I am not convinced starbow is a deathball game.
but for the sake of argument, lets say it is.
to make a deathball game a non-deathball game you need to encourage small engagements over large ones.
let us pretend that we have X marines fighting 1 marine with infinity health, the infinity health one will obviously win, buthow much damage will the mortal marines deal before dying?
well, X = 1 means both sides will deal equal damage, so (assuming a marine have Z health) the infinity health marine will take Z damage.
X = 2 means the two will deal double damage until one marine dies, then an additional Z damage will be dealt, 2Z + Z = 3Z
X = 3 means the three will deal triple damage until one marine dies, than an additional 3Z damage will be dealt, 3Z + 3Z = 6Z
X = 4, 4Z+6Z=10Z
5, 5+10=15
6, 21
7, 28
8, 36
9, 45
10, 55
100, 5050
as you can see, simply by having units which can be massproduced the game automaticly becomes centered around having as much stuff as possible, since having 2 of a unit is not twice as good, but three times as good, thus deathball. this is entirely unavoidable.
however, starcraft combat is balanced around more than just units, the key to making a deathball game a non-deathball game is two-fold:
casters and defenders advantage.
these factors allow a smaller force to defeat a larger force, or at the very least even out the odds. it is very much tied to what I said about spells of supreme power:
if all casters had only spells of supreme power, why would anyone build ordinary units? they make little difference in the battle.
that situation is the exact opposite of a deathball game, its a micro game (much like MOBA games which, as you may note, are not deathball games).
however, as with all things extremes are generally the worst possible options.
the current balance between casters and units feels fine, casters do enough to even the odds of a small army against a big army, but not so much as to nullify the army advantage of the enemy. obviously this depends on how you use them.
I started with saying "you need to encourage small engagements over large ones" and I kind of sidetracked from there, but I began with what is most important.
small, strategically correct and tactically executed engagements must be more efficient than storming the first weak point you see with everything you have.
that is the basis of what I think you try to achieve. (do correct me if I'm wrong)
what is interesting is that this have less to do with the game than you think. it has to do with the map.
months ago I said this:
PM from Roblin to kabel:
...
there are 3 main features which I believe would enhance the experience, namely:
smaller and "deeper" chokes than normal: if chokes are tiny and long, such as corridors, then it is very difficult to move big armies through them, thus telling the players "use smaller armies on more places", smaller chokes also allow small amounts of units to hold against big armies for extended amounts of time, thus forcing the attacker to rethink his attackpath, I mean chokes so small that no more than 2 dragoons can pass it at a time, 3 lurkers will easily die to 10 dragoons, but 10 dragoons will suffer heavy casualties if they try to pass through a choke guarded by lurkers.
open areas become even more open: this in combination with smaller chokes makes certain strategies heavily favor certain types of terrain, tanks are best put where they guard chokes, zerglings, marines and zealots want to wait until the enemy leaves the choke before surrounding and destroying, the chokes is a "nono zone" for zerglings while those same chokes is the favourite position of a reaver or lurker.
more bases: if every base lasts longer until it depletes, that means starvation tactics will be less effective and thus less viable, this would be counteracted by "rewarding" the player with more mapcontrol with more possible bases to take.
...
and that is what I based the map "forbidden lagoon" upon
it still holds true.
note:
this sentence: "more bases: if every base lasts longer until it depletes," references the economy system which was in use at the time, it has no connection with the current economy system. rewarding mapcontrol with more bases is still important to note though.
do you not agree that the games we played on forbidden lagoon was incredibly epic? games felt like they lasted forever, the average game lasted 40+ minutes, bases were taken left, right and center, small groups of units was constantly moving somewhere on the map, but it was always the player that kept managing the economy that ended up winning.
the map was the first I made, it looked crappy, Im sure it had balance issues, and was bad in almost every respect.
but it was good in others.
the chokes were small enough to let 10 supply of units fight against 50, there were space for epic battles (remember when I dropped 10 lurkers, 30+ lings and a bunch of hydras into your defending siegeline? you won the battle with 2 tanks and a couple of marines left, and this was before darkswarm was in the game!)
and thanks to those factors, there were not a single deathball in any of those games.
the reason we saw no deathballs is because the map was tailored to give massive defenders advantage. it was never worth it to send your entire army at the enemy because both players had small groups of troops raiding everywhere, and the enemy could force battles to drag out for several minutes if need be. leaving something undefended would be suicide.
in conclusion: its not the mod, its the maps.
if maps give lots of defenders advantage and have lots of bases (7+ per player), that discourages deathballs very effectively.