|
I don't really get the all-in point. I you see, that your contain don't do enough damage because the opponent has already enough tank/coloss, you can just retreat. There is no need to wait till you get overpowered (and if he can defend them without taking damage, he will overpower you for sure).
They may not have special synergies with other zerg units, but they aren't dead supply either. The locusts deal there high damage when used together with the rest of the army on open field.
And SWs are surprisingly good in harass. Drop them in the enemy main, let go 1-3 waves of locusts and retreat before the enemy army approaches. They'll deal great devastation due there lowhp highdps mechanic.
Swarmhosts are not tanks nor lurker. They are a truly new unit that didn't exist in scbw or WoL. And because of that there doesn't exist experience how they will perform longterm and metagame wise. I personally like them how they are and hope they won't change there mechanics (minor changes if they seem overall to weak oder strong are ofc ok).
|
Excellent thread. You've understood and successfully described the problems with the current Swarm Host. Conceptually, I do like the Swarm Host, but as it is used, it creates a bad kind of game play, just like you have described. The Swarm Host is currently a death-ball unit, more effective when massed and clumped. So, what is the best way to prevent the death-ball syndrome? Non-stacking AoE damage. Following this I will present my solution to this problem.
Possible solution: Each Swarm Host spawns one locust that has increased durability, increased size and does AoE damage in an area in front of it. I'm thinking the same kind of AoE that the ultralisk or hellbat has. This AoE damage does not stack (much like storm), meaning if you have more locusts attacking the same target, it won't do more damage than 1 locust attacking that target.
This solution would make any kind of death balling of the Swarm Host very ineffective. While just a few Swarm Hosts could hold their ground and hold chokes effectively. This would also enable you to use Swarm Hosts as defense, like holding a ramp from a clumped up force, or even defending drops.
On December 17 2012 01:32 Goj wrote: I don't really get the all-in point. I you see, that your contain don't do enough damage because the opponent has already enough tank/coloss, you can just retreat. There is no need to wait till you get overpowered (and if he can defend them without taking damage, he will overpower you for sure).
They may not have special synergies with other zerg units, but they aren't dead supply either. The locusts deal there high damage when used together with the rest of the army on open field.
And SWs are surprisingly good in harass. Drop them in the enemy main, let go 1-3 waves of locusts and retreat before the enemy army approaches. They'll deal great devastation due there lowhp highdps mechanic.
Swarmhosts are not tanks nor lurker. They are a truly new unit that didn't exist in scbw or WoL. And because of that there doesn't exist experience how they will perform longterm and metagame wise. I personally like them how they are and hope they won't change there mechanics (minor changes if they seem overall to weak oder strong are ofc ok).
The problem is, that going for a low count of swarm hosts ensures that they won't do anything at all. This is because once the locusts are cleaned up (which they are going to be), the opposing army will just be able to attack the swarm hosts directly. When this happens, there can be two outcomes: 1. You leave your swarm hosts where they are - they get killed. Getting more swarm hosts at this point would just lose you the game, since the opposing army could easily beat your original ones. 2. You unborrow your swarm hosts and retreat into a defensive posture. You have to get more swarm hosts in order to make them more effective, OR you have to transition into some other kind of unit. Doing the latter makes your first swarm hosts completely unnecessary, and next game you just don't build them in this fasion.
|
I don't agree with the OP at all, especially this part:
"The Zerg player knows that 5-6 swarm hosts won’t create a critical mass of locusts needed to break the Protoss players defensive position with Storm or Colossus on the way, and in response he/she proceeds to make 15 swarm hosts."
The Locusts of 6 Swarm Hosts will make it past and get in range of 6 Colossus with Thermal Lance. That's pretty good i'd say. If the Colossus don't move, they will take damage. If any units are in front of the Colossus then those units will take damage. You're not going to kill any units with the first wave, but as long as you do damage it's worth it, especially against Protoss. You can micro the Colossi back to prevent taking damage, but that still means you're putting pressure on the Protoss.
Also, using Storms to defend against Locusts is a bad idea. I'm sure every Zerg will be happy to have free units die to storm, that's one less Storm that can be cast on units that aren't free.
And yes, Swarm Hosts will be overrun if they are by themselves. But that is no different from Siege Tanks or Broodlords. They're not supposed to kill the opponent by themselves. They're supposed to put pressure on the opponent, possibly do some damage and force a reaction.
|
United Arab Emirates439 Posts
On December 17 2012 02:07 Elp wrote: I don't agree with the OP at all, especially this part:
"The Zerg player knows that 5-6 swarm hosts won’t create a critical mass of locusts needed to break the Protoss players defensive position with Storm or Colossus on the way, and in response he/she proceeds to make 15 swarm hosts."
The Locusts of 6 Swarm Hosts will make it past and get in range of 6 Colossus with Thermal Lance. That's pretty good i'd say. If the Colossus don't move, they will take damage. If any units are in front of the Colossus then those units will take damage. You're not going to kill any units with the first wave, but as long as you do damage it's worth it, especially against Protoss. You can micro the Colossi back to prevent taking damage, but that still means you're putting pressure on the Protoss.
Also, using Storms to defend against Locusts is a bad idea. I'm sure every Zerg will be happy to have free units die to storm, that's one less Storm that can be cast on units that aren't free.
And yes, Swarm Hosts will be overrun if they are by themselves. But that is no different from Siege Tanks or Broodlords. They're not supposed to kill the opponent by themselves. They're supposed to put pressure on the opponent, possibly do some damage and force a reaction.
Yes, not to mention if you have map control against the Protoss, 6 Swarm Hosts can attack in multiple locations, meaning he can't just take his 6 Colossus, plant them right where you are, and defend. He has to split his 6 Colossus up into how many possible locations you can assault from, just to defend from your 6 Swarm Hosts. Or he has to gain air dominance, use Hallucinations, or perhaps Observers, to spot where your Swarm Hosts move to.
Everyone in this thread who is saying all you can do with Swarm Hosts is plant them down in one location and try and overpower your opponent are using the unit so wrong. That's only good in a "I've won this game, stop turtling and gtfo" situation.
|
Swarm hosts are not a mid game power unit.
They are strong in mass.
you you litterally have to "swarm" the swarm host to make something work.
The thread i made Talked about a new upgrade that would not only make it a strong powerful unit, but would also solve the massing problem.
Now the locust life(which would effect both types of locust) upgrade would have to have a slighty cost decrease so that people would want to get both upgrades and invest in swarm hosts.
The problem with the swarm host is that it only feels swarmy when you mass it, so basically its underwelming, because its needed in mass, so design wise its a burrowed BL.
Now my upgrade would make it unique,cool, and make fully upgraded swarm hosts a presense felt on the battlefield, even in a small group.
|
The OP does a very good job at identifying the issues with the SH. Any siege unit should force a reaction, even in small numbers. If 2 tanks show up in your natural, that forces some kind of reaction, they provoke the opponent to engage. 2 SH can basically be ignored and this is the main reason why they fail as zerg's siege unit.
Any suggestion to make SH work as intended, should follow this premise: Once as little as 2-3 SH are burrowed in front of a base, they should do real damage, forcing the opponent to deal with them.
The OP suggestion is clearly a step in that direction and definitely deserves testing. I also think that anything that helps locusts survive until they engage is a good measure. So on possible suggestion would be something like giving locusts an undetectable burrowed movement and they would unburrow at the rally location or once they find a structure. Bonus damage vs structures could also be interesting.
|
very good write up. i think increasing the locust size a lot and maybe increasing its hp might help make them more viable in smaller numbers since they wont be killed by AoE so fast while if you want to mass them they get worse since the now bigger locusts block each other so there would be no point in massing them. if they are too strong after hp and size increase just nerf their damage so they actually reach their target better and tank for your army but do less damage by themselves.
|
In my experiment with SHs, you will have to make at least 8 and continue to pump roach/hydra/ling from behind to make the contain effective.
But I think despite the all-in part, the thing that I cannot take is that, for that much of a investment (1100 gas to make 8 SHs with upgrades), you will not have any anti-air to plan this attack, once you go into your opponents base and try to set up a contain, you saw "oh shit, voids/banshees/tempests" then you lose everything. It is just not worth the cost in my opinion.
I am thinking of the following solution:
1. nerf SHs slightly, but drop the cost (resources/supply), so that we can incorporate hydras in order to set up a contain.
2. keep the current stat, and maybe nerf it a bit, but give locusts a anti-air attack. That will make sure the opponent to choose the engagement wisely/scout better and making SH pushes not so much of a toss-up.
|
United States7483 Posts
I've played some games vs. a roach/hydra/viper composition where the zerg player mixed in a few swarm hosts late game to force me to engage him at a bad angle or do nothing with a decent chunk of my army. I just had to sit there at my fourth base killing locusts for a while, attempting to drop but failing because of his absurd spine/spore count. Meanwhile, he was abducting whatever units he could to pick them off, and I had no chance of securing a fifth. I couldn't push out due to his positioning, and I couldn't circumvent him due to the swarm host pressure. I wound up splitting my army and then he just pushed in and killed me.
|
Russian Federation4295 Posts
2. keep the current stat, and maybe nerf it a bit, but give locusts a anti-air attack. That will make sure the opponent to choose the engagement wisely/scout better and making SH pushes not so much of a toss-up.
Never! ZvZ will become TvT
|
I would say it promotes non-deathball play. If you clump all your swarmhosts and rally them towards the opponent tank/collossus-line, then they wont do shit (or win the game if you are way ahead). If you park a few near a lightly guarded base with cannons/a couple of bunkers. You will deal damage and force a repositioning of their army. Then you can pull back.
These are the scenarios I can see: either you split your army badly and dont keep track of his units: he can kill your army in two bites. Or you split your army well-putting pressure on one of his expos while forcing an engage through locusts vs your main army. Or he splits his army well, forcing your harass away while killing of the locusts spawning from your main position.
If I would advocate a change, It would be to make them faster so you can retreat more easily (but I dont know how well they retreat on creep for ex.).
I think it's a great idea for a unit, promotes completely different strategies due to the unique mechanics.
|
Russian Federation4295 Posts
(but I dont know how well they retreat on creep for ex.). Locust speed on creep is 2.65, which is 40% faster than offcreep at 1.88 speed.
|
On December 17 2012 02:34 ZjiublingZ wrote:Show nested quote +On December 17 2012 02:07 Elp wrote: I don't agree with the OP at all, especially this part:
"The Zerg player knows that 5-6 swarm hosts won’t create a critical mass of locusts needed to break the Protoss players defensive position with Storm or Colossus on the way, and in response he/she proceeds to make 15 swarm hosts."
The Locusts of 6 Swarm Hosts will make it past and get in range of 6 Colossus with Thermal Lance. That's pretty good i'd say. If the Colossus don't move, they will take damage. If any units are in front of the Colossus then those units will take damage. You're not going to kill any units with the first wave, but as long as you do damage it's worth it, especially against Protoss. You can micro the Colossi back to prevent taking damage, but that still means you're putting pressure on the Protoss.
Also, using Storms to defend against Locusts is a bad idea. I'm sure every Zerg will be happy to have free units die to storm, that's one less Storm that can be cast on units that aren't free.
And yes, Swarm Hosts will be overrun if they are by themselves. But that is no different from Siege Tanks or Broodlords. They're not supposed to kill the opponent by themselves. They're supposed to put pressure on the opponent, possibly do some damage and force a reaction. Yes, not to mention if you have map control against the Protoss, 6 Swarm Hosts can attack in multiple locations, meaning he can't just take his 6 Colossus, plant them right where you are, and defend. He has to split his 6 Colossus up into how many possible locations you can assault from, just to defend from your 6 Swarm Hosts. Or he has to gain air dominance, use Hallucinations, or perhaps Observers, to spot where your Swarm Hosts move to. Everyone in this thread who is saying all you can do with Swarm Hosts is plant them down in one location and try and overpower your opponent are using the unit so wrong. That's only good in a "I've won this game, stop turtling and gtfo" situation.
I think you and that other guy are severely overestimating the strength of 6 SH, their pushing power will likely kill nothing, a zealot or stalker here and there at most and will offer almost no map control. The thing is, having SH in a way reduces a zergs overall map control in favor of controling one path, if your opponent find a way around or does good drop play you have not much tech which can do anything about it. Mutas give waaaay better map control since you can actually respond to drops and kill them and use mobility to fight anywhere on the map. If sc2 is like chess then the SH is a Rook.
|
I to feel that swarm hosts don't really function the way that they should. doing some thinking on the subject, and trying to figure out how I would like to use/engage with a mid game army, i came up with a different solution to the "weak in large numbers, strong in small numbers" conclusion that the OP suggested.
in my opinion making the locusts beefier and less damage dealing will only serve to make the unit different, maybe it contributes more to the core army as a way to soak up damage, but it lacks in being able to be useful in creative and interesting gameplay. going back to the mutalisk example, what makes mutas so effective? its how fast and efficient they are. they can easily get in and get out and deal some damage/take map control. however when mutas are massed what happens? in zvt the thor will shut down mass muta play with its anti-light/splash anit-air, the siege tank do this to hydras as well. however in both of these examples better micro can make them more effective (magic boxing and spreading+speed)
what i suggest for the swarm host is to make the locus more like a glass cannon, high damage (or maybe just +vs light) and low health with a moderately fast move speed, (keep the timed life though), as well as lower the cost of the unit, as it stands they are too expensive for what little they do. what does this do?
1) it allows for the unit to have harassment potential b/c it would be good vs. workers/structures, this might make swarm host drops much more effective, and increase the incentive to make a timing push with them on thirds. Even into the late game the unit could be left near an expansion (yours or theirs), while the main army pushes for example.
2) with a high number of locus on the field the opponent can easily tech to a good aoe dealing unit to make short work of your swarm hosts, as colossi and tanks will both be good at dealing with them. however, what if the zerg player spread his swarm hosts before burrowing? that would make them way less vulnerable to concentrated splash. in fact in the beta i've seen a few pushes like these in zvz, where the other zerg player has 3-4 swarm hosts in their army, and i didn't go for swarm hosts. Because i don't have access to splash it forces me to retreat and try to out maneuver their army until i have a way to deal with the swarm hosts. this kind of dynamic would be great if we could bring it to the other matchups.
3) with a lower cost/build time and a very different attack potential I think timing pushes could develop around the swarm host, particularly timings where 2 base zerg could pressure a terran/toss trying to take a third or a even another zerg trying to take a 3rd or 4th. in addition, it could also make lair rushs more common vs ffe toss, if the zerg could have swarm hosts up before the infrastructure of the toss kicks in, then that kind of early play could become more common, as it stands now, zerg has very very few ways to be effective at tier 1.5-2 vs the other races.
these are just my thoughts
|
Here is a replay of a diamond zvp, demonstrating my use of swarm hosts. As depicted in the OP, they delay my hive, and take out of my other ground army. However, i was able to make use of small amounts as a small anchor for my army, as tanks are for marines. Yes, it was a little effective, but it was a mssive investment without little payback, had i not done my dropping strat.
I'm tired, and ill just leave the replay here without further detail: And yes, this not a pro game, but it still makes a point. It also shows some ultra use. http://drop.sc/284991
Anyway, I think a proper solution is to buff locusts' health, damage, and nerf their range. This will cause small numbers of them to be strong, and large numbers will lose effectiveness because due to reduced range, locusts cannot all be attacking at the same time. Just like ultras in a choke.
|
Excellent post. When I opened 'possible solutions' I expected an expansive post though, so I guess I was wrong there.
Anyway, in my opinion, I think Swarm Hosts would be a lot more interesting if they actually did some close to guaranteed damage, BUT that their damage would be decreased vastly.
There are different ways to implement such a mechanic. The first one, and probably extremely hard to balance way to achieve that would be to have invulnerable locusts. Yeah, yeah imbalanced. But what if their damage would be something like 3 instead of the 14(?) they do now AND their life time would be decreased so you can't run them as far, making it so you actually have to place them in pretty risky situations.
What would the consequence be? First, since their damage is so low, they wouldn't be able to be the backbone of your army. If you mass them, your opponent moves out and kills you, simple as that. On the other side however, they will do guaranteed damage, making it so your opponent can't stay in his base forever, or eventually the Zerg's army WILL be bigger as is, because even though he isn't taking terrible terrible damage, he's still taking some damage, and can't keep it up forever.
Another similar solution would be to decrease the dps of the locusts, but make the locusts themselves a lot quicker (and possibly give them eitjer extra range OR hp), so you're again guaranteed to do damage. Since again, the dps would be decreased bu quite a bit, they wouldn't be as powerful, but it would give the swarmhosts more of a supporting role making it so that when you have a few, they actually do damage, but if you build too many, your main army will have lost quite a bit of dps.
Those are my thoughts on the unit and how I think it would be more interesting to use.
|
Not saying nothing is wrong with the Swarm Host, but... Don't see how they get stronger in numbers, besides being easier to control at that point, since it's most of your army. Don't see why it must never be all in to make Swarm Host contain, if you Tank contain and everything dies, counter attack will be strong. You can set up defense at home, but so you can with Swarm Host. The instance you say, I must get down the third, or I'm screwed, you're allin and so the big investment doesn't have to pay off always.
|
How do we make the swarm host good in small numbers, but not IMBA in larger numbers?
I can think of only 2 solutions to this problem.
1) Make swarmhosts melee, think slow chunky zerglings with alot of health. eg. zerglings are not imbalanced in large number unless they are being throws from the skys, *hint hint blizz*.
2) Anyone remember this thing from red alert 2 http://cnc.wikia.com/wiki/Terror_drone_(Red_Alert_2). The terror drone latched on to enemy units killing, or severely damaging then. Due to the design, one single drone could attack an enemy unit at a time, blizzard could adjust the "latchon range" to balance this attack. Also could be interesting if they could leap over FF, but this could possibly be imbalanced.
|
Hi everyone,
I have expanded on the Solutions section. I offer my thoughts on the subject, as well as those of Engineer and Existor. I also included an explanation for why I have not pushed for a re-introduction of the Lurker in place of the Swarm Host, as that has been a common question.
If anybody feels I have left out some good solutions that were brought up here or in another thread, please let me know.
|
These are my suggestions:
1) Increase the effectiveness of Swarm Hosts in small numbers. An easy way to do that is to have them release bursts of Locusts instead of a steady stream. The cooldown would be increased to compensate.
2) Decrease the effectiveness of Swarm Hosts in large numbers. An easy way to do this is to increase the size of the Locusts so that they become much less effective in choke points.
3) Increase the harass ability of Swarm Hosts by making them more mobile. This makes them more effective at harassment and keeps the opponent guessing if he doesn't have vision.
4) Make Locusts die on impact but do a lot of damage. No splash.
Let's examine what these effects will do to the Swarm Host:
Controlled bursts but long cooldown with mobility allow the unit to harass at range. But sustained damage isn't as good so you won't want to build too many of them. It forces your opponent to split his forces to deal with the harass. You, as the Zerg player, get mobile harass that is relatively safe but it eats up supply from your main army. But if your opponent doesn't use enough defenses, he'll take substantial damage. If you try to build too many Swarm hosts, they bunch up at choke points and become far less effective. Not to mention a single forcefield stops them from going up a ramp.
Unlike muta harass which can target workers down, the Hosts should do substantial building/unit damage instead at a relatively safe range. It gives Zerg another option for harass. The extra speed allows them to join an army but the increased size of the locusts and lack of sustain DPS means more than a few swarm hosts is a bad idea.
It's just an idea based on theory (since I can't test said changes) but I believe this might be a good solution.
|
|
|
|