|
Free units are a load of it
So what you often hear is that free units are bad for a thousand reasons but mainly because they allow you to kill things without taking damage. That's not really what free units do, that's what 22 effective range does. If a unit spawns free units that only travel 7 range before it dies it will be taking damage. Also, the units aren't actually free, everything in this game costs a resource, in this case you pay a cooldown for the unit. A cooldown is a resource that is regenerated over the cause of its duration and can't be stacked in how you store it. It's effectively similar to energy except that it can't be stored further than the amount it required. If templar only had storm and couldn't store more than 75 energy. Then storm was effectively on a 135 second cooldown barring things like emp. You spend your cooldown every time a set of locusts pop out. Just as a roach spends its cooldown every time it fires acidic saliva. And in a lot of cases you want to be mindful where to spend it on, roaches overkill gigantically so that cooldown is best spent well. If you spend it on an immortal that just jumped into a prism then you wasted the resource, in that sense, attacks of any unit aren't 'free', they cost a cooldown. But as said, it's a resource that can't be stored so if you're not using it you're wasting it. And this is exactly what getting a concave does, it ensures you do not let this resource go to waste by getting as many of your units to fire at the same time.
So a swarmhost spawns "free units" to fire at stuff from strategic range, just like a tempest spawns "free bullets" from strategic range. But the major catch is of course that bullets don't tank damage except from PDD's and locusts do tank damage. But ehh, how exactly is it a bad thing that people are forced to consider what they want to focus down rather than just amove? Please tell me, how is it bad for the game to force people to focus fire or at the very least make the decision when it is better to attack the locusts and temporarily decrease damage or kill the swarmhost standing in the locusts and permanently decrease it? In ZvZ there's often a decision of 'Do I want to waste banelings on locusts, units that are going to die anyway?', well, as it stands this is in many cases indeed the correct decision to waste banelings on locusts but it's still a decision to make. I don't actually see why swarmhosts effectively firing slow moving bullets that tank damage is a bad thing on paper. It creates decisions and forces micro. You can't attack normal bullets, it's one less decision in that sense.
Swarmhosts have a sickeningly long cooldown
This is very good for a setup unit. The point with siege tanks is that their cooldown is fairly short for a setup unit, which means that you waste cooldowns relocating them. Thereby it becomes optimal to keep them where they are, moving them around wastes fire. The thing with swarmhosts is that swarmhost-kiting is an effective strategy, there is almost no reason ever to keep them in the ground when they aren't off cooldown and ready to launch a wave of locusts which takes 25 seconds. This makes them a setup unit which rewards active use. Firecake may not do this, but in the hands of Soulkey and Stephano we see people constantly relocating swarmhosts if you have the apm to do so after each wave.
Swarmhosts are needed
Let's face it, swarmhosts are no matter how you twist or turn it needed. Without swarmhosts Z has no way to break heavily turtling Protoss or Terran players, but this enters a complicated issue. What would you replace it with? It has to be able to break a lot of siege tanks. The simplest solution is a unit that outranges siege tanks. Well, that would just make siege tanks kind of useless wouldn't it? A setup unit that can be outranged easily is very hard to make work. Which also means that if it didn't outrange tanks it would still be kind of useless to break it. Enter the genius of free units, it can "outrange" siege tanks in practice but because 90% of its bullets are dead before it arrives it gives the opponent ample time to prepare a response and adapt in some way, it isn't quite as binary as 'Oh, I am doing damage to your army and you are not to mine, this should be over soon.', if it were just bullets, no matter how many siege thanks Terran had, it didn't matter really, whereas in this case more siege tanks means more bullets get shot down and the engagement doesn't become as binary.
Now, you can obviously also just make this high range air unit against tanks like tempests, carriers, broods. But here's the issue, those are all pretty high tech for a good reason. If BL's or tempests came out at the same tech frame as tanks, then tanks would again be kind of useless. Which is ultimately what the swarm host was supposed to do, give Zerg an option to end the game without having to tech to hive when they are ahead against a player who just turtles after being behind, and that's what they ultimately do very well. Free units is a very nice compromise to achieve this goal that doesn't break the game.
Swarmhost are exciting.
I'm sorry but the majority of games we've seen with swarmhosts were ultimately pretty exciting, yeah, there's this one game with FireCake and Mana which everyone talks about but that's like one game, the majority of games featuring swarmhosts, especially with players who have the multitasking to continually relocate them have been good games. It's like saying Ravens are a broken unit because Avilo just turtles and techs to them creating the most boring games, that's not the only use of them. The bad swarmhost games have also been on this hyper turtly split-map maps where in general these kinds of things tend to happen no matter the matchup or unit composition so maybe people need to look at not giving people three free bases instead of swarmhosts.
  
|
It´s the synergy with other units/static defense that makes the swarm host boring, because instead it becomes a means of ultimate turtling. That said, I do like the swarm host, I just wish they were easier to reach/kill, so you could send hit squads at them effectively, instead of using everything to fight of wave after wave.
|
Back at it eh, with the Devil's Advocate? I'll bite. I'm in two minds about Swarmhosts. I've had a couple of games where I came up again the mass turtle behind SH style. Those games were a bitch. I tapped out both times, once out of boredom and once out of the fact that I had lost. As a specific Z strategy or style, I don't know, though, it may be too soon to tell as it may take a few months before Protoss gets to grips with it - should it become popular.
Harstem on Meta did raise the interesting point of buffing SH movement speed and burrow speed, and having more of a lag between locust rounds or faster locust death so that more interactions around the unit were opened to the opposing players. It could then be used more actively by the Z player too. I thought it an interesting idea.
|
swarmhosts imba, you are so wrong it hurts
Sorry, I just had to after all those posts from your FF are good thread :D
On topic, I think you make some good points especially with regard to the SH/siege tank relation.
Furthermore, I agree that Swarmhosts have the potential for super interesting games if the oposing player realizes early on that they are being made and the player has superb multitasking. Still, SH have the tendency to lead to super turtle games in most other cases - especially for casual or lower lvl gamers (ie everything below top masters) it is just frustrating to see that you cannot advance against a SH static D line. The major problem here is that locusts screw up the pathing of units and act as high dps way blockers. That is why I think free units are a bad design: If SHs would just attack (with range 22, same dps etc) one could decide to take the losses and close the distance to kill them. This, however, doesn't work (especially in combination with static D) because units simply get stuck and then die.
|
I think what I don´t like about locusts is that they behave like normal units. There´s no downside to them: They´re fast, deal lots of damage, and tank damage pretty well. Especially their speed is annoying. Infested terrans could be avoided by micro and it would be nice if Locusts could be as well. If they e.g. moved like a wave in their initially set direction, unable to turn around, they´d be a lot more interesting, because they´d have to be aimed and could be dodged. Anything really that allows you to out-micro them.
|
On January 15 2014 17:13 aZealot wrote:
Back at it eh, with the Devil's Advocate I'm going to make more, I think it's kind of a good idea to take the most unpopular stance you can find once in a while and try to defend it. For the record, I really dislike any form of setup unit including swarmhosts. I absolutely despise it when I need to use them to win which sadly does happen quite often in ZvP.
I'll bite. I'm in two minds about Swarmhosts. I've had a couple of games where I came up again the mass turtle behind SH style. Those games were a bitch. I tapped out both times, once out of boredom and once out of the fact that I had lost. As a specific Z strategy or style, I don't know, though, it may be too soon to tell as it may take a few months before Protoss gets to grips with it - should it become popular. Turtle behind swarmhost isn't that good though because you're not picking away at the enemy, containing behind them works better obviously because you're pressuring the opponent.
|
Swarmhost is the unit I got most excited about when hots is announced, a cheaper free unit spawning unit?
Sounds good to me, especially when I theorycrafted how awesome it would be to have bio tank vs swarmhost ling style.
And in fact, it worked really well in the beta, it was refreshing and interesting, there were more drops and I had to move around swarmhosts a bit more to prevent the bio running in and kill them all.
Sadly I got really disappointed when I tried it in ZvP and once the HotS metagame stablised, bio tank just disappeared.
I think the biggest problem with swarmhost only occurs in ZvP is due to the fact that there is no standing army as strong as the Skytoss.
I have posted a blog about skyterran, what if zerg couldn't finish off the terran and he got up a skydeathball which is pretty invincible?
Should there be a deathball that can't be beaten by design, even with a lot more economy and resources.
Swarmhost deathball allows Zerg to have an army as strong (if not stronger) as the toss which isn't necessarily a bad thing.
The problem of swarmhost deathball imo is equal to the problem of sOs triple stargate mass void ray style.
It's a rush to these deathballs. Almost like infestor broodlords.
The best ideal game on how it goes would be:
Mid game aggression from zerg, ling hydra and forces more colossus production.
Toss gets up some colossus and a third base, zerg on 5 bases and starts trading.
Zerg techs up to ultra tech or roach hydra viper or tech switch to muta, leading to more stargate heavy toss as toss get more gas up.
Zerg eventually goes into swarmhost broodlord style with toss on similar economy with skytoss army.
I think this is some of the more interesting games that we see, rather than some rush into skytoss vs rush into swarmhost deathball.
The fix must come from the mid game somewhere to encourage more of these transition periods
|
United Kingdom1381 Posts
What I find annoying about SwarmHost is the APM disparity. One player has to micro their units to their maximum potential while the other simply sets a rally point.
I'd like to see a test mod where the Spawn Locust ability was manually cast.
|
Swarm Hosts don't make for boring gameplay.
Their problem is that they are strong when the opponent goes for boring gameplay (Mech turtling, Protoss ultimate deathball turtling). And that they are weak when the opponent goes for interesting gameplay (drop play, mobile play).
Swarm Hosts create the problem that they are an immobile/turtle response to immobile/turtle gameplay, while it should be the other way around to create interesting games. Just that Zerg mobile tools lack the power for that and thus SHs are required to be played.
|
On January 15 2014 19:57 iHirO wrote:
What I find annoying about SwarmHost is the APM disparity. One player has to micro their units to their maximum potential while the other simply sets a rally point.
I'd like to see a test mod where the Spawn Locust ability was manually cast.
Only in case you use them defensively. On the offense, the player using them has to constantly reburrow or they will die against any sensible opponent. And they're hard to micro as well, because they are even more stupid than infestors. There's always one or two thinking he should join up with the locusts and visit Tossbase. And on the defense, they're only really strong against slow stuff.
|
Extremely interesting read.
I think that Swarm Hosts are fundamentally not broken. There are many ways to fix them, I just understand that many fix may very well introduce new imbalances.
There are two things I would like to point out :
It's effectively similar to energy except that it can't be stored further than the amount it required
Well, unfortunately, energy makes for a more interesting mechanic because there is a choice by the player. You choose to keep energy. Do you storm those 3 marines? Or can you deal with them and wait for a better occasion? Swarm hosts have no choice, every 25 seconds, you use your "spell", no reflexion needed.
Swarmhosts have a sickeningly long cooldown + Show Spoiler +This is very good for a setup unit. The point with siege tanks is that their cooldown is fairly short for a setup unit, which means that you waste cooldowns relocating them. Thereby it becomes optimal to keep them where they are, moving them around wastes fire. The thing with swarmhosts is that swarmhost-kiting is an effective strategy, there is almost no reason ever to keep them in the ground when they aren't off cooldown and ready to launch a wave of locusts which takes 25 seconds. This makes them a setup unit which rewards active use. Firecake may not do this, but in the hands of Soulkey and Stephano we see people constantly relocating swarmhosts if you have the apm to do so after each wave.
I feel that Blizzard does not reward enough players that use Swarm hosts dynamically. Relocating swarm hosts is a good thing, but the problem is that being Firecake is still very good.
A simple change to the Swarm Host that can, in my opinion, make it much more interesting, because being dynamic results in better results : swarm host produce a single locust every 12.5 seconds. Thus, if you burrow/unborrow them all the time, you would get exactly the same results as you would today. But if you don't, then you produce waves that are twice more frequent, but twice weaker. Which is much less effective, and rewards dynamic play.
Of course, you can go further, swarm hosts produce a locust every 10 seconds, release them immediately if borrowed, but the swarm host can hold up to three locusts on his back while unborrowed. Thus this really encourage using nydus, doing swarm hosts drop, etc... You would effectively be attacking every 35~40 seconds from a different angle to be very effective. But, you would be rewarded for using them well.
|
Swarm host are a lazy last minute design. And they are shitty. They don't do an effective job at what they are supposed to do. They do not fit well with any zerg style of play or units besides massing towers. And no there wasn't 'just that one game with mana' There have been lots of games where swarmhosts just suck dicks while the zerg player has the entire map mining and can't even break the opponent. The other races spell casters are much much much better than zerg's. Vipers are great if they actually worked well. The problem is that in order to abduct or cloud they have to be in a close range of thors, ravens, feedback, vikings, tempests, blink stalkers, etc. They are essentially nullified in almost any lategame situation. Infestors are pretty much the same thing.
And all your argument about outranging tanks or whatever, there was already broodlords and they make 'free' units as well. So do infestors. Neither of which really played much of a role in seigeing or breaking defenses because of the things I mentioned already.
Imho, if the swarm hosts fired much faster and the locust had much less life and moved faster and lowered damage they might actually be kinda useful. Because in the late game situations the locust hp is owned by splash or high damage units anyways, their hp is almost inconsequential, so they should be faster, spawn faster, or something and just lower their HP so the early game they aren't so cheap.
Anyway, I feel like you don't actually believe what you are saying you are just trying to argue something that you know is bad for your own understanding or others which is cool.
|
Should just do a devils advocate "starcraft 2 is awesome".
|
On January 15 2014 21:34 fezvez wrote:
A simple change to the Swarm Host that can, in my opinion, make it much more interesting, because being dynamic results in better results : swarm host produce a single locust every 12.5 seconds. Thus, if you burrow/unborrow them all the time, you would get exactly the same results as you would today. But if you don't, then you produce waves that are twice more frequent, but twice weaker. Which is much less effective, and rewards dynamic play.
Of course, you can go further, swarm hosts produce a locust every 10 seconds, release them immediately if borrowed, but the swarm host can hold up to three locusts on his back while unborrowed. Thus this really encourage using nydus, doing swarm hosts drop, etc... You would effectively be attacking every 35~40 seconds from a different angle to be very effective. But, you would be rewarded for using them well.
This is pretty nice. It's very similar to how players can kite and otherwise delay attacking at the end of the attack cooldown with their BLs in order to wait for the second broodling to spawn, and make the attack that does come twice as powerful.
As a general impression, I don't like the notion of locusts respawning so quickly that the hosts have a new wave ready to go as soon as the old wave times out. It doesn't give the opponent enough of an opportunity to dart in and pick off Swarm Hosts. If they spawned locusts every 12.5 seconds, but only starting once the old locusts have timed out, that emphasizes the need to reposition them, instead of just leaving them in place and going off to put some old pizza in the microwave, and maybe pour yourself a cocktail. Perhaps the enduring locust upgrade could be repurposed into an upgrade that caused locusts to be iteratively respawned quicker, lowering the 12.5 seconds into 8 seconds or something. At any rate, I definitely disapprove of the spawn locust ability being on autocast.
|
I've been away from ladder for too long to really comment on these posts but I like reading through the discussion they generate. Also it's good to see people really question the hate trains that people sometimes jump abord without even thinking about it.
|
United States4883 Posts
Free units and super cost-efficient energy units are soooooooooooooooooo bad for the game. Here is a list of units I hate:
- Swarm Hosts
- Broodlords
- Ravens
- Old Infestors (newer ones are much better)
- Ghosts (with EMP in particular)
- Tempests
The problems with these units is that they encourage turtling and "sparse trading" to occur. That is: you expend energy or free units to cost-efficiently trade with units one at a time. This is bad for RTS. If Blizzard wants to encourage more aggressive gameplay, they need to remove these hyper cost-efficient units and allow for more trading, micro, and positioning to occur between two real armies. I'm all for being able to hold an expansion with two templar on the high ground and a few cannons, but I don't think the Terran player should be able to hold the center with a few ravens and BCs and literally never "get out of position" or be threatened. The same thing occurs with Broodlords or swarm hosts hanging out in a bunch of static defense. They can't be touched or threatened because the free units trade effectively with real units and it's very easy to just never be out of position with that army.
I favor storm over EMP and fungal because storm is dodgeable and it's also very easy to detect where templar are coming from, as opposed to the cloaked ghosts or burrowed infestors. It also deals damage over time instead of flat damage (for the sake of argument, fungal does flat damage because a fungaled unit WILL take full damage from the fungal); this means there's a direct correlation between reaction time and the actual amount of damage taken. Another similar spell with good design is Irridate from BW. Irridate deals (splash) damage over time, can be dealt with using good micro, and has the ability to (slowly) snipe a single mid-tier unit. EMP in BW is very different from EMP in SC2 because of smart-casting. In BW, it was very difficult to land a good EMP and get a large pack of units; in SC2, it takes no ability to deal thousands of damage to the Protoss army instantly, and there's no way to dodge it.
The idea of semi-free units is a good one though. Units like interceptors for carriers or the skeletons for Undead in WC3 are great units with legitimate drawbacks. Interceptors cost minerals and time, but are otherwise "free" if they don't die. Skeletons in WC3 required you to kill a creep or enemy unit in order to use (graveyard aside). If we can imagine a world where Swarm hosts have to spawn locusts for money or inidividually remake them, we can see a much better swarm host. It can be cost efficient without being cost efficient at no cost.
Anyways, rant over. Fuck free units  .
Other than that, I love the swarm host and think it has some of the highest micro/multi-tasking mechanics in the game when used correctly.
|
|
|
|
|
|
EPT
