines will fire and you will lose some, things will become congested with wreckage and then perhaps half the second line will fire, by the time the back of your deathball enters combat 20 seconds after your front line... well damage has already been dealt and small harrasments will have dealt their attrition. Also mobile artillery deals good harrasment, an option not really available in starcraft (for instance by the time a siege tank has deployed you've simply moved, in this you might notice by the time the first volley hits), similarly carpet bombers might only get one run but if you concentrate on massing as much dps into the smallest space you can that first lot of bombs will hurt a lot.
    Last edited: April 1, 2013
  13. asgo

    asgo Member

    Messages:
    457
    Likes Received:
    21
    my impression is, that with multiple battlefields/planets the deathball issue wouldn't be that much of a problem in PA in general. If you can't wipe the whole map with your deathball it's probably a waste of resources since the units are missing elsewhere. If you have the resource superiority to buildup deathballs on every contested planet, then you would probably have won anyhow.

    If additional mechanics like friendly fire/line of sight and formations are applied sensibly, you can probably avoid the deathball problematic altogether or at least reduce the motivation to use such tactics considerably. At least such mechanics are intuitive enough so that the player can easily reason out, why his big blobs are less effective than used in smaller portions.

    What I wouldn't do is messing around with unit stats to get a change on that level. Unit stats have to many global implications on different balancing issues to make them a good choice for modifying the combined behaviour of groups of hundreds of units.
  14. amphok

    amphok Member

    Messages:
    139
    Likes Received:
    13
    easy fix, just add more units with enormous AOE damage, Brood War(starcraft 1 expansion) it's the only game that got rid of death ball, in an effective way
  15. yogurt312

    yogurt312 New Member

    Messages:
    565
    Likes Received:
    2
    Thats what i mean, the game is fundamentaly non-conducive to deathballs.

    They cant really ball up, they are easily harassed by artillery and bombers and they can't apply their full compliment of death.
  16. bobucles

    bobucles Post Master General

    Messages:
    3,388
    Likes Received:
    558
    A nice introductory thread explaining the death ball and how it can be avoided. It's a bit thin on solutions, though.

    The hardest part of solving a death ball is creating valuable reasons to spread out. PA inherently has some solutions in the form of extractor points, volatile construction, energy demand, large footprints, and super weapons. Most of these answers help prevent turtle fortresses and limit the ability to concentrate power in any one base. Solving death balls with mobile units will be a whole different beast entirely.

    I have posted many possibilities in other threads, mostly AA utilities and restricting mechanics for air units. Achieving the same goal for ground units will be very difficult.
  17. cola_colin

    cola_colin Moderator Alumni

    Messages:
    12,074
    Likes Received:
    16,221
    The initial post is good, veta adds a few point I wanted to add.
    Mainly:
    - The Unitdesign does not matter very much
    - Having resources spread all over the map (TA/FA-Style) extremely discourages Deathballs.
    - Larger forces will always easily kill smaller forces. That has nothing to do with deathballing.
    - At some point forces get so big in FA (TA?) that they cannot all shoot at the same target at the same time. That's the point where it wont hurt much to split up your army.

    I really don't see a problem with this. TA and FA both had not problem whatsoever with deathballs, thanks to the fact that you needed to hold mex spread all over the map.
    Starcraft 2 has as many deathballs as it has because you only need to hold a very small area. Starcraft 2 with mexes would look very different.

    That's all about ground, though.

    Air is much harder imho: Controlling the air with airunits tends to be all deathballing, because air units stack on top of each other perfectly. I can't see how you could possibly solve this without radically change the way air-units behave.
  18. superouman

    superouman Post Master General

    Messages:
    1,007
    Likes Received:
    1,139
    From my understanding of Starcraft II, deathballs were the consequence of the pathing which let players to have a large number of units on a small area.
    When you add a quite large range of the units, the entire ball of units could shoot at the same time which so the DPS per m² (if we can call it that way) was very high and it was not worth having multiple armies because a single one was much more efficient.


    Ps: I was crying of joy when i was watching the stream about flowfield and the groups of units moving in line.
  19. Devak

    Devak Post Master General

    Messages:
    1,713
    Likes Received:
    1,080
    I don't entirely see the problem of the Deathball. In many cases in FA and TA did i have armies so large, the front row was in hefty battle and the back end didn't have the enemy in it's "vision range". Playing around with FA, with no T3/T4 enforced, shows a very different game from the normal "build super experimentals". For one, Deathballing quickly becomes ineffective due to splash damage and the balls reaching "critical mass".


    The problem for FA completely lies in Air blobs, and that is a problem the Uber people don't want and they're fixing that for PA. There were very different problems in Supcom 2 so that doesn't really matter.
  20. Malorn

    Malorn Member

    Messages:
    82
    Likes Received:
    14
    I think that TA style gameplay does indeed weaken the idea of deathballs. Still, this post is effective and elegant in making it's point, a point that ought to be made. Nuff said.

Share This Page