There Be Dragons: Slaying the Deathball

This thread... Has been so hijacked by itself that the people arguing on the whole don't seem to realize that there are two arguments going on... Probably due to Ledansi having really lengthy posts and people tl;dr ing him.

Blob of loads of units: probably gonna happen, it's good end game play (on one planet at least), and there are various, already balanced ways of approaching it, nothing needs to be done to stop it, and anyone who thinks it needs to be stopped probably doesn't realize that it is generally a hallmark of having lesser skill than your opponent.
Most, not all, of the people who are arguing against Ledansi are arguing about that.
But that's not what Ledansi is talking about.
EDIT: Well ****. It looks like I changed my own mind in the course of this post.... You can ignore this blip, it's pretty contradictory to the rest of my post...

Ledansi is talking about specific unit combinations that only get EXPONENTIALLY more rewarding the more units you pump into them.

To take it from what people have said about Samsons and Peewees, from what is said about them, in a 1v1, a single Samson would win, but come away heavily damaged from an encounter with one Peewee. Let's say that is because the Peewee gets into it's own attack range after 4 hits from the Samson, but dies after the 6th hit. To win with the Peewees, you'd need 2.

So now, if you have 6 Samsons, you can suddenly start destroying Peewees in a single shot and destroying up to 3 Peewees at a time before they get into firing range. So suddenly the bare minimum of Peewees you need to even damage 1 of those 6 Samsons becomes 4 at once, and even then the damage that they cause to that 1 Samson would be minimal. And with Perfect micromanagement, that means you'd be able to do the same damage with 6 Peewees, but when I say perfect, I mean six separate forces entering the range of the moving band of Samsons at almost the same time, to counter pointing and clicking with those Samsons and telling them to fight along the way. So effectively, I'd say it would take more like 24 (estimation, but not exaggeration) Peewees to overwhelm the Samson's rate of fire enough to do a similar amount of damage as the initial circumstance. You'd need 32 Peewees to win this encounter.

That event alone is sorta alright, the real problem is that the benefits don't stop there. Scale that up a little, at say the middle of the match, and quadruple the number of Samsons to 24, at this point it starts to get obviously broken. With 24 Samsons you have the ability to one-shot 4 Peewees at once, which means you can destroy 8 Peewees in 2 hits, 12 Peewees in 3 hits, and a whopping 16 Peewees with 4 hits each simultaneously before taking any damage to even a single Samson unit. But now, that mob of Samsons has, most likely, doubled, tripled or quadrupled their effective firing rate, from differences in position staggering their detection of oncoming forces and because how small a portion of the group is required to eliminate the threats. And with that change the number of Peewees it would take (without micromanagement) to overwhelm the rate of fire to damage 1 Samson will have doubled, tripled, or quadrupled what it was down with 6 Samsons. I'll go minimalistic and say firing rate has only doubled. So with brute force and no micromanaging it now takes about 288 Peewees to do the same percentage of damage that 1 Peewee does against 1 Samson. To win this way you'd need 384 Peewees.

But, the issue there becomes multiplex. For one thing, with these numbers, micromanaging is a really easily possibility for anyone playing and 288 Peewees would most likely be enough for anyone to win that battle, against 24 Samsons, with a little thought, so skilled players could push that number down to a force 90 to 80% that size. But on the other hand, 1 or 2 spare Samsons could irreversibly turn the battle back in the Samson's favor.

The real scary part of the last iteration is that 24 Samsons could easily survive through a dozen or more skirmishes with forces of Peewees that number even up to 100, and with small refreshments at a regular rate, those 24 Samsons could likely do that even if their enemy was micro managing his units like an expert in every encounter. So depending on the cost of Samsons compared to Peewees, the Peewee force could take over more land, make more units, have a better economy, be an expert at micromanaging, and still be utterly wiped off the face of the map if the Samsons ever managed to get an offensive force of 24 units, and a defensive one of 6-12 to guard a relatively small and simple base set up to give him the bare minimum to pop out a Samson every minute or so, before the Peewee side managed to amass any force large enough to contest that.

Now apply that to cost of the units, if Samsons cost twice as much as Peewees, then you'd only have to build 6 of them before they become at least doubly efficient as Peewees, so there's no reason to not just set up pods of 12, 24, or god forbid 48 of them wherever the hell you please, because they are so much more effective that spreading them out makes you less vulnerable (assuming that their only enemies are Peewees at least). And if you make Samsons worth quadruple the cost of Peewees, the same thing happens, but the first efficient iteration is now raised to 12, which is still easy enough, so the problem persists, and continues to persist indefinitely. Even if the cost is 12 times that of the Peewee, it's still efficient to make 24 of them, just really hard and it would feel rather pointless because you would rather be the side that made 288 Peewees instead because until you got 24 Samsons, your base is fair game for your enemy to destroy. Meaning that the way Samsons and Peewees interact is balance destroying and cannot really be fixed, just shoved around leaning towards one side of the other. (It can, easily, in fact this situation balances itself the instant that both sides can make both units, because this is just a classic example of range versus power, but for illustration purposes, just go with it)

tl;dr #1: In a world of only Samsons vs Peewees, Peewees become useless unless Samsons are so expensive that they are useless.


Now take that information then imagine that instead of 6 Samsons, you can have any combination of units that exhibit similar problems of greater and greater return with every additional iteration (what our author has somewhat aptly called deathballs), and instead of Peewees you have all the other 90% of units and their combinations. That is why he wants to avoid the deathball, and what a lot of you are misunderstanding is that a deathball could be made out of as little as a dozen different units if an exploitable circumstance slipped through the cracks, they are fundamentally far more efficient than the alternatives in a straight battle sense, and they only continue to become more and more efficient the larger they become, but they could also potentially last through dozens of battles, unscathed, only exacerbating their efficiency. Rebalancing the costs and DPS and HP of what goes into that small deathball, to make it larger, clunkier and more expensive, just shoves the problem to later in the game.

When structured like that, with deathball being the most efficient way to take over a planet, it becomes a game of economy. And for that purpose territory is only important for resources, if you can stop your opponent from getting off the planet, then territory has a definite limit to it's usefulness, if you cannot stop them from getting off the planet, then you just need to make sure to destroy all of their remnants so that you can exploit the resources of that entire planet in relative peace. They might attempt to raid or to establish a foothold, but with a standing million man army and the resources of an entire planet at your disposal, it's a minor annoyance at best, so you just have to worry about asteroids crashing into your planet.

This is the cool aspect of PA's potential, that having a million man standing army on one planet is just this side of useless if your enemy is off the planet. Especially if we include some of the cool ideas I've seen in this forum, like orbital drop factories. So I agree heavily, that what everyone has been saying about masses of units is absolutely correct, and I imagine this game will make that standpoint even more absolutely right than it has been in the past. So the easy call to make is that, if you are worrying about deathballs of thousands of units being game breaking, then you are worrying about the wrong thing, because by the time someone pulls together an army that size on the ground, you could have easily built one of your own, or set up on an asteroid to destroy that planet, or one of a dozen other things.

I get it though, because deathballs of that size would sap all the fun out of the ground end game on individual worlds, it would all become "who can build up the resources to make something unstoppable, first" and an economy game, with your major skirmishes being over crippling your enemy's ability to get off the planet, and attempting to cripple their war machine, and territory control would become something you only need mid-game on the ground, by the end game of individual planets, all those skirmishes and territory control matches would stop as massive balls of death move across the map, the losing side would likely be smart enough to turn tail and run and they could most likely get off the planet, but this wouldn't fix the problem of the end game on the ground. (because that is the bulk of this game, there is no space conflict, and other than destroying planets and bases, the only reason for setting up outside of a planet is for effecting the combat on the ground)

tl;dr #2: But, the real reason I've gone and looked at all of this, is because I mostly agree with Ledarsi in the fact that this is a problem, and that end game deathballs are just as broken and painful as early game deathballs, they still represent a fundamental flaw in balancing, it's just that the difference in scale means that there's more interesting "balanced" things happening before the end game where broken mechanics surface.


However, calling that Broken is wrong, because the competition up to that point is real, cutthroat, and well developed, and generally speaking, equal opportunity is given to both sides.

What it does do however --no matter how you split it up-- is obscure two critical parts of strategy and tactics that have NOTHING to do with economy: Positioning and Formation Tactics.

In a game where winning at strategy means winning at economy, there is no reason to fight enemies off of their mexes or to destroy their generators beyond messing with their economy or expanding your own, which is just Territory, scrap and wreckage that represent common battlegrounds and become the frontlines don't add much to Positioning (they do mean that a frontline is a great place to make a last stand, which is Positioning), they primarily polarize Territory. The idea of an advanced base, closer to the field of battle is barely useful, and having fabricators that aren't heavily defended and centralized is pretty much only a detriment when your enemy has a death ball. And Formation Tactics are obscured because they become relegated to micromanagement that on a small scale is far too difficult and completely useless, and on larger scale, when deathballs start popping up, the micromanaging becomes more useful and easier to use, in that it's much less like a micro game, but the nature of death balls, is that they compound the difficulty of that micromanagement, (which I actually addressed earlier, in my example, slightly indirectly, sucks to be you tl;dr'ers) and they lower the likelyhood of any reward for it, because what you'd have to micromanage would be so much like a deathball, that the couple extra seconds to pump out enough to complete it are usually worth it. (since deathballs are slow)

Why should you care? Because end game in regular RTS titles that take place on one plane, in PA, is just mid game or early game.

The funniest thing, is that TA, or at least Balanced Annihilation, seems to have already addressed and solved the issue of deathballs, and PA will likely follow a similar scale, but both sides disagree on how TA solved it. However, from what I've read (never played either, Nerd^3 and the Kickstarter video made me an instant convert) TA doesn't seem to address Formation Tactics or Positioning, beyond that.

tl;dr #3: Yeah... If you didn't read it all, I don't blame you, I've got no clue how all of this came out of my head. I've only ever played one RTS game before, (Impossible Creatures, not the best game, but fun) and I wasn't very good at it. I'm likely wildly wrong, or condescending on elements I don't know that much about in this post, and after reading a few of the latest comments, it seems that what I'm addressing here has likely been addressed elsewhere, better... But I just dumped probably two hours of my life into this, stream of conscious, and I think I've got some good, solid points, so I'm posting it anyway.... lol at starting out by saying Ledarsi did some tl;dr posts, then posting one as long as his opening post and less organized...

EDIT2: If you ever saw another EDIT2, it never existed. (It was just me being confused by another topic)

An other problem deathballing has with spread out resources is, that since it can't defend them very well, grabbing them is very difficult in the first place. (Unless the other player is passive, which he shouldn't be because being passive is bad.)
Basically deathballing isn't only sacrificing map control, but also economy, which means, that it either has to win with a well-timed push or it will simply lose horribly since the other player can utilize his better economy and catch up on army strength.

The strong link between map control and economy may not be the silver bullet, but it rips pretty heavily into deathballing.

Also, a deathball who is sacrificing everything for an attack by neglicting defense can be defeated by first killing the undefended economy therefore halting production and spamming static defenses to slow it down and chip away at it while constantly denying the engagement until it is weakened enough to take it out.
Of course you can't stop any size of a deathball this way, but getting one big enough for this to not work should not be possible because of the lacking map control/economy.

But this is very theoretical, we need the game to see how it will go. And I still think deathballs in PA will have enough weaknesses to not make them dominate the game.

Yes, I realize this, that's why I said;

Imagine Starcraft with spread out resources, on the small and medium maps, you likely still get a degree of deathballing because even if resources are spread out, they are still in relative close proximity.

The single biggest thing that helps against deathballing IMO is scale, but it isn't the silver built all on it's own either. Scale, resources, Unit design, all these things work together to solve the problem, none of them individually are 100% effective(or without thier own issues).

Mike

This guy gets it. If units become exponentially more effective when grouped together, then the obvious best strategy is to have a single force with all your units, which is impossible to defeat or even effectively damage unless your opponent has a single army with at least as many units.

Ideally PA would be designed with a general concept of "efficiency ceiling" for number of units of certain kinds engaging together as a mass. You certainly can use a single blob larger than this, and it will obviously be stronger than a smaller group, but it is less efficient than having two smaller groups, especially since they can be active in different regions, or in different positions in a theatre working together, but not necessarily all firing at the same enemies or fighting together simultaneously.

Ignoring for the moment how this would be achieved, a subject on which there appears to be contention, if we suppose it was achieved then players would use armies of that size. Players would use multiple groups at the efficient size threshold acting independently out on the map. However if for some reason such a group lacks the mojo to get a specific job done, that player would have two or more groups operate together for higher absolute strength, knowing they will be less efficient as a result. If possible they should surround a target army or base to avoid getting in each others' way and maximize their efficiency.

I do agree that my posts in this thread after the OP are not as well organized as I would like. Still, length is not something I will apologize for. Some subjects require more words to adequately cover, such as extrapolating effects on gameplay from this or that system, and why one choice is preferable over another. I do wish more people would commit more to a thread than "Wouldn't it be cool if there were SPACE BEARS!" with regards to making suggestions.


"Interesting"

What I personally find "interesting" is indeed a very weak reason for any particular feature or design in PA. While I am a very experienced RTS player, any opinion of this type is anecdotal.

However when I say that having more armies is more interesting than having a single blob, I am actually using the word "interesting" in the game theory sense of the word. Under this definition, not all decisions are interesting. For example, choosing between $10 and $100 is not an interesting decision. There is an obvious optimal decision for the actor. We might make this an interesting decision in many ways- by having a time delay for the larger amount, a risk, or some other factor that balances out the otherwise strictly superior choice. For example, would you rather have $10, or a 1 in 10 chance of $100?

So the question of whether a deathball or distributed forces is more interesting is actually an objective analysis. Not my own subjective opinion of which I prefer. A deathball reduces the player's choice to the previous case where they have to choose between $10 and $100- you can have 10 utiles by having 10 groups spread over the map, or you can have have 100 utiles by consolidating all those units in one group. This is strictly less interesting than having 10 utiles in 10 groups and 10 functionally different utiles from your other choice of having a single large group. Now the player has to make an interesting decision about whether to consolidate or spread, rather than a blanket principle "always consolidate" from exponential scaling in a deathball.

Exactly. This is a systemic issue with the canonical RTS formula of HP, damage, range, and definite properties and costs. However if you're careful with the game design you can avoid the gameplay problems.

First off, thank you, by the time I finished that post I was more unsure of it than anything.
Second off, direct caps on efficiency don't fit in with the theme of PA based on what Neutrino has said. They squick me out a bit too. But since you've mentioned that you don't know how that will be solved, then, point taken.
Thirdly, thank you oh so much for the Game theory definition of Interesting, I can't believe I've never heard it before.
Fourth, I really enjoy this topic title and subject, since this whole deathball concept and how to defeat it, is the stuff that needs to be talked about Now, during the pre-alpha stage, and it's something I hope that the developers have already been thinking about for the last few months.

I don't know how you would implement a "direct cap" on efficiency, and it sounds like something I would be against. But having units consume space, stack loosely, not shoot through one another, deal friendly fire damage, and so on and so forth, it becomes possible to make units inefficient with huge numbers.

Combat inefficiency already happens with many, if not most, unit types. Peewees are a great unit in the early game because they are a great unit full stop. However their short range and very low HP makes them less efficient to use in extremely massive groups once your economy is larger and you could theoretically afford it. A Flash moves faster and has more HP, making it more efficient to use in bigger groups, but it too has an efficiency ceiling that it hits when you have so many in a single place. It's better to either split up the Flashes, or else switch over to a more resource-dense unit type.

If you only had Peewees on a sufficiently large map, then you would be forced to have many smaller groups of them because having a single really tremendous army of Peewees would be extremely inefficient.

Hmmmm, well... to that effect, making everything more like Peewees or Flashes (in the terms of economic efficiency caps) might just be throwing the baby out with the bathwater. If you make all units paper then the deathballs that manage to come into existence would just be rarer, most likely counter-intuitive, and even more powerful in comparison. I feel like I'm using incorrect terms here, and this idea isn't as well formed, just a kind of gut reaction.

Short range and low HP are not the only ways to make a unit less efficient to use in huge numbers in the same battle. PA does not have to make all the units like the Peewee or Flash in this regard. However PA should make all units like the Peewee and Flash in that they are inefficient to use in really large tight groups, even if the method used differs from unit to unit.

Indeed, that counters single unit Deathballs and kills them pretty consistently, but focusing on it like that blinds to multi-unit deathballs, which, though brought up in the context of units countering each-other (which all in all I think is a good idea if it is a natural consequence of the systems used, but that's one of many pipe dreams we don't need to discuss), I feel could possibly arise, unintended, as a consequence of per-unit deathball countering. I've got no proof, and not even a clear train of thought as to what suggests this to me, but if it does manage to happen, those deathballs would become huge balance breakers, in crippling, game changing ways that couldn't be fixed, as balancing one deathball composition may destabilize your initial balancing or lead to a new deathball.

With proper pathing, wont there be a death line instead where army's line up and shoot each other, leading to one army to stacking on one side of a line and wrapping around the enemy's line, consuming it?

Defining that as proper pathing is a little ambiguous. But that is true, an awesome visual, and has almost nothing to do with the topic at hand. Awesome to think about though.

That would be the objective. To have forces with geometry instead of being a large blob, singularity, or point.

You can do lots of interesting things with a line of battle. You came up with one of them yourself- attacking its flanks to roll up the line. Crossing the T will give the forces on their flank a very bad day.

If that's the objective, then I agree that your idea would create it.

Mkay... I'm confused now.

Pathfinding is a small portion of combating deathballs. What igncom suggested now that I've thought about it, would be a way to use pathfinding to better combat them, but, better pathfinding as a goal of combating deathballs?

The concept of an efficiency cieling is a good one (please excuse spelling im on my phone). Obviously none of us want arbitrary limits because thats not what the game is about. so what other limitations work towards an efficiency ceiling?

Range vs unit footprint
a big one is that if a deathball has units that are out of range of things on one side of it then adding more units to the deathball is only going to increase the total hp of the ball instead of its dps or alpha strike damage. its a ceiling of sorts. Every extra unit in a group has an effectively smaller range than the previous one until that range becomes zero. One of the largest problems with things like restorer balls was that they could all occupy a very small area and immediatly apply all their dps because of their effectively non existant footprint. the inclusion of a footprint for air units solves much of that particular problem but for land units the key is the ease of reaching this ceiling. An army you churned out in 5 minutes with spare metal could be at this ceiling, it wont beat a bigger army but it will inflict actual damage as its not destroyed instantly, your opponent is also at the damage ceiling. coming with multiple armies increases the width of the battle and your damage ceiling.

Maneuverability and centralised objectives
Particularly in starcraft you can see deathballs squeezing through the tinyest of gaps. they are fast enough to have a 1 minute response time from just about anywhere on the map meaning you can afford to keep all your units in that blob. If it will take longer to get back home you cannot afford to keep your units in that death blob because by the time it can respond you are crippled or worse. Similarly if you have more than one location to defend smaller groups can just hit you were your deathball isn't, creating a need for armies accross the map. Furthermore if having a large group of units creates trouble moving them through gaps and otherwise slows down the movement speed overall, this is a further incentive to create smaller forces and increases the response time of people against these armies. the ceiling here is that you will tend to only arrive with certain parts of a blob past a certain size, because it has to split up to travel, or you can only afford to put in so many units while you defend.

Anti-blob weapons
This isn't just about splash damage, although that is a part of it. but take artillery for example, its only good against big things not just because it tends to deal large splash damage, but because it never really hits what you aim it at. Weapons that miss suck against one target, but against groups this loss of damage can become mitigated. with weapons that miss a lot either by inaccuracy or change in direction of the group during flight a large group is still easily hit, bringing the otherwise pathetic effective dps of such weapons much closer to its theoreticall maximum. Similarly weapons with a very large area of effect reach their maximum dps much more easily as it is simple to get many units under their areas instead of just the fringe of the group. the larger your group is the longer these weapons will be at full efficiency with laying waste to the resources you put into that army.

all these things can work together so that the larger a deathball gets, in many ways the worse it gets.

One thing I did like about the terrible pathing in supcom 1 was that it would cause units to actually clash together and get stuck, making a good formation impenetrable.

Technically, if a formation is tight enough that a unit can't fit inbetween your units, it's still technically impenetrable, unless you're playing SupCom2 where units can slip and slide on top of clipping to get into that tight fit.

Mike

Yeah....formations are useless in supcom2.

im not trying to push experimentals in the game (you'll see why)
but arent experimentals "DeathBall" killers
i know that there comes a point that a deathball is leaded by an experimental or a deathball
with experimentals only
but that is also the point when nukes and planetcrackers come in
wich lead to the title
PLANETAIRY ANNIHILATION
*problem solved* (wich wasnt really a problem at the first place)

Not strictly speaking, or rather, for an experimental to be Anti-Deathball it still needs to be designed to fill that role just like any unit. Just being bigger than regular units doesn't directly affect that as much as you think.

Mike