Welp. Guess it is time for another of those Design Blogs. Truth be told, I had the idea for this blog back in HotS beta, but lost interest in writing these sorts of blogs around that time. (I still have another blog fully written that I never bothered posting. It was just missing pictures.)
Purpose:
This time I thought I would look at the Swarm Host and also the Lurker. Not because the Lurker would actually work in SC2. They tried putting it in the game and apparently it did not work. I do not really know why, but I have an idea that the Tanky Trio would be a large reason why Lurkers would never work in SC2.
Ain't nothin' to a Boss.
Rather I want to look at the type of gameplay that follows from the specific design of the Swarm Host and compare it to the Lurker.
Regardless of whether they are 'meant' to be compared, the Swarm Host was explicitly designed for the specific goal of "board control." And 'board control' or controlling space is something the Lurker was very good at.
Attack Pattern Design
On the surface the Swarm Host and Lurker have some very obvious similarities. 1) They cannot attack unless burrowed 2) Therefore they must be moved into burrowing position to attack 3) They have a long attack range from that burrowed position.
But the differences between the type of attack (what I'll call their Attack Pattern) are where many of the differences in gameplay originate.
Swarm Host Given their immobility when attacking and defencelessness when moving, you generally want keep the actual unit far away from the enemy (where they will be vulnerable) and attack from far away.
The Swarm Host's Attack Pattern is a small arc in front of the Locust that moves forward with the Locust. The only place where the Swarm Host damages is immediately in front of the Locust
The SH spawns Locusts that have a range 3 attack. The Locust have 15 seconds to attack or die, but given the low hit points of the Locust, you don't really need them to be alive for very long. This means there is an ideal range between maximum range (where the Locusts will die before reaching the enemy and where the Locust will last just long enough to do it's damage before being killed by the enemy (or tanking damage for other units to get into range.) I guess we could call that the Maximum Effective Range. Closer than that and the SH is needlessly close to the enemy without any discernible benefit.
Swarm Host Attack Pattern
For the Swarm Host there is little advantage to go any closer than the Maximum Effective Range. As long as there is something there to kill or soak up damage, it should stay at the MER. If the front lines dies, then you can frog hop them forward, or maybe reposition to block some other route, but where ever the SH are positioned, the Locust (when in MER) does the same damage whether it is close to the SH or far away. And the closer the SH is to the enemy, the more vulnerable the SH is to being sniped.
Essentially then, the design of the SH creates Passive gameplay for the Zerg when using them. The Zerg wants to keep the SH at Maximum Effective Range and slowly creep the SH lines forward step by relentless step. Yes, you can micro the Locust themselves, but this is not a unit that promotes sudden, flashy plays. The Locust spawn time is constant, the Locust speed is constant, the SH is immobile when attacking. So you win by adding more of them, splitting Locust slightly better and win with SH by the incessant dripping of Chinese Water Torture. If there is any flashy play with the Zerg, it will be using SH in conjunction with something else. Not with the SH themselves.
If I could think of one word to describe the Swarm Host it would be Safe. It is unit designed to play Safely. To keep it Safe and make Safe game decisions.
Lurker Attack Pattern So how is the Lurker any different? After all it also is supposed to be a 'board control' unit and therefore, passive play should be the expected result.
In fact, the Lurker has an odd tension between wanting to be far and wanting to be near. It is true that similarly the Lurker would prefer to sit at Maximum Range in order to be safe from being sniped. So you will often see Lurker killing fields trapping a player in their base (ZvP is my bane.)
However, Maximum Range is not Maximum Effective Range for the Lurker. It does NOT attack the same far as near. The Attack Pattern of the Lurker is actually a line of spikes that damages (with splash) everything in a line from the Lurker to its maximum range.
Splash damage down the attack line
Therefore, the Lurker's maximum damage is actually when the lurker is in the centre (or at least directly in front) of the enemy.
Lurker's Projected Power
Gameplay Implications
1) Stealth Bombs This is why stop lurkers works so well. (This is a trick to prevent Lurkers from attacking until manually commanded- usually when the enemy is right on top. Think burrowed banelings.)
Savior's Stop Lurkers
Of course this is also where the Lurker most vulnerable. This tension between playing safe, but less damage versus sacrificing safety for damage creates dynamic gameplay. There is are reason for safe decisions and there are reasons for risky decisions.
2) Pushing the Enemy You may see Lurkers retreat from high ground to high ground (because of high ground advantage.) But there comes a point, where the Zerg launches forward. Retreat, retreat, then the trigger is pulled... HERE COMES THE ZERG! Zerglings in first to tank damage, but Lurkers close behind to get right in amongst the enemy, to bury and slaughter anyone foolish enough to stick around.
For the equivalent, think of the old endless retreats that SC2 Zerg did, sacrificing base after base until, trigger... SO MANY BANELINGS!
This dynamic allows for big plays. Risky plays. Plays that will make the audience scream. Plays that will make commentators scream.
This is the aggressive, quality that the Lurker has that Swarm Host will simply never have due to it's passive, Safe design. Banelings and Lurkers have a reason to close the gap, the Swarm Host rarely does. This is sort of quality makes the unit an exciting unit to use and an exciting unit to spectate.
Artosis screaming SO MANY BANELINGS will never get old. I doubt any commentator will scream SO MANY SWARM HOSTS with any degree of sincerity.
3) Counter-play not just Composition In addition, because Lurkers attack in a line, they can actually be dodged (similar to marine-baneling splits.) Therefore the attack pattern of the Lurkers creates Play and Counter-Play. Something extremely important for spectator sports rather than simply making the right composition.
NaDa marines vs 7 lurkers
More NaDa (plus Lurker pushing in.)
Yes you can control the Locust, split and focus. But because the lurker attack is actually melee, you can dodge it entirely. Because it is splash the difference between dodging and not can be catastrophic or epic. The lurker attack is just jam packed with Spectator goodness.
4) Deadly Harass Furthermore because of the splash damage in a line. 1-2 Lurkers in a worker line can wreak havoc in a very short time. In other words, the design of lurkers give it many different roles in the game and gives many options for a clever Zerg player to figure out the best way to get the raw killing power of the lurker into range of the enemy to let the slaughter begin.
(On a side note, lurkers would be come irrelevant late game due to Critical Mass ranged units, but Dark Swarm keeps them relevant to the end of the game. Melee will always need the ability to close the gap versus Ranged in the late game.)
On Banelings, Swarm Hosts, Lurkers I guess the question could be asked, well why does the Swarm Host need to have that aggressive quality, when the Baneling fills that role?
And I suppose it doesn't. But just how good is the Swarm Host at board control anyways?
Board Control
aka Controlling Space
Absolutely, the Swarm Host in Critical Mass can stop the enemy cold in the middle of the map. Furthermore, by moving back and forth from choke to choke a Zerg can control the middle of the map. But critical mass is the key. You need a lot of them to be effective.
This also goes back to design. Two lurkers sitting on top of a ramp could hold a base because 1) high ground advantage, 2) splash damage in a line, 3) difficulty getting up ramps, 4) no Tanky Trio (marines get massacred.) 5) Morphing lurker eggs were heavily armoured and could actually physically block the ramp so that the Zerg could bring in reinforcements. A very cool play.
Due to the design of the Swarm Host, I don't think we can ever have Swarm Hosts that are effective individually because they mass way too well. Infinitely Spawning units puts the enemy on a brutal Timer where something MUST be done than a so-called infinite attack such as Siege Tank or any other ranged attack.
The reasons for this are Spawned units 1) Are an infinite supply of hit points that attack 2) Block movement
The reason this is brutal should be obvious to anyone that has played the sorts of Hack'n Slash arcade games like Gauntlet Legends. It really doesn't matter how many enemies you kill, you HAVE to kill the spawn point. Every damage you take from the infinite unit is unanswered damage that will only weaken you further and further and never weaken the enemy.
start at 6:25 The real enemy are the gates
Furthermore, the spawned units physically block the path required to get to the thing that is spawning all those units. In the SH's defence, I think this is one of the reasons the SH works better than Lurkers in SC2. It is much harder to get to the SH when you are being physically blocked by units (the problem is compounded if Broodlords are involved.)
There is a reason the Locust spawn time is slow. It just scales too out of control otherwise. If the SH is good as an individual unit, the game would quickly turn into a only make Swarm Host game.
So SH by design is a passive safe unit, and cannot really be used in small numbers. That leaves slow pushes with massed numbers.
Strangely, the Lurker seems to fit in the exact middle of the Baneling and the Swarm Host as though it's qualities were split it into two separate units in SC2. Banelings want to close the gap the same as Lurkers. However due to their kamikaze nature, they cannot be reused and are therefore terrible at controlling space. On the opposite end of the spectrum of one shot and you're done is the infinite unit spawn that is Swarm Host. The SH is great at sitting back on slow pushes and holding off the enemy indefinitely, but have no reason to push forward.
Conclusion: I like Banelings. I always have. Especially because of the marine-baneling split dynamic. Probably since the middle of Beta I have not particularly liked the design of Swarm Hosts. Back then I felt the design promoted passive, Safe play. Since then, on the few streams and tourneys I have watched and listening to Idra and a couple others, I have seen very little reason to change that opinion.
The Swam Host can indeed shut down middle of the map movement. So in that sense a 'board control' unit has been achieved. And I do think Zergs were missing something that could take and hold territory that the rest of the Zerg arsenal could not achieve short of making zillions of Spine Crawlers or just sacrificing endless numbers of bases until they could remax and pull the trigger again. However, I think inherently the Swarm Host promotes passive, stale gameplay because it lacks that up and at 'em quality of Banelings and Lurkers.
There is little reason to risk moving forward and so the SH will continue to whittle away safely from the back. (Or if they don't, then they get cut down by the enemy.)
Thanks for the perspective, always enjoy your design writeups.
I heavily disgaree with the backlash to swarm host. I think they are a very interesting unit in HotS and it will be more interesting to see how their play evolves over time. I do agree that they do promote more 'passive' gameplay, but I think that's a positive thing for Zerg at the moment. In my opinion, a swarm host transition is always a risk due to the cost and immobility.
I think Zerg needs options that reward a 'board control' type play. I do agree that risky play can be awesome to watch, but I don't think zerg needs any more 'risky' alternatives, but rather just a fairly interesting unit for safer options. The swarmhost fills that role.
This blog is giving me a lot to think about so I'm eager to see the responses
Well I don't think passive play is necesarily bad. but a combination between passive and aggressive is better. For instance traditional Mech Play might be considered passive, but because of vultures there is a very active and aggressive arm of the style play. Lurkers also create passive play, but also can be used aggressively. But I don't see that versatility in the SH.
I don't know much about HotS but I do like that you touched on the concept of counter-play. When we consider both counter-play and "damage maximization" there's another behavior that is more or less true of any set of units (but is amplified in the Lurker): maximum damage depends on the formation of the enemy. Of course this is true in everything from basic arcs vs limbo lines or basic sources of circular splash damage, but the Lurker is unique in its niche of a line splash that also has a projectile travel time. Anyways the reason why the Lurker damage "sweet spot" seems to be right next to it is because a Lurker attack traveling through an entire enemy formation inflicts way more damage than a Lurker attack hitting the enemy's front lines.
Again I'm not well versed in SC2, but I feel like the Swarm Host can also provide counter-play in terms of enemy unit placement and formation. The Locusts seem like they can also provide a dynamic movement impediment. Still, your conclusion does seem fair from my perspective as a reader. I would hesitate to say that just because a unit behaves passively with regard to its own mechanics, that it also promotes passive gameplay. The BW Siege Tank also has no incentive to close ground like the Lurker did, but it's still core to a number of aggressive Terran builds (and defensive ones too.) A unit should be regarded in terms of the entire game in order to make judgments on gameplay.
I really miss the lurker, but if they're insistent on keeping the swarm host, it would be nice to see them experiment with Carrier-like mechanics (leash micro, locusts costing 25 minerals each, burrowed movement). Would still allow them to be a siege unit while going further away from requiring 'critical mass' effectiveness.
Edit: You should post this on reddit to facilitate further discussion.
I like this blog! Thank you for writing it. I never thought about how the tanky 3 might be the reason why lurkers were removed from WOL alpha. Very interesting! As for the swarm host, I actually like it. But I do agree that it promotes passive play. Micro could be used but it really isn't necessary for swarm host to be successful. That's why I feel the locust time upgrade should be removed and replaced with a speed upgrade that only affects off creep speed. That way, swarm host have a larger time window where they are vulnerable and promotes a more mobile swarmhost play where a player can launch locusts and run away. Maybe increase swarmhost movement speed by a little as well.
Well I don't think passive play is necesarily bad. but a combination between passive and aggressive is better. For instance traditional Mech Play might be considered passive, but because of vultures there is a very active and aggressive arm of the style play. Lurkers also create passive play, but also can be used aggressively. But I don't see that versatility in the SH.
Thanks for the feedback.
It is possible to have interesting, aggressive play with swarm hosts. There have been games where the zerg will spawn a set of locust to attack one base and then move them to attack another. The problem is that is far too easy to just sit back and slow siege your opponent to death.
@imallison Well I think that's like a lot of things it isn't quite so black/white. I probably should have tempered my words with 'tends to promote passive play' or 'naturally leads to.' I don't think aggressive play is impossible per se, but I do think the design lends itself to sit back and slow seige. And I think there are designs that lend themselves to more dynamic and versatile gameplay. But I do think there is a limit to the creative use of the unit. We might see a bit from the top players. But if it is designed so that it lends itself to dynamic play, it will filter down to lower echelons of players
@EchOne You are right that unit interactions matter and can turn a passive unit into something more. But in regards to the Siege Tank, there is more going on. For one Siege Tank trades mobility for tons of damage. Particularly BW tanks with their slower rate of fire. That simply is not possible with SH due to the whole infinite unit spawn. Furthermore Overkill and spider mines creates an interesting unit interaction with zealot bombs and the like. I don't really see that potential just in in the way the Swam Host attacks.
edit And yes the Tanky Trio units are Roaches, Marauders, and Immortals. Or as I called them the Triple Tanky units in my A-move by Design blog, but that's a rather obscure reference I'll admit.
Although blink stalkers wouldn't be much better vs Lurkers now that I think about it. Lurkers rely on units coming towards them to do damage. Blink-stalker close the gap so quickly that they would probably snipe Lurkers way too quickly to be of much use.
On August 15 2013 04:51 Falling wrote: This time I thought I would look at the Swarm Host and also the Lurker. Not because the Lurker would actually work in SC2. They tried putting it in the game and apparently it did not work. I do not really know why, but I have an idea that the Tanky Trio would be a large reason why Lurkers would never work in SC2.
Don't forget the colossus.
Though I don't think lurkers would be too out of place in zvt given how the matchup is currently played. Lurkers would make neglecting detection, tanks, and (to a lesser degree) marauders in order to stream modest groups of marine/mine across the map much riskier.
Yeah Collosus too. Dang there is a strong correlation between things that would kill Lurkers and units I don't like (stalkers excepted.) It almost goes without saying that Collosus is not a good unit. But yeah that crazy long range and mobility would murder Lurkers. (Maybe that's why they had that huge range upgrade they were going for before they nixed the Lurker. What was it range 10 -12?) But basically Lurkers would never work ZvP.
ZvT lurkers could possibly work, but I think it would just force more marauders and then they'd die.
Interesting read. The swarm host to me...is well kinda not zergy to me. Sure it has swarm and it gives you free units. But its boring and it doesn't feel fast paced or overwhelming powerful. You just have to place them in a good spot and let the locust do there thing. Of course you have to protect them. But from a viewer point of view watching someone die this way is pretty boring. As a Terran player facing this, I would just make my own siege tanks and just blow the locusts away and than of course drop there expansions. The unit kinda sucks as a whole if you ask me.
I don´t think the SH has been explored as of yet, so it´s hard to assess. And while I´m not a fan of free units, I think SH have shown great potential. They have been used to... - contain an enemy in his base - zone out an enemy´s deathball by making trading inefficient - to "connect" bases defensively - to harass enemy bases with one wave of locusts and then retreating
I like especially the last and I´m sure there´s plenty of more interesting uses for it, so one the one hand it´s sad that it´s used rarely in pro-games and on the other hand I wonder why, because it has been heralded as kind of a game-breaker numerous times. If it´s really that map dependent, then that would also be interesting to me, because it would promote more specialized strategies.
What I don´t like is the scaling. A few SH seem pointless, while large numbers can become an annoyingly powerful a-move deathball. There should be a way to incorporate them into an army in a meaningful way. Maybe it´s just because they haven´t been figured out, yet, but I guess it´s because of the space the locusts need and hence take away from other, more useful Zerg units.
On August 15 2013 14:40 Falling wrote: Yeah Collosus too. Dang there is a strong correlation between things that would kill Lurkers and units I don't like (stalkers excepted.) It almost goes without saying that Collosus is not a good unit. But yeah that crazy long range and mobility would murder Lurkers. (Maybe that's why they had that huge range upgrade they were going for before they nixed the Lurker. What was it range 10 -12?) But basically Lurkers would never work ZvP.
ZvT lurkers could possibly work, but I think it would just force more marauders and then they'd die.
marauders are already present in many counter builds so i dont see why it would be an issue.
if anything i feel the important difference with lurkers is the much increased dps vs sh, even roaches. i would gladly give up endless free locusts for a chance at a much more "exciting" "aggressive" unit.
lurker-ling, lurker-hydra would be amazing compositions to play with. im not talking about design or balance, i mean it would be just hella fun, running around the map just like we did in bw.
and terrans could actually get a chance to use siege tanks against lurkers. because holy shit siege tanks vs swarm hosts is the worst fucking most boring thing i have ever witnessed.
Falling I think a lot of us had this idea. There are only a few scenarios where swarm hosts are viable unlike lurkers. It's funny because we see so many players try to make the switch to swarm hosts and all the sudden they lose. I only know of a few players who know how and when to use swarm hosts in their strategy. It's painful to watch.
Well at first I didn't like swarm host much, but I changed my view.
Swarmhosts and vipers make very interesting ZvP games, so I like them now. ZvP dynamic is very strong now, most units can be utilized and it is without a doubt my favorite matchup to watch.
Reading this made me feel like I don't need to watch or even playmuch starcraft to get better, I just need to read really smart people's analyses of units.
I really liked the nada micro video. It looks almost like the zerg is bad, but still, nada plays it so perfectly ..
The lurker was So epic. If you were playing this game when you were like... 13 or so that thing just seemed so imba. i remember losing like 100s of marines to lurkers and just like.. losing.. that game was too hard.
The gauntlet analogy was good but what i liked most is that the clip reminded me of the days when games had such bad graphics that they took on a sort of otherworldly quality.. or maybe i was just young.probably that.
EPT
![[image loading]](http://i.imgur.com/Tfh8WiO.png)
![[image loading]](http://i.imgur.com/uEDHO1w.png)
![[image loading]](http://cdn.makeagif.com/media/8-14-2013/tXghGU.gif)
![[image loading]](http://i.imgur.com/ipgS4wR.png)
