[D] What changes could help with death balls?

On October 05 2012 15:07 Alex1Sun wrote:
A simple solution to break death balls is to give all races huge AOE (the whole screen), high range, no friendly fire units with stackable low damage that would be cost- and supply-inefficient against harass or small task forces, but would force the opponent's army to be spread over more than one screen or split into several far-away groups for multi-pronged action.

Additional damage vs massive would also likely be necessary to break high hp death balls consisting of colossi or thros. Lowered damage to workers might also be required.

On October 05 2012 16:27 Stow.Wif wrote:
Good AOE for each race, no need to make them too huge. If you look a it, there is not much deahtball effect in the matchups where the 2 races have efficient AOE againt each other :

TvT : tanks are good, not much deathball (lot of small moves around the map).
TvZ : tanks/hellions are good, fungal and banes are good againt marines, not much death ball except againt mech (banes and fungal not as effective) and in the late game (broods forcing tanks to unsiege).
TvP : tanks and hellions are bad, so toss is encouraged to make a deathball, to which terran responds with a bio deathball. Note that emp cannot stop a group of unit by itself.
PvZ : banes/fungal are not good enough dps -> same effect as above
PvP : this one is tricky, but i believe AOE are not efficient enough in toss vs toss, there is no possibility to defend a position with 2 templar or 2 colossus.
ZvZ : when it comes to roach, banes/fungals are not good enough : deathballs of roaches

Of course, AOE is not the only factor to the deathball effect. Also come in mind the possibility of using efficiently a small squad of unit to achieve a goal, and this point is tied to races but also to maps.

Some idea : buff storm against toss (adding an extra effect to shields ?), buff tanks (maybe the hellion buff is enough) with a drawback, make bane drop more efficient againt deathballs.

On October 05 2012 17:28 ledarsi wrote:
In order to remove the deathball, it is necessary to add tools to each race that make non-deathball play drastically more effective than using a deathball. This means positional units that are very powerful. It means non-stacking powerful spells like dark swarm. It means board control like zero-supply mines. It means adding tools to each race that make deathballs suboptimal.

These things give the player options, and allow them to split their army up and INCREASE its effectiveness, rather than decrease it. If your army is strictly less effective in smaller groups, then a bigger ball will kill the small groups one at a time with few casualties in each battle. What needs to happen is a small group needs to be more efficient than a larger group. A larger army might beat a smaller group, but will suffer enough casualties that multiple small groups will defeat it.

On October 05 2012 22:35 SC2John wrote:
I have been going on and on about space control in many topics over the past week. I'm 100% positive that the main issue with SC2 is the fact that space control takes a lot of units; there is no way to defend an area against a maxed army unless you yourself have a maxed army.

Here is one of my old posts:

On October 05 2012 17:28 ledarsi wrote:
More units (by reducing supply costs of units) will greatly encourage splitting forces up. Colossi are a 6 supply splash damage dealer. You cannot, and should not, have very many of them. Compare this to a 2 supply siege tank in BW. You can have way, way more of them. The simple fact that you can have more stuff encourages using smaller groups, rather than a single group that is as large as you can make it (right up to the supply limit). Even in SC2, really huge armies have diminishing returns on each additional unit added. It's just that you don't have enough units where those diminishing returns start to kick in. Modifying the units so these diminishing returns kick in sooner would also be good. Two forces of half the size resulting in a net increase in fighting effectiveness, rather than asking to lose, is mandatory.

On October 05 2012 18:33 Zaurus wrote:
Why attack separately when you can attack together? To solve deathball, limit control group to 12..... Easy solution. I m just worried for P, Protoss units don't work well in small numbers.

On October 06 2012 04:44 XXXSmOke wrote:
The best way to eliminate a death ball is reset the AI to only selecting 12 units per hotkey.

You can change the units all you want, but letting them all ball up and just go 1a is going to keep death balls active.

On October 05 2012 19:53 Roth wrote:
I think Blizzard could also do something about the maps.

There was once a thread about smaller bases with just 6 mineral patches and 1 gas. I think this could also help to limit deathballplay because you do not have that many ressources to build up a big deathball. So to get a big army you first have to spread yourself out and take many bases which you also have to defend.
The thing with this change is that not every map needs to have smaller bases. You could make a map pool out of ~3 bases with small expansions and ~ 3 with big expansions. So there would be a lot of different play, especially if you would mix them up in a BoX series.

Another point I want to add is to make the maps bigger and the distance betweeen the expansions bigger so you have to stretch your army out to defend multiple locations. In this scenario you could not have a big army together that effective because you would neglect your expansions.


Just some thoughts I had. I think the positive thing about these changes is that no major changes have to be done. It would all be about the creation of maps.

On October 07 2012 03:12 kcdc wrote:
You'd have to rebalance the whole game. What you need to do is make attacking with small task forces more rewarding. That means forcing players to spread their defenses more thinly over more bases and having those bases more exposed so they're easier to attack. Currently, all the modern maps are designed so that you can hold 3 bases with forcefields because that's a requirement for game balance in WoL. That layout makes it easy to defend 3 bases with one army without worrying too much about splitting forces or attacks from multiple angles. You could open the maps up more, but Protoss would be underpowered.

On October 05 2012 23:12 NukeD wrote:

Dynamic unit movement.

On October 05 2012 22:48 puissance wrote:


I second this, also I dont know if it has been brought up, but highground advantage is another point which allows few units to hold key positions. E.g. 1-2 Tanks + Mines behind Supply Depots or some Spines, Lurker and a Defiler (or even just the Nydus and 2 Lurker).

On October 05 2012 23:26 AzraelArchontas wrote:
A mild high-ground advantage to make defending slightly easier
say a -1/2 range to low-ground armies
If you want to make this more noticeable
Add a +1/2 range to high-ground armies as well

On October 05 2012 23:26 SarcasmMonster wrote:
Overkill.

a) With no overkill, every unit deals damage at 100% efficiency. Even in a large deathball, no overkill means every unit is 100% efficient.
b) With overkill, damage efficiency drops steadily as the size of the deathball grows.
c) Hence overkill slows down the growth of power of a deathball.

ALTERNATIVE EXPLANATION

a) Overkill is a nerf, hence units with overkill can be stronger than their hypothetical counterpart without overkill (to compensate for the overkill nerf).
b) Overkill does not affect small groups of units as much as large groups of units.
c) Hence balancing around overkill is a buff to small squad attacks and a nerf to deathballs.

On October 05 2012 23:52 Perscienter wrote:
On topic: additionally to many other suggestions, an improved targeting AI might help. I can't access replays and videos right now, but had the impression, that colossi are often targeting the nearest, single unit. I assume, that the AI can be programmed to maximize the damage against groups of units, thus aiming at the damage-maximizing centre of a lump.

On October 06 2012 04:31 Doko wrote:
In the specific case of tanks a different approach over say buffing damage / splash / fire rate etc to improve territory control / force deathball splits is the following.

A B C
X Y Z


-Tank-

If xyz are marines / lings zealots whatever 1 thing is for certain. If target priority is equal for ABCXYZ... the front row (xyz) will always get hit before the back row (abc). For simplicity lets say Y gets hit first. The tank shot will deal full damage to Y, most of the time X B Z (the ones 90 degrees off the point of impact) will take around 50%. A and C (the ones 45 degrees off the point of impact) will take around 25%.

The problem with this is the fact that a huge portion of the siege tank aoe damage is not being utilized AT ALL. units in the front always get hit unless target fired by you, due to having other more important things at the time target firing is not always possible.

What if instead of just buffing damage, fire rate or siege time of tanks you give them a piercing shot that deals full damage in a line to Y and B, 50% to X Z (90 degrees), and 25% to A and B.

You now deal a stupid amount of damage to units trying to advance in narrow terrain or if the opponent doesn't bother to pre-split some units in front to take the first shot while he advances, at the same time vs an opponent that creates a huge concave to attack into you utilizing the fact that you are on open terrain only receives a small penalty.

Obviously this might force you to nerf the damage if it became to strong or promoted extreme turtle games but changing the numbers is not the only way to look at it.

If for example this proved to be too strong in the specific case of a tank they could try applying it to void rays, tempests or whatever unit requires a boost. (those 2 just came to mind cause the effect of an energy beam / gigantic ball of energy piercing stuff would be "cool" and I personally consider voidrays one of the coolest units in the game that failed miserably at being useful outside of cheese).

On October 06 2012 08:37 ledarsi wrote:
All the people suggesting limiting selection size or control group size are being silly. That will increase the mechanical difficulty of utilizing a deathball, but as long as the keeping all your units together in one place is ideal, that is what players will do. Even if we make it more difficult to execute. The only way to solve the deathball problem is to make deathballs suboptimal, such that you gain an advantage by doing something else.

Regarding the diminishing returns point from earlier, Alex1Sun is absolutely right that units with longer ranges are more effective in larger groups. Short range units, and especially melee units, suffer more serious diminishing returns as the localized unit count rises. Note that in Brood War, there were relatively few long range units, and those units had very serious drawbacks. The Siege Tank is immobile when sieged in return for its range and firepower. The Reaver can hardly move at all without a shuttle, and is best used for harassment, not en masse as a group. And the Guardian is relatively weak in a main combat situation. These units don't make for strong deathballs, despite their range. By contrast, units like Colossi and Broodlords are excellent in a main combat situation, and their low unit count (small surface area), high damage, and other features, such as mobility, cliff walk, spawning broodlings, etc. make them powerful and survivable in numbers on a main combat footing, unlike Reavers.

And to address the number of units issue in greater detail, there is a serious problem of fungibility of forces if your army is composed of big units. Higher supply costs means you have fewer options for how you determine your composition, and fewer choices about their distribution over an area. Suppose a zerg is building mutalisks, and the terran goes for thors to respond. Each Thor is a large investment, with a long wait before it is finished. And each Thor functions at 100% capability until its HP reaches zero, at which point its strength is zero. Compare this with a 2 supply goliath. You can build three of them for the same supply cost as one thor. This gives you options for how to distribute your supply that the thor does not. You might choose to build two goliaths and an additional tank (2 supply in BW) rather than go for three goliaths.

Sidenote on purchase economies + Show Spoiler +

Furthermore, once you have three units, such as three goliaths instead of the one thor on the board, you can distribute them. You can keep the three goliaths together, or split them up to different areas of the board, such as different areas of your base to deflect mutalisk harassment. There is no way to split up a thor- it is impossible. Even worse, if you lose the one thor, you lose a larger chunk of strength than if you lose one goliath. And it is not possible to split up a thor. So, because each thor represents a larger investment and chunk of your supply which is not negated unless it is totally destroyed, you are incentivized to keep all your thors together so they mutually protect each other. Cheaper, weaker goliaths benefit from this same process too, of course. But when they fight they will suffer more casualties, with a more continuous effectiveness dropoff. Rather than a thor going from full strength to zero instantly, you lose one goliath, then the second, and then the third. Apply this to a large army and this continuous-strength-loss-with-damage factor makes a tremendous difference in battle.

Smaller, cheaper, weaker units (such as marines) when used in groups will sustain casualties in battle, even if they win decisively. This weakens the army, and the units must then be replaced to bring it back up to strength. Larger units means wider variance in casualty figures for any particular battle. A force of thors that has sustained serious damage, but no actual casualties is still basically 100% effective. This becomes a serious issue when these larger units are like Colossi, which have a relatively noncontinuous strength dropoff, more like Thors, and which counter units which do have a more continuous strength dropoff with damage. Imagine a micro-Colossus which costs 2 supply, with suitably less power such that three of them is as strong as an SC2 Colossus. This micro-Colossus would be a vastly more interesting unit just due to the greater numbers, greater fungibility of production, army distribution over space, and casualties sustained during battle.

On October 06 2012 04:42 Cloak wrote:
DPS density is the core issue, or written another way, DPS/Surface Area. You either tackle ways of lowering the DPS, lower range, over kill, AoE that kills the DPS indirectly, or you lower the operable surface area, pathing, unit range, or physical space occupancy. Any other changes wouldn't have an effect really unless they lower that ratio down for all 3 races. SC2 is just a lot cleaner and smoother, so you get these unnaturally fluid army dynamics. We've hit the uncanny valley of army simulations, so now we need artificial blemishes to make it more interesting and aesthetic.

On October 07 2012 20:54 FeyFey wrote:
I don't think that buffing AoEs or only being able to select one group really would stop deathballs or that space control units would. Not that I mind deathballs after TvT, PvZ is my favorite BW matchup and that revolves around the toss making 1 huge army and then rolling over the Zerg, while the Zerg has no such means and needs other ways to stop it. So one side being able to deathball is awesome.
Right now especially the maps are at fault that deathballs are yay. If there is only one position you have to attack, then you don't need to split armies. And well 3 bases + production, can there be a better target. In BW there also was a ton of deathball play. But you could play against it and since one race had the better deathball most of the time, you had to work on beating the deathball in another way. There was a problem for the one going for one giant army in bw though. A deathball army was first of all slow and second of all you had to kill one base after the other, which gave the opponent enough time to attack at different positions and slowly kill off the deathball one by one.
In Sc2 if your deathball arrived at one base you are at the doorstep to the production and all the other bases. And especially the toss deathball is freaking fast.

What really prevents deathballs is multiple positions far away from each other being important and the means to defend them, which means passive defenders advantage. So choke points vision advantages and all that stuff. Right now only the main and natural base of a starting location has those advantages, every other base on the map has just the defenders advantages that you build there.
In BW as Zerg you didn't wanted to spawn cross on some maps against a Toss, if the Toss spawned North, you wanted to end up East because it meant you could expand on the West side of the map. If the Protoss attacked one location it didn't mattered because you had another equally good location and they exposed their tech and production to your army if they attacked.

If you don't want deathballs, there have to be multiple location on the map both players have to fight for, otherwise deathballs will happen, but thats what free defenders advantage is for. gives you the option to slow down the Deathball with less units. BW is a good example here with the highground mechanic, allowing a zerg to slow down a terran midgame push for quiet some time each highground that lies between the bases.
If you hate deathballs with passion though and I have no idea why, look at Dawn of War2 for example, they eliminated deathballs, because you need to be scattered over the whole map or you lose.
There are enough examples out there that show how to do it. But there is also another thing that has to be kept in mind, people want to play deathball style, because everything else is harder to play.

And well introducing new maps is a huge problem atm, since its easier if every map plays the same and tournaments testing new maps also gets problematic. Especially since maps often favor a matchup or the other so you need a veto system. New maps normally are a cheese-fest if forced or never get picked.

And even an instant nuke won't prevent deathballs, you will rather see those units protecting the deathball. I mean even Vortex doesn't stop Zerg from going full deathball, it just protects the toss deathball.