|
Zerg’s “Inject Larvea” (I’ll just refer to it as Inject from here) has been a hot topic lately, and for good reason. It’s safe to say blizzards recent attempts to remove macro mechanics were brought on by inject. The other two mechanics, Chono and Mules, seem like fairly healthy mechanics- with the potential for abuse, which needs to be addressed. Inject, however, has some issues. It creates a problem for zerg by denying their ability to micro consistently, and is extremely punishing.
Before we can start fixing Inject, we first need to understand both what it’s trying to do, and what its problems are. Let’s start with its purpose: (Keep in mind creepspread is also a factor in accomplishing some of these goals)
Injects encourage interaction with the macro components of zerg for both sides- a simple formula blizzard has put into the game. You don’t just build production facilities and have them make units, you have to nurture them. (Protoss has to manage power for buildings as well as warpgate units, chronoboosting structures, and dealing with both mass production and full cycle production- Terran has add-ons, rally points, and more types of production facilities)
Injects add complexity to an overall simple production system. Without injects, the larvae concept ends up being fairly dull and boring, as well as extremely lopsided. Without some way to distinguish larvae rate from just-establishing-a-second, and fully-saturated-two-bases the larvae concept would be very weird to manage, and would eliminate the need for managing larvae.
Make up for the mechanical difficulties zerg’s DON’T have to deal with. Zergs don’t have to move their camera as much to build their supply depots, and can worry about where to place them when they have to APM to spare. Zergs also don’t have to worry as much about carefully balancing their production vs income. Zergs also don’t have to manage building placement (as much).
Now let’s look at the problems Inject creates:
Inject forces an unreasonable amount of APM in a single instant. Inject both forces you to look at and click on each base you own- either forcing you to instantly recognize which bases don’t have a queen, or forcing more manual methods of finding hatcheries. At the same time, at same instant the inject pops, you need to spend that larvae. The longer that larvae sits, the longer you lose the larvae that could be produced from the hatchery itself. This is in addition to the things that could be happening out on the map.
Inject pigeonholes the rate of zerg’s production. Let’s say a 3 base zerg needs 15% more larvae to spend their money. While a protoss or terran can add a few building to get to that perfect number, zergs instead have to invest in 33% more production- way overkill in this case. This also leads to my next point.
Inject, and larvae, aren’t scoutable. There is no inbetween production rates, so there is no way to tell if a zerg is going for a larvae inefficient build, or a very larvae efficient build. The only way to tell what a zerg is building, is to actually see units. This has led to a balance state where aggressive builds are uncommon and rarely effective. Add a scoutability, and we can get a much more diverse range of builds.
Inject dips too many times into the punishment bucket. Miss an inject by 5 seconds, and your larvae are forever 5 seconds behind. Those who miss injects are generally also maintaining a lower APM, and thus likely also have a hard time spending larvae, so they end up losing potential larvae from the base hatcheries as well. These stack up very quickly.
Overcoming poor injects (by, say, a lower league player) is counterproductive. If you need an extra hatchery to have the larvae to spend your money, likely due to lower APM, you end up with one more hatchery to inject- you now need more APM because you don’t have enough APM.
Inject forces a spiked production, not a smooth cycle like the other races. Zergs end up with large waves of units spawning, not a staggered stream of units. This is less a balance issue, and more a basic feel/look for the game. Its far from a major problem as much as an annoyance.
Please, let me know if and why you disagree with any of these points- or if you think I missed any here.
Based on this, I think blizzards concept of autocasting inject is a poor fix. It removes the interaction with their macro on the players side (though it is still there for the opponent), it removes any compensating for other mechanics zerg’s don’t have to deal with, but it does still offer some scaling/complexity to the production system. As far as fixing the issues inject has, autocasting does hit pretty much every point- almost too many, honestly. The result is we’ve fixed the issues, but we also haven’t accomplished the goals inject should be going for.
While I’d like to mostly start the conversation from here, so we can all work to address these issues, I do have two suggestions for fixing many of these problems. Keep in mind any numbers I put here are purely to represent the idea. They would obviously need to be tuned to reach a balanced state.
Suggestion 1
This one is simple, would be easy to implement, and would lead to a relatively small change in balance. If it takes 30 seconds for a queen to gain enough energy to inject, let’s have the inject only take 20 seconds to pop. This adds some forgiveness, as you can eventually make up for missed seconds, as well as an in-between for increasing production, as you can create an extra queen to bounce around hatcheries while the normal queen doesn’t have the energy (this also creates an element of scouting information). In theory, this forgiveness can also create a more staggered production rate- this is hard to accurately predict.
Suggestion 2
This change is a bit more drastic, but more easily balanced. Let’s remove the idea of inject all together. Instead, hatcheries naturally produce larvae faster depending on the number of queens currently on the map. (Let’s say a hatchery naturally produces 1 larvae/ 10 seconds. With a single queen on the map, that changes to 1/8 seconds, two queens 1/7 seconds, etc) This would obviously have to be capped at some point.
(Note we’d likely have to nerf creep spread to make up for this)
From here, there are a few directions we can go. My current though would be to have a global increase in the maximum number for larvae at any base, based on lair/hive tech. (Lets say hatch tech zergs max at 3/base, lair makes all hatcheries max at 6, Hive to takes it to 12) Keeping the larvae/base low compensates for the lack of inject by forcing zergs to spend their larvae as it comes out. This way, we could also force more APM to creepspread if needed.
The end result here is a smoother climb in production rate, as well as opening options for getting to that desired production rate. Interaction from the players perspective becomes more tactical, and spreads out the APM requirement, while keeping- if not exaggerating- the interaction for the opposing player.
|
This is something I have talked about for awhile now and I think it is time to put it into action, so here it is. We remove all macro mechanics and the way we balance Zerg is with the HatchaQueen, the HatchaQueen is a Queen that is spawned on top of the Hatchery, able to spawn a creep tumor beside the hatchery for creep spread and also having a defensive ability which can be activated on cooldown.
How about Suggestion #3 be the HatchaQueens
|
Suggestion 1 is already the case. Injects pop at 23 energy.
|
The problems with inject I see are:
1) requires too many clicks, Zerg generally takes the whole map; combined with creep spread, scouting, overlord placement, dealing with spread out harassment, managing many bases..., blah blah blah.. the APM is already taxed
2) during battles it is difficult for Zerg to keep up timing on injects; especially drawn out micro intensive battles or harassment.
3) The timer for inject is longer than the other mechanics, so the ebb and flow is easier to overlook / mistime. You can't establish as easy of a rhythm. There is no great way of checking how far along the inject is to completing
4) It gets disproportionally harder to use as the game goes on compared with other mechanics
5) The larvae mechanic is far less forgiving than the other two mechanics
Blizzards new change partially addresses some of these
|
Regrading the problem that injects create production spikes, it can be easily fixed by making injects a skill that buff larva spawning rate.
I think this change combined with some of the suggestions OP mentioned can be a potential middleground in this MM debates .
|
I spend more time hassling with my creep tumors than I do with my rax/depots or pylons/robotics facility so it should be worth mentioning that too.¨¨
How about if you set automatic inject you get only 2 larva and when you ahve manual you get 4 larva?
I find hots chronoboost A LOT easier and more fun.
Terran hots mule was really fucking fun when you could just make "mule base" in corner of map and spam mules in that base.
I played
No you don't and you shouldn't in any matchup. Sometimes it's good idea to bank larva and check wtf enemy is doing. Basic 101 when defending proxy 2rax. In late game you bank maxed larva and then trade armies.
Miss an inject by 5 seconds, and your larvae are forever 5 seconds behind.
Only in early/midgame where it's indeed easy to manage. 5 seconds is incredibly a lot in real time and less in larva time. Yes that's in contrast huge fuck up but only relevant in early/mid game. In late game it's about banking the larva. A lot of the times you wonder if you should even go back to base to inject instead of spreading 10 tumors because you know you have 40 larva atm.
Inject, and larvae, aren’t scoutable.
In ZvZ/ZvT you check if cocoons pop drone or zerglings. You check if he saves larva at some points of game.
If you know your enemy goes zerglings you know it's inefficient larva strategy but it's totally irrelveant because zerg will himself compensate it by building more queens/hatcheries.
Inject forces a spiked production, not a smooth cycle like the other races.
I find it fucking cool about zerg. It's very smooth, elegant and it's the only production cycle that requires thought. You don't blindly make 8 drones. You make intelligent decision to do it or make 8 roaches instead. Every zerg has lost game when they made 20 drones and then see enemy moving on map.
The problems with inject I see are:
1) requires too many clicks, Zerg generally takes the whole map; combined with creep spread, scouting, overlord placement, dealing with spread out harassment, managing many bases..., blah blah blah.. the APM is already taxed
I'm sure most people agree that it's more fun when there is more actions required to manage EVERYTHING than human can do. Inject itself is manageable at every relevant spot in game.
|
On September 13 2015 13:10 Draddition wrote:
Suggestion 2
This change is a bit more drastic, but more easily balanced. Let’s remove the idea of inject all together. Instead, hatcheries naturally produce larvae faster depending on the number of queens currently on the map. (Let’s say a hatchery naturally produces 1 larvae/ 10 seconds. With a single queen on the map, that changes to 1/8 seconds, two queens 1/7 seconds, etc) This would obviously have to be capped at some point.
(Note we’d likely have to nerf creep spread to make up for this)
From here, there are a few directions we can go. My current though would be to have a global increase in the maximum number for larvae at any base, based on lair/hive tech. (Lets say hatch tech zergs max at 3/base, lair makes all hatcheries max at 6, Hive to takes it to 12) Keeping the larvae/base low compensates for the lack of inject by forcing zergs to spend their larvae as it comes out. This way, we could also force more APM to creepspread if needed.
The end result here is a smoother climb in production rate, as well as opening options for getting to that desired production rate. Interaction from the players perspective becomes more tactical, and spreads out the APM requirement, while keeping- if not exaggerating- the interaction for the opposing player.
This is BY FAR the most elegant and intuitive solution that has ever been proposed on the subject. I've been saying this for weeks now. The fact that Blizzard hasn't tried just removing inject and tweaking hatcheries is outrageous. I feel like they want to make queens remain mandatory. But why would they want that? Queens should be an "option" like every other unit.
Also, hatchaqueens is a bad idea. It has no nuance, and it feels contrived and bandaid-y. Like the mothership core garbage that we're now stuck with.
|
I think a more extreme version of #1 is best
Spawn Larva is Instant, but it costs 50 energy, only gives 3 Larva, and can only be cast on a larvaless hatchery(/lair/hive)
So larva are 'banked' in the form of Queen energy, but Queens aren't particularly better for production rate compared to hatcheries.
Would require weakening the Protoss and Terran Macro as well to compensate
|
Your first suggestion would just mean you build 2 queens so you have plenty of energy for 20 second injects.
I like the currently testing solution of overlapping injects, but it still has the problem of being mechanically annoying and complicated.
What if instead of requiring you to move the camera, select queen, cast, and click hatchery, we just had a button ON the hatchery for "inject" that caused the nearest queen to come cast it?
Then, just like other races, you could just select your grouped hatcheries and hit one key without having to move your screen.
Alternately, a single keypress on the queen could cause her to inject the nearest available hatchery in range, instead of requiring a targeted click.
|
Here's what my zerg friend suggested:
He doesn't want inject to be too easy to go with as it is zerg production and provides good amount of skill differentiation between near equally skilled players, but it is bit hard hitting on lower leaguers; but he wants there to differentiate from lower to higher zerg macro in terms of inject
His suggestion is to allow queuing of Zerg inject up to 3 times that autoinject but at a cost; each spawning spawns 1 less larva as queuing is used, where it would give 4-3-2 larvae. But if you inject before que starts (small buffer due to queen having dew second energy buffer or built in buffer), it is reset and counts back to 4.
What about that?
Early game where not much is going on, good inject is must to maximize zerg on done hatch asap or so he says. But later on as game goes, more lenient queing could be afforded to eventual lategame spam 3 and forget for good time.
|
|
|
|
|
|
EPT
