|
On December 30 2011 06:50 Felnarion wrote:
I would argue the overall cost is actually more than the sum of the corrupter+morphing.
Why?
Because in order to get brood lords, you have to go through corrupters. Your argument is that I pay essentially half up front for a corrupter, half at delivery of my broodlord, therefore the cost feels less.
I could see how that would be true, if corrupters had a purpose. Unfortunately, just about anything a corrupter can do, an equal cost of mutas can do better and faster. Nevermind that your ultimate goal is brood lords, which is obviously an anti-ground unit, which the corrupter is not.
What I'm trying to say is that, while you're trying to get this anti-ground unit, you have to spend a significant amount of time with a purely anti-air unit which is useless to you.
The amount of time is as low as 0 if you time out your resources correctly. Even if Corruptors are near useless they're still less useless than no unit at all. Especially since if you have BL's you can expect your opponent to likely have vikings, medivacs, colossi, or their own corruptors/bls.
|
As a very high masters zerg, i think broodlords are a terrible designed unit. Its sad that in the late game, versus protoss or terran your only option is broodlord / infestors.
Yet... once you do get out a good numbers of broodlords, you give up any map control you had. On a big map like taldarim, broodlords means protoss can have a huge amount of stalkers, blink around, kill 2-3 bases before he even engage your army, In the meanwhile, if his econ wasnt too bad, he can also expand, and rebuild all his gates to another base if your broodlords are killing his main.
You also cant really split your army with broodlords, if you leave them undefended, without roaches/infestors, or infestors queens corruptors, they will get killed by a bunch of stalkers or marines ghosts vikings. So it leaves you with yes, one decent deathball, but super slow, and very hard to control. A few misclick or seconds of inatention, they get too far ahead of your army and get picked off, etc.
Basicly if feels like,* well lets hope hes stupid and goes straight for my army*, but in most cases its not true, terrans will take map with pfs and turrets, turtle with tanks and ghosts.
Overall, i feel like broodlord is just a finishing move unit when your already very far ahead but cant finish a turtling terran or toss. You never get the same feeling as having a protoss max deathball where your like * Alright, im in pretty good shape, i can do whatever i want now*
I really cant wait to try the new Hots units to see if it changes the dynamic, i would love to have ultralisks a little bit more powerful, or rather, ghosts, archons, stalkers, immortals and mauraders a little bit weaker versus ultralisk. Ultralisk is a really great fun unit, very mobile and fast, good for map control, if only theyre were ground bases army with ultralisk that are as powerful as a Protoss deathball or a terran army with lots of ghosts and siege tank.s
|
Not entirely sure what questions your asking? Are broodlords worth it? In some situations yes. If you have established air control, broodlords in large numbers are very effective. Not so much as other transitions or other builds. Every unit in the game has a purpose, and is effective when used correctly at the correct timings. The broodlord is no different then the ultralisk in that respect.
It is my opinion, that many zergs use ultalisks incorrectly. Their dps is not very high for their cost. They are not usually very cost efficient in and of themselves. They are amazing damage absorbers though, and if your strategy already involves a lot of lings and bannelings to begin with, then ultralisks make a lot of sense because they benefit from the same upgrades. Ultralisks in ZvP, PvZ should be used to break forcefields and direct collossus fire away from lings. In ZvT they should buffer seige tank shots. In my opinion, ultralisks should not be the main component of your army, but a subset used for a specific purpose. It makes me sad when zergs, even good zergs, make 10 ultralisks and the game goes south as a result. Make 3, 4, maybe 5. And then make other stuff. Just my two cents.
|
On December 30 2011 06:53 chestnutcc wrote:
Felnarion: For that to be true, there must be an active cost of having corruptors instead of other units. What you're saying is that the opportunity cost of the corruptors must be fed into the total price. I'm not sure what to say to this, since it is v situational, and should there be colossi or void rays, or should the P/T attempt to counter with voids/vikings, corruptors obviously have great utility here. One does not usually see mutas perform this function.
Mutas are often used to kill voidrays and colossi, in fact, I would argue Mutas are better against VRs than corrupters.
And any zerg will tell you, corrupters biggest cost, is their opportunity cost. We would much rather have other units in our mix. Because once the colossi are dead, the corruptors have almost no value until they're morphed into brood lords.
Of course, your analysis of this cost assumes the Protoss is going Colossi at all.
|
If you're losing to broodlords it's more often than not a result of bad scouting. If you have an obs in his main and you see the greater spire starting. JUST PUSH. He'll have like 20 corruptors that will be wasted until broodlords come. If you feel like you don't have an army strong enough to push, just take like 3 bases, add on fleet bacon and make mothership. Also, get a warp prism or two just to keep him busy until mothership has enough energy and you have enough archons.
|
On December 30 2011 06:37 chestnutcc wrote:Show nested quote +On December 30 2011 06:33 straycat wrote:On December 30 2011 06:20 chestnutcc wrote:
I would disagree on this point, imo broods are much cheaper than their additive cost.
Hmm, could you explain further how they can be less expensive than their additive cost? Or did you argue that they are somehow "cheaper" than another unit with the same mineral/gas cost but that has less utility? I.e. that they have higher "value" per gas/mineral cost. Or do you mean something else? I tried reading the OP again but I could not quite follow. I also wonder what you mean when you say that the time to morph a brood lord should be considered as less than the sum of making a corruptor and the morph to blord. The idea is simply that of 'Do you prefer a dollar now, or a dollar tomorrow?'. If you have multiple payments at different points in time, their total value cannot always be simply the grand sum of all the payments. This is known as time discounting of payments and is the fundamental principle behind returns on bank deposit or investments of any sort.
The total value will be more than the sum of the payments, not less.
This is assuming you are rushing for brood lords, in which case, yes you will probably die. If you were rushing for broodlords, the time between payments is minimized and the total cost is effectively additive, so its a huge investment. My post was more about a normal game, where these payments would be staggered significantly.
Sure, you can try to get away with taking hits of 200/200 here and there, spread across a long time period, in order to increase your survivability. But this increases the cost of your broodlord investment - you're putting in the same money, but now it takes several minutes more for it to pay off in the end.
And all of this money invested equals units that you are not making in the present, as an alternative to teching up. If you don't go for BLs, you could add another handful of roaches and infestors to your army and do more damage in the short term. Or you could hatch a bunch of mutas and go pick off workers and do harrassment damage.
You have to give up those short-term advantages (which could potentially make a huge difference in the game) in order to tech up to BLs. IMO the time cost is too high, and the morph time for G.Spire and BLs should be cut down.
It takes longer for a saturated 2-base 100+ supply zerg on hatch tech to get a broodlord out, than it does for terran to get a banshee with cloak when starting a new game with 6 workers.
|
mothergoose729: I was pondering the cost structure and macro factors that lead to the choice of brood lords over ultralisks. and what macro decisions other races can or should take to combat them, not just on the battlefield, but also in their economy.
|
I agree with those who have said they can't really count the cost of the tech in the price of the unit. Your entire gameplan should have transitions to get to later stages of the game. Getting up expansions, siezing map control, getting up expensive and time consuming tech or upgrades, and doing so safely. Whether by turtling, harassing, pushing, or poking the front, it revolves around the time you have bought to buy that advantage.
I get what you mean about this being a split cost making it more efficient than having to pay all up front, but I believe that to be an intentional design choice for the Zerg by the developers. They have stated numerous times that they want the three races to be asymetrical, each with unique advantages and disadvantages, and the ability to morph banelings and brood lords from other units is one such aspect.
|
+ Show Spoiler +On December 30 2011 07:05 Chocobo wrote:Show nested quote +On December 30 2011 06:37 chestnutcc wrote:On December 30 2011 06:33 straycat wrote:On December 30 2011 06:20 chestnutcc wrote:
I would disagree on this point, imo broods are much cheaper than their additive cost.
Hmm, could you explain further how they can be less expensive than their additive cost? Or did you argue that they are somehow "cheaper" than another unit with the same mineral/gas cost but that has less utility? I.e. that they have higher "value" per gas/mineral cost. Or do you mean something else? I tried reading the OP again but I could not quite follow. I also wonder what you mean when you say that the time to morph a brood lord should be considered as less than the sum of making a corruptor and the morph to blord. The idea is simply that of 'Do you prefer a dollar now, or a dollar tomorrow?'. If you have multiple payments at different points in time, their total value cannot always be simply the grand sum of all the payments. This is known as time discounting of payments and is the fundamental principle behind returns on bank deposit or investments of any sort. The total value will be more than the sum of the payments, not less. Show nested quote +On December 30 2011 06:34 Chocobo wrote:
Broodlords cost a lot, especially considering that you're taking a serious risk to attempt to tech to them.
I think hive tech upgrade and time should absolutely count, because it's very rare that you'll upgrade to hive for a reason other than broodlords. Starting with lair tech already, it costs 200/150 (100 sec) for Hive, 100/150 (100 sec) for G.Spire (and add on another 200/200 if you didn't already have a spire, sometimes the case vs protoss). Then 150/100 (40) for a corruptor and 150/150 (34) to morph to BL.
Grand total investment, from the time you decided "I think I'll try getting some broodlords": 600 minerals, 550 gas, and 4 minutes 34 seconds.
That is a looooooooong tech time considering you are already starting at tier 2! Also, you're not going to tech up to make just 1 broodlord... so with that and the potential cost of a spire if you didn't have one, you could be investing 1200/1200... maybe 1500/1500 or even more, and it doesn't pay off at all until the entire process is complete.
I know I've lost my share of games by making this investment and then dying in the next 4:30 because my opponent spent those resources on army and attacked me, and I fell just short because my money was wrapped up in BL tech.
So for this kind of investment... they damn well better be strong units without a mindless hard counter to them. This is assuming you are rushing for brood lords, in which case, yes you will probably die. If you were rushing for broodlords, the time between payments is minimized and the total cost is effectively additive, so its a huge investment. My post was more about a normal game, where these payments would be staggered significantly. Sure, you can try to get away with taking hits of 200/200 here and there, spread across a long time period, in order to increase your survivability. But this increases the cost of your broodlord investment - you're putting in the same money, but now it takes several minutes more for it to pay off in the end. And all of this money invested equals units that you are not making in the present, as an alternative to teching up. If you don't go for BLs, you could add another handful of roaches and infestors to your army and do more damage in the short term. Or you could hatch a bunch of mutas and go pick off workers and do harrassment damage. You have to give up those short-term advantages (which could potentially make a huge difference in the game) in order to tech up to BLs. IMO the time cost is too high, and the morph time for G.Spire and BLs should be cut down. It takes longer for a saturated 2-base 100+ supply zerg on hatch tech to get a broodlord out, than it does for terran to get a banshee with cloak when starting a new game with 6 workers.
This is the opportunity cost point of view, in which the opportunity cost of the investment must feature in its cost. I will say off the bat that I am not considering rushing to brood lords, or 2 base zergs or anything like that because those are extremely situational and obviously incur huge opportunity costs due to the fragile nature of the build.
My argument is that broodlords typically appear when the zerg needs them i.e. when they need to break seige lines or run over a turtling toss. Rushing for them is not what I consider here. In this sense there is no opportunity cost to morphing the brood lord. The opportunity cost of getting corruptors out I'm not so sure, in PvZ with colossi out, corruptors don't incur a significant opportunity cost, but in TvZ corruptors may indeed have such a cost since the entire corruptor and morph cost is usually done within a short time span.
dasbif: The baneling is another interesting example, thanks for bringing it up.
|
BLs are costly but the fact that they don't cost 6 supply is very valuable for zerg. You can get more unit to support them than with an Ultra based army.
|
From my non-numerical feelings on my own games, brood lords just FEEL like a lot cheaper of an option than ultras. Let me explain.
Occasionally I reach a point in the game where I say to myself "no matter how big my economy gets, I can't win on lair tech. I need to go to hive." I'm faced with the decision, brood lords or ultras. I'm not going for both because unless I'm on 5, 6 bases with an absurd number of drones (say, 100 mining after I make/cancel spines for extra supply) that's not gonna be economically viable.
In almost 100% of my games, you'll see my answer be "brood lords." Why? Because of the following excerpt from the OP:
On December 30 2011 05:55 chestnutcc wrote:
the disappearance of several broodlords does not affect the next remax, if the game is on even footing
This means I can keep being my zergy self and remax and remax again and again until the game is over. If I build Ultras, for one reason or another, and I couldn't tell you why, this doesn't happen. I get ONE army of ultras and if they don't win I'm gonna be hard pressed not to die right after they fail.
EDIT: Upon further thought, it strikes me that this may be why Blizzard buffed the build time of ultralisks; when they had ridiculously long build times you absolutely could not remake them after losing them because by the time they were ready to spawn, their eggs were liable to be the only things left for the Zerg player on the map. The reason the dramatic build time cut didn't increase their popularity too much was probably due to the fact that, again, brood lords are simply better and don't feel like a one hit wonder after which all hope is lost if enormous damage isn't dealt.
|
It's actually way easier than this broodlords perform better in most combats. broodlords are the only unit that zerg can build to force combats. broodlords can be morphed from corruptors, so you can get them and their protection really easily/naturally. broodlords are supplyefficient while ultralisks are not (so you are not forced to fight with them right away). Broodlords have very few and rather specialized counters while ultralisks get countered by nearly everything on their own and still get countered by good positioning with infestorsupport.
Conclusion? Broodlord>Ultralisk in most scenarios.
|
dunno about zvp but ultralisks are amazing in midgame zvt...
|
On December 30 2011 07:45 KawaiiRice wrote:
dunno about zvp but ultralisks are amazing in midgame zvt...
well but i think that OP was specifically asking about times when you have the opportunity to decide between ultra or broodlord and as far as I know there is no strategy that allows you to go for midgame broods.
|
@chocobo: Usually, future payments are discounted downwards, not upwards, so the sum is cheaper, not more.
@bigj: I ignore the direct utility of broodlords in my post. I'm interested in why they perform better than ultras in the macro scenario.
@felnarion: True, but your scenario assumes that the toss does nothing about the harass so you're inflating the opportunity cost. This is sort of why I ignore it since its too volatile a facet. If you want to go mass muta, you don't make corruptors, and your game plan is to win via attrition. Brood lords are a more head on approach, so the two are mutually distinct imo so I'm not sure what to say about that.
@kawaiirice: Yes, its an interesting contrast to the usual corner of shame ultras find themselves in. Imo this has to do with the unit compositions used against terran and how ultras and broods are suddenly on even footing, since corruptors are created in this matchup more or less for the sole purpose of morphing broodlords, so the separate costs are no longer staggered. Also reactored vikings are much most cost efficient counters, but that deviates from the zerg macro part.
|
reactor vikings are extremely cost inefficient vs zergs in high counts of brood lord and when you overproduce vikings zerg can overrun you with an ultra switch due to lack of army supply or just kill them all off with infinite fungal / transfuse
|
from what I've seen from how protosses cope with my brood lord army is from least to most effective: attempted phoenix/voidray switch < archon toilet < blink stalker archon. Seeing as these are the most likely base compositions you'll see as a zerg when you go for the classic 'ultra switch' their utility isn't as much as staying with brood lords; it can simply throw off the opponent
|
On December 30 2011 08:02 KawaiiRice wrote:
reactor vikings are extremely cost inefficient vs zergs in high counts of brood lord and when you overproduce vikings zerg can overrun you with an ultra switch due to lack of army supply or just kill them all off with infinite fungal / transfuse
These overlapping considerations are sort of why I choose to ignore all but the zerg's point of view. You would be able to comment on this better, but imo the overall game plan is made roughly in isolation, with deviations appearing as required (reactionary play); my concern is more with the plan in isolation, ignoring possible deviations.
|
On December 30 2011 07:55 chestnutcc wrote:
@chocobo: Usually, future payments are discounted downwards, not upwards, so the sum is cheaper, not more.
@bigj: I ignore the direct utility of broodlords in my post. I'm interested in why they perform better than ultras in the macro scenario.
@felnarion: True, but your scenario assumes that the toss does nothing about the harass so you're inflating the opportunity cost. This is sort of why I ignore it since its too volatile a facet. If you want to go mass muta, you don't make corruptors, and your game plan is to win via attrition. Brood lords are a more head on approach, so the two are mutually distinct imo so I'm not sure what to say about that.
@kawaiirice: Yes, its an interesting contrast to the usual corner of shame ultras find themselves in. Imo this has to do with the unit compositions used against terran and how ultras and broods are suddenly on even footing, since corruptors are created in this matchup more or less for the sole purpose of morphing broodlords, so the separate costs are no longer staggered. Also reactored vikings are much most cost efficient counters, but that deviates from the zerg macro part.
Hm, so you mean wether i go for ultras are broods if i have no direct use or some sort of "counter"-argument?
Well then I guess my answer would be neither of them, due to broods not being mobile enough to be versatile and ultras being not supplyefficient so i dont want to sit on a bunch of them and wait for my opponent to max on supplyefficient units and then lose in one big engagement.
Though if i have enough support units (very likely) to hold most harras scenarios i would definatly try to get a huge corruptor/broodlord army with infestor support as i think that if you control (unrealisticly high
amounts of) them well they should not lose to anything but a superspecific hardcounter composition anymore in a direct engagement.
|
On December 30 2011 05:55 chestnutcc wrote:
This is made worse by the ultras role as a huge tank and damage dealer, so they will inevitably be the first things to die, and tanks are pointless without something to tank for.
One could argue that broodlords also add tanking, in the form of broodlings. Those pesky little buggers seem to be bullet magnets, absorbing a lot of damage. And they're free.
Related question: Will the ultras' burrow charge make them more viable, due to increased life expectancy?
|
|
|
|
|
|
EPT
