|
Hey all,
Background check - my name is DinoMight and I'm a high Diamond Protoss player on NA playing for team NEFN (hell yeah team NEFN!).
![[image loading]](http://img.photobucket.com/albums/v143/daftboi/nefnclanlogotrans2_zpsc2f3f3fd.png)
I also happen to work in the finance industry in NY and spend a lot of time assessing risk and valuing investments. Today I decided to apply some of my knowledge to something that a lot more people might appreciate - StarCraft.
StarCraft, much like investing, is a game of strategic risk taking. For example, a Command Center first build says "the return I get from having more economy than my opponent outweighs the risk that I will die to cheese." A Dark Templar opening says "the damage my DTs will do to his economy outweighs the risk of investing in useless tech." In each of these cases, the player choosing the strategy is assessing how much risk the build involves and executing it. And much like in the finance world, what differentiates the top players from the rest of the crowd is the ability to properly value risk. "Should I pull all my SCVs and try to end the game now?" Well, that depends on your assessment of his defensive capabilities. MVP in his prime was amazing at this - he could time attacks to hit seconds before Storm finished because he just knew. MarineKing, on the other hand, is an AWFUL judge of risk.
The value that something has is affected by the risk involved in obtaining it. Say I offer you $1 now or $1 a year from now. You'd take the dollar now, obviously. This is because you have no way of guaranteeing that in a year I'll keep my promise. This means that the further out in time you receive something, be it a benefit, or a product, or a monetary sum, the less you should be willing to pay for it... because there is a smaller chance that it will actually live up to what you expect.
Let's look at a High Templar, for example. High Templar can cast Psionic Storm. Storm takes energy, and energy regenerates over time. In theory, a single Templar could be said to have infinite value because he will continue to provide Storms until the game ends. But now factor risk into it. How many storms do you realistically expect to get out of one Templar? Maybe that number is 3? Maybe that number is 5. Or 1. Whatever that number is, that number impacts how much you're willing to pay for a High Templar. Blizzard has effectively said that given the average life expectancy of a High Templar and the number of Storms that they're going to land, a High Templar should cost 50/150 and 2 supply. Sure, Billy the hero Templar could go on to land 20 Storms and single handedly win the game. But to a Wall Street investor and a StarCraft pro-gamer, what are the 19th and 20th storms really worth? Zero (or some infinitely small fraction near zero).
They may not think about it economically or in these terms, but Pros are running through this assessment every single time they decide to make a unit. How come we almost never see Battlecruisers in competitive play? You could say it's because they take too long to build, or because they require too much tech, or because they have too many hard counters. But at the end of the day, their cost (which is really high) does not justify the benefit that they provide given how long they must survive to provide that benefit.
So as I said before, StarCraft is a game about assessing risk and spending finite resources based on this assessment. Here's a quick demonstration of what risk can do to the value of something.
Say you received $100 today. Or you received $100 at the rate of $10 a year for the next 10 years. Which is worth more? Well, the first option is worth $100. Say that the current interest rate is 2% (interest rates are a measure of risk). Well, your $100 over 10 years would only be worth $89.8 (the $10 in the 10th year is only worth $8.2).
Say the risk of collecting the $10 increased dramatically (I'm a very shady guy with a lot of debts and I probably won't pay you). Maybe in that case a 10% interest rate is more appropriate. The value of that $100 over 10 years drops all the way down to $61.4
But what if the risk decreased. What if there was absolutely zero risk? Well, $100 in 10 years would be worth exactly the same as $100 today. Going back to that Templar argument. What if you had a Templar that could cast unlimited Storms during a game and that could never die. How much would you be willing to pay for it? 50/150? 150/450? In theory, you would pay an infinite amount of money (of course there is the practical point that you must be able to accumulate this much money before dying).
Now let's get to the 800lb Gorilla in the room.
THE SWARM HOST.
In a game where strategic assessment of risk differentiates players to such an extent that a player with nearly broken wrists playing Terran can lift 5 GSL trophies while another Terran player with some of the best mechanics on earth remains an underperforming fan-favorite (if only for his mechanical prowess)....
...deep breath... that was a long bit of sentence...'
Can we really have a unit such as the Swarm Host?
The Swarm Host generates unlimited free locusts (same as the high Templar). Now, I know the Zerg players in the room at this point are thinking:
"Silly DinoMight, you've just taught me about the time value of money and about how a benefit that I receive in the distant future is worth almost nothing to be now. How can you be so hypocritical as to suggest that the 1000th Locust is worth just as much as the first?"
To which I would say "Wow, I'm really impressed by your grasp of finance so far"
JKJK.
But here's the point that I've set this whole blog post up to make. In our example, the 10th payment is a lot less valuable than the first payment because a lot of time elapses between now and then and the risk that the payment will not be collected is high. With Swarm Hosts, that risk is much less.
Swarm Hosts, with the enduring locusts upgrade, have the longest range in the game out of any unit. They can attack from nearly half the map away. And on certain maps, they can sit comfortably at the Zerg base while denying mining at several key expansion locations. In addition to that, they are protected by being invisible when burrowed. Finally, locusts have incredibly high DPS and are quite swift on creep. This means that in order to get to the Swarm Hosts, you have to travel very far into enemy territory with detection and enough firepower to not only kill the locust waves, but kill them at a fast enough pace that your army can make forward progress. Keep in mind that while Colossus can kite endless numbers of locusts, a small enough Colossus count will never actually reach the Swarm Hosts.
Now combine Swarm Hosts with spore crawlers to repel observers and the proposition now becomes "I need to have enough firepower to clear out the locust waves at a high enough rate that I also have time to kill static D in between waves of locusts and maintain forward progress." n practice, we see how this develops on the battlefield. Once critical mass is reached, Protoss players simply can no longer attack into Swarm Hosts, regardless of army size. With a 200/200 army supply cap, Swarm Hosts are so efficient that nothing the Protoss makes can kill them in a fight.
Let me try to avoid what some may perceive as balance whining at get back to finance. Recall the Templar. How many Storms did we realistically expect out of that guy? And how his value changed when we essentially turned him into Goku from Dragon Ball Z? Now let's consider the cost of a Swarm Host (200/100). How many locusts is a Swarm Host expected to realistically release over the course of a game, and how much damage is each one of those locusts going to do?
If you're a Swarm Hosting player, your Swarm Hosts will in many cases live from when you spawn them until you lose or win the game (fuck, I lost the game). Or perhaps a little bit less than that. Compare that unit's longevity to that of other units in StarCraft and you quickly begin to realize the problem...
The Swarm Host dramatically, and to a further extent than any other unit in the game, takes risk out of the risk/reward balance in a game based entirely about economics and risk/reward.
But the Swarm Host, BECAUSE of its longevity, is really really vulnerable to changes in risk. This is a financial principle.... the longer someone takes to pay you, the more changes in perceived risk will impact your valuation of that payment. Let's look at a Swarm Host mathematically. Over a 25 minute game a Swarm Host will release 60 waves of locusts. Let's assume then that this is 60 payments of "locust." Say each wave of locusts is worth 10. If we assume that there is zero risk that you are going to collect all your payments, the value is 600.
If we allocate the same risk as the US government...this is a bit of a stretch financially speaking because there's no basis but let's say 2.49% (10-year US treasury). That value drops all the way down to 317!
Now let's assume really really high risk. Let's say, Greece risk. 6.39%. With Greek risk, the value of your Swarm Host drops all the way down to 152.7.
These are made up numbers for illustrative purposes just to show the impact that risk can have on the value of something. In practical cases you would be looking at the cost of the Swarm host 200/100 and whether or not that is too cheap for a unit that can provide such cost efficiency.
Let's look at the situation that occurs once critical mass of Swarm Hosts is on the field (as described above, the point at which the Protoss can not fight it head on).
Either:
1) The Locusts are damaging buildings and eventually will shut down the economy of the Protoss (and this is a game about economy).
or
2) The Locusts are fighting the Protoss army defending said buildings (or elsewhere on the map) and trading efficiently. This means that Zerg will bank minerals at a higher rate than Protoss will.
Now I know many of you might be thinking "Well, yes, nobody is dumb enough to fight Swarm Hosts head on (well, maybe they are sometimes). The key is to abuse their immobility to damage the Zerg economy while Protoss gathers enough forces to afford an army that can finally kill the Zerg."
And yes, absolutely! That's how you SHOULD play against them. But here is the problem in my view:
Maps have finite resources
This is where economics come back into the equation. Swarm Hosts and the associated static D/ Vipers trade cost effectively against every Protoss army. We've seen the resources lost tab on enough Snute games to know that. So the goal for the protoss is to outmine their opponent at a ratio high enough to be able to afford the extra resources it will take them to combat the Swarm Hosting player (mathematically speaking, if Swarm Hosts traded at x:1 with Protoss units you'd need to mine X times more to fight them).
Here's the catch. By the time the Swarm Hosts are out... it may not be possible to do that. Swarm Hosts can reach the bases farthest from where the Protoss player starts on most maps just by sitting there and rallying out. In addition, there may simply not be enough resources left on the map, in terms of absolute numbers. So effectively, the Protoss has a timing window in which he has to outmine the Zerg by a ratio high enough that he can win the game before resources run out (in which case he will trade inefficiently until he has no more units).
And in a strategy game where controlling scarce minerals to create an army is the key objective, this effectively means that the only way the Swarm Hosting player can lose is if they are outplayed.
I welcome your comments.
Thanks for reading.
-Dino
EDIT
Takeaways:
-Given the low risk involved in using Swarm Hosts to their maximum possible utility, they are mispriced. They should either cost more supply, so that critical mass can't be achieved, have less range so that there is more RISK in using them, or simply cost more minerals - delaying critical mass and requiring a longer amount of time for them to pay back their investment (and allowing Protoss some wiggle room).
-Protoss has to economically outplay their Zerg opponents in order to win agaisnt Swarm Hosts. This is only true once the Sarm Hosts are produced and if the game began on even footing. Zest vs. TLO is not even footing. A game where the Zerg went 3 hatch before pool and got away with it is not even footing.
-This is all theoretical. I've beaten Swarm Hosting players many times and lost to them many times. It's not meant to be a balance whine, only an analysis of a unit that I see as unfitting of an economically driven strategy game.
 
|
|
|
|
|
put this guy out of his misery please, delete his blog already.
btw, maps having finite resources is not an issue for economic growth.
as the game goes on every remaining unit becomes more and more cost effective.
so there is no inherent connection between resources and unit production
|
United States4883 Posts
I've very against super cost effective units in which the best possible strategy is simply to mass them.
This includes swarm hosts, (WoL infestors), ravens, ghosts (in TvP), tempests. The problem with these units is that they are so cost effective that they have no gigantic weaknesses and discourage unit diversity. The best possible strategy with such a unit is simply to mass it because it does well against everything with no major weaknesses.
BLords, marines, and void rays are units that are a slight step down. They are super cost efficient against almost anything, but they have enough of an inherent weakness that their strengths are compromised and they have to be supported by other units.
To me, the biggest difference between these two sets of units is "free damage". Swarm hosts and tempests provide free damage due to their range while ravens, ghosts, and WoL infestors rely on energy, which is essentially free after ~20 or so units. The biggest problem that these units create is that they create a late game environment in which things are decided solely on cost effectiveness, not on the amount of resources being mined; this is the biggest difference from BW, where expansion management was the key to whether or not your were ahead or behind. In SC2, getting to 5 bases before your opponent is rarely a radical advantage because the late game is not based on how much money you have but rather, what your army looks like and how well you are trading your "free damage" vs your opponent's "free damage".
Example of a perfect unit: The Viper. The viper is a perfect unit, by far the most well-designed unit in SC2. Its spells do literally nothing on their own, so it promotes diversity of units. 10 blinding clouds are not any more useful than 2, same thing with abduct, meaning that vipers are most effective in small numbers. It has unique strengths and weaknesses and can be just as powerful as it can be fragile.
EDIT: So, yes, I pretty much agree with what you're saying here (I think?), but I think the problem stretches past more than the swarm host. I think the swarm host is just the easiest example of what's wrong with SC2.
|
so there is no inherent connection between resources and unit production
maps having finite resources is not an issue for economic growth.
...
uh...
so anyways...
it's a very well thought out post Dino... I dont know if I agree with your assessment though. I mean it makes sense in a vacuum. But there's way more variables at play.. you just throw in spores as if you don't need absurd creep spread or drones to even make that viable.. and to get creep you need queens and so on so forth. There's also zero mention of splash or other 'cost effective' methods of managing defense vs. swarm host armies. Like Collo attacks, MotherShip cloak, tempests, upgrades, warp prisms and warp tech... i mean.. there's a lot of cost effective things you can do to combat it.
while I agree the design should probably be looked at... I don't think it's quite as bad as you paint it.
i did enjoy the read though
|
Ridiculously long winded blog to make a simple point.
The issue is that many Z are finding that swamhost is the only way to play against P late game compositions. Until you address this point your blog just seems like whining.
|
Thanks for reading, guys.
Like Ignorant said, I am ABSOLUTELY looking at Swarm Hosts in a bubble. This is a very academic way of looking at things and I know that in most practical applications it's not relevant. However, what I'm addressing are those few games that we see where the two players are fairly evenly skilled and the Protoss player simply runs out of minerals trying to defeat the Swarm Hosting player. Sure, it doesn't happen all the time and it's not a huge cause for alarm. But my view is that a game fundamentally based on economy management shouldn't come down to that.
I know that Zerg since the Infestor nerf and the introduction of the Tempest has had a hard time with certain Protoss compositions and that Swarm Hosts are one of the only ways to deal with it.... trust me I get it. But I think that the concept of a long rage siege unit that needs to be dealt with proactively can be better executed than it is currently without allowing for the ridiculous endgame scenarios that we face today. Simply put, on certain maps, once the game reaches a certain state there are literally not enough minerals on the map for Protoss to win.
Again... to clarify this is only in a game between even matched opponents where the game has been even up to the point where Zerg made Swarm Hosts. This almost never happens. But scientists and scholars spend many hours theorizing about things that never happen 
|
On September 30 2014 09:28 nunez wrote:
put this guy out of his misery please, delete his blog already.
btw, maps having finite resources is not an issue for economic growth.
as the game goes on every remaining unit becomes more and more cost effective.
so there is no inherent connection between resources and unit production
I don't understand this. One Immortal vs X roaches - for example - the Immortal will kill a certain number of roaches then die. This number will never change based on how many minerals are left on the map.
Therefore the rate at which Protoss trades resources vs. Zerg will not change.
If Protoss needs to trade 3:1 and there are only enough resources to allow a 2:1 trade, then by definition the Protoss can never win. Yes, it's highly theoretical... but in theory it's correct.
|
I think playing around the reosurces you can eventually acquire is a very interesting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you often as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravens or templar etc.
Or as a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests).
|
On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests).
I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army.
But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games.
Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it.
|
On September 30 2014 21:32 DinoMight wrote:Show nested quote +On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests). I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army. But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games. Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it.
Well, to be completely honest with you. I think the reason why Protoss players try to get a resource advantage over the Zerg is mainly because that strategy is a winning strategy on all (unsplittable) 3-4p maps and works quite well on 2p maps too. However, we have seen instances of mass Tempest strategies on splittable maps like Habitation Station or Heavy Rain, in which case the Protoss does not work with a resource advantage, but trades pop-for-pop with the Zerg.
HasuObs vs Jaedong at IEM was such a case if you are looking for a specific game.
I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat).
The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses.
|
On September 30 2014 22:40 Big J wrote:
I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat).
The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses.
I think in theory you're correct. But the problem is in practice, that is SO MUCH HARDER to do than what the Swarm Host player has to do. The best example I can think of right now is a series played between CJ Hero and Snute where Hero was brilliantly controlling his Mothership/Tempests/High Templar/Oracle etc. It was such a good game.. but in the end Hero lost because he made tiny little mistakes in control here and there and every time it cost him a little bit of money. Eventually he ran out of money and Snute won.
I just don't think games should be decided like that. Basically it puts the burden to win on the Protoss.
|
On September 30 2014 22:46 DinoMight wrote:Show nested quote +On September 30 2014 22:40 Big J wrote:
I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat).
The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses. I think in theory you're correct. But the problem is in practice, that is SO MUCH HARDER to do than what the Swarm Host player has to do. The best example I can think of right now is a series played between CJ Hero and Snute where Hero was brilliantly controlling his Mothership/Tempests/High Templar/Oracle etc. It was such a good game.. but in the end Hero lost because he made tiny little mistakes in control here and there and every time it cost him a little bit of money. Eventually he ran out of money and Snute won. I just don't think games should be decided like that. Basically it puts the burden to win on the Protoss.
Well, but we are talking theory here, right? In practice, I'd say unless it is a splittable 2p map, just go with the Prism + Stalker/Colossus. Run rampant and outexpand the Zerg. I think on those maps P has an advantage in the lategame.
On splitable 2p maps, sure, I can see that the Zerg has an advantage with such a SH playstyle if played well.
|
United States4883 Posts
On September 30 2014 22:40 Big J wrote:Show nested quote +On September 30 2014 21:32 DinoMight wrote:On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests). I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army. But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games. Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it. Well, to be completely honest with you. I think the reason why Protoss players try to get a resource advantage over the Zerg is mainly because that strategy is a winning strategy on all (unsplittable) 3-4p maps and works quite well on 2p maps too. However, we have seen instances of mass Tempest strategies on splittable maps like Habitation Station or Heavy Rain, in which case the Protoss does not work with a resource advantage, but trades pop-for-pop with the Zerg. HasuObs vs Jaedong at IEM was such a case if you are looking for a specific game. I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat). The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses.
The counterpoint to "well, in theory SHs serve the purpose of...because otherwise tempests would just make it impossible for Zerg to win..." should not be that only one of the units requires close inspection. They both do.
Like I said in my earlier post, the Tempest actually works on the same principle as the swarm host. Cannons + tempests + a few storms are the most cost effective composition for Protoss. SH + mass static defense + a few vipers/corruptors is the most cost effective composition for Zerg. Do you see the parallel here? The argument for "the swarm host is a flawed unit because it relies on trading free units for resources, thereby undermining the economic nature of a finite resource game" cannot be stated without the mention of the tempest as well. They are both two halves to a greater whole of why PvZ late game is fucked because neither can exist independently without the game completely breaking.
The "glass cannon" argument is not a strong one. Although there are some slight imbalances in the way that "glass cannon" units like the colossus or the marine work, the bottom line is that you still have to trade resources for damage. Undoubtedly, this leads to a "deathball syndrome" of sorts, which only superior pathing, high ground advantage, and unit versatility tweaks can manage (we won't get into this), but the greater problem that no one wants to address is the underlying flaw of "free damage" in SC2. From the ghost snipe to the WoL infestor to tempests and SHs, the ever present problem is there, but we still keep running into walls with it without actually addressing it.
|
|
|
The original quote doesn't quite go like that
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests).
but rather just gives the example of the tempest being such a unit.
I think if you cut both, the Tempest and the SH - and balance the game somehow - you will just eventually reach the Colossus, the Templar, the Infestor, the Broodlord or the Viper who then start to be the ones that trade too well.
I think at the end of the day it is a question of range. If you have a Swarm Host and a Tempest against each other, that might work out. If you have a Templar and a Viper, that might work out as well. The only thing that stands out here, is that in the SH vs Tempest case, you literally only have those two units with such action radius, which makes it so streamlined.
In the Case of the Templar and the Viper, we actually also have a Colossus and a Broodlord and an Infestor and a Carrier, who might have a saying in the combat, so things get trickier.
I think that is actually the crux what makes these superlong range units so "ultimate" and what creates "free damage". They have much fewer counterparts, and it becomes quite obvious very soon around the max - where range counts more than anything else - what the combat interactions are.
In case of 8-10range units, there are many more variables, which creates many more scenarios. In the case of 4-6range units, we suddenly have so many variables that it becomes very dynamic.
|
On October 01 2014 02:50 SC2John wrote:Show nested quote +On September 30 2014 22:40 Big J wrote:On September 30 2014 21:32 DinoMight wrote:On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests). I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army. But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games. Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it. Well, to be completely honest with you. I think the reason why Protoss players try to get a resource advantage over the Zerg is mainly because that strategy is a winning strategy on all (unsplittable) 3-4p maps and works quite well on 2p maps too. However, we have seen instances of mass Tempest strategies on splittable maps like Habitation Station or Heavy Rain, in which case the Protoss does not work with a resource advantage, but trades pop-for-pop with the Zerg. HasuObs vs Jaedong at IEM was such a case if you are looking for a specific game. I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat). The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses. The counterpoint to "well, in theory SHs serve the purpose of...because otherwise tempests would just make it impossible for Zerg to win..." should not be that only one of the units requires close inspection. They both do. Like I said in my earlier post, the Tempest actually works on the same principle as the swarm host. Cannons + tempests + a few storms are the most cost effective composition for Protoss. SH + mass static defense + a few vipers/corruptors is the most cost effective composition for Zerg. Do you see the parallel here? The argument for "the swarm host is a flawed unit because it relies on trading free units for resources, thereby undermining the economic nature of a finite resource game" cannot be stated without the mention of the tempest as well. They are both two halves to a greater whole of why PvZ late game is fucked because neither can exist independently without the game completely breaking. The "glass cannon" argument is not a strong one. Although there are some slight imbalances in the way that "glass cannon" units like the colossus or the marine work, the bottom line is that you still have to trade resources for damage. Undoubtedly, this leads to a "deathball syndrome" of sorts, which only superior pathing, high ground advantage, and unit versatility tweaks can manage (we won't get into this), but the greater problem that no one wants to address is the underlying flaw of "free damage" in SC2. From the ghost snipe to the WoL infestor to tempests and SHs, the ever present problem is there, but we still keep running into walls with it without actually addressing it.
While mass Tempest, Raven and a few other units exhibit some of the signs of Swarm Host syndrome, in my view only the Swarm Host eliminates risk to such an extent that it is TOO favorable. The inherent issue with the Swarm Host vs. those other units is that it can do damage without ever exposing itself. Ravens can get feedbacked, Templar can get EMPd. Tempests get abducted. Swarm Hosts live practically forever.
My view is that to fix Swarm Hosts (and address the issues that I brought up, that in a theoretically even game the Protoss player must ouplay the Zerg to win) Blizzard should add more risk to using Swarm Hosts. 25 second Locust spawn times and 25 second Locus lifetime means there are always Locusts. You could shorten the lifespan of the Locusts so they can't travel as far BUT also shorten the respawn time of the Locusts. Essentially the only difference would be that the Swarm Hosts take more risk by being closer to the opponent.
This would force the Zerg to reposition them more often (exposing them to enemy fire) and demanding more of the Zerg player's attention (higher APM requirement). It also adds player risk (the Zerg player could fuck up and lose a lot of his Swarm Hosts if he isn't careful). In my view this would alleviate the issue of Protoss having to outplay a Swarm Hosting player (again, given a theoretically even game up to that point).
|
Canada13404 Posts
On October 01 2014 02:50 SC2John wrote:Show nested quote +On September 30 2014 22:40 Big J wrote:On September 30 2014 21:32 DinoMight wrote:On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests). I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army. But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games. Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it. Well, to be completely honest with you. I think the reason why Protoss players try to get a resource advantage over the Zerg is mainly because that strategy is a winning strategy on all (unsplittable) 3-4p maps and works quite well on 2p maps too. However, we have seen instances of mass Tempest strategies on splittable maps like Habitation Station or Heavy Rain, in which case the Protoss does not work with a resource advantage, but trades pop-for-pop with the Zerg. HasuObs vs Jaedong at IEM was such a case if you are looking for a specific game. I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat). The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses. The counterpoint to "well, in theory SHs serve the purpose of...because otherwise tempests would just make it impossible for Zerg to win..." should not be that only one of the units requires close inspection. They both do. Like I said in my earlier post, the Tempest actually works on the same principle as the swarm host. Cannons + tempests + a few storms are the most cost effective composition for Protoss. SH + mass static defense + a few vipers/corruptors is the most cost effective composition for Zerg. Do you see the parallel here? The argument for "the swarm host is a flawed unit because it relies on trading free units for resources, thereby undermining the economic nature of a finite resource game" cannot be stated without the mention of the tempest as well. They are both two halves to a greater whole of why PvZ late game is fucked because neither can exist independently without the game completely breaking. The "glass cannon" argument is not a strong one. Although there are some slight imbalances in the way that "glass cannon" units like the colossus or the marine work, the bottom line is that you still have to trade resources for damage. Undoubtedly, this leads to a "deathball syndrome" of sorts, which only superior pathing, high ground advantage, and unit versatility tweaks can manage (we won't get into this), but the greater problem that no one wants to address is the underlying flaw of "free damage" in SC2. From the ghost snipe to the WoL infestor to tempests and SHs, the ever present problem is there, but we still keep running into walls with it without actually addressing it.
To be fair, depending on the matchup, I disagree to a certain extent.
Tempests are weak to anything that shoots up. If you sit them behind mass cannon walls I can understand the issues in ZvP but thats the only scenario for tempests being too strong.
Ravens again ZvT, only because you can't remove the energy or reduce the effectiveness of a raven.
In all matchups however the swarmhost offers two things ravens/tempests/ghosts don't.
-- You don't need much to get to swarmhosts, and the window in which you can lose is not as large compared to going ravens, ghosts, straight tempests.
Effectively you can gamble on "greedy swarmhost" and get lucky or unlucky but once you have them the game changes completely.
The issue remains - not only does protoss need a lot of money to straight up fight back the swarmhosts waves, the protoss also needs to have this money in addition to the money necessary to harass and slow down the zerg economy to begin with.
So as per dino's example for protoss to deal with swarmhosts you need to mine X:1 + harass cost, or if you want to consider harass cost in the outmine X ratio, then it balloons quite a lot.
|
United States4883 Posts
On October 01 2014 03:17 DinoMight wrote:Show nested quote +On October 01 2014 02:50 SC2John wrote:On September 30 2014 22:40 Big J wrote:On September 30 2014 21:32 DinoMight wrote:On September 30 2014 17:53 Big J wrote:
I think playing around the reosurces you cqn eventually acquire is a very interedting strategy. In fact it is exactly what we see in other matchups as well. The prime example is TvZ where you oftwn as zerg do not attack to break the Terran, just to limit him to a certain amount of bases until he eventually does not have any money left - in thise scenario neither player even uses swarm hosts or ravena or templar etc.
Or a s a Mech player against Zerg on a 5-6player map. You just try to live through the storm and eventually seeker him to death until he is out of resources.
What your analysis misses is that this kind of trading is nothing bound to special abilities and can be achieved by things like dropping and picking up easily as well. Or just having enough buffer for you glasscanons to wreck a whole army - colossi..
SHs obviously have an advantage over other units in this, however, cut them and then units that come second to SHs can suddenly apply such a strategy (e.g. tempests). I see your point. I mean it makes sense - denying bases is 101 of outplaying your opponent and winning with a larger army. But the difference between what you're talking about and what I'm talking about is that in the Swarm Host case, all the Zerg has to do is not die until the money on the map runs out. Granted, this is highly theoretical, but, we sometimes see it in games. Because resources are finite, units should not have infinite value. Otherwise, there is incentive to only make THOSE units. A lot of other units have "revenue streams" (a certain number of storms etc.) but when you discount the value of the really uncertain 19th and 20th storm for example you have a finite value. A Templar is worth X storms. Unfortunately the Swarm Host breaks this equation and just incentivizes massing it. Well, to be completely honest with you. I think the reason why Protoss players try to get a resource advantage over the Zerg is mainly because that strategy is a winning strategy on all (unsplittable) 3-4p maps and works quite well on 2p maps too. However, we have seen instances of mass Tempest strategies on splittable maps like Habitation Station or Heavy Rain, in which case the Protoss does not work with a resource advantage, but trades pop-for-pop with the Zerg. HasuObs vs Jaedong at IEM was such a case if you are looking for a specific game. I think in theory, if you have a mothership cloaking your army, you keep track of the Zerg army with an oracle and use a combination of Colossi/Void Ray/Tempest/Templar you can go the distance. The Mothership makes you immune to SHs, the VR/Templar win every combat with whatever antiair the Zerg brings and the Templar prevent any abducts from happening. The rest is Tempests slowly whittling the Zerg down (granted, you need enough Tempests to one-shot targets so that queens cannot heal them and make it an eternal combat). The zerg would constantly have to sacrifice infestors/Overseers to get vision of you. I guess, theoretically, given enough resources, the Zerg could send 1000 Overseers, but even then I'm not sure if the damage done while having vision would justify that investment, since the Protoss has all the resources the Zerg spent still in the bank to rebuild any losses. The counterpoint to "well, in theory SHs serve the purpose of...because otherwise tempests would just make it impossible for Zerg to win..." should not be that only one of the units requires close inspection. They both do. Like I said in my earlier post, the Tempest actually works on the same principle as the swarm host. Cannons + tempests + a few storms are the most cost effective composition for Protoss. SH + mass static defense + a few vipers/corruptors is the most cost effective composition for Zerg. Do you see the parallel here? The argument for "the swarm host is a flawed unit because it relies on trading free units for resources, thereby undermining the economic nature of a finite resource game" cannot be stated without the mention of the tempest as well. They are both two halves to a greater whole of why PvZ late game is fucked because neither can exist independently without the game completely breaking. The "glass cannon" argument is not a strong one. Although there are some slight imbalances in the way that "glass cannon" units like the colossus or the marine work, the bottom line is that you still have to trade resources for damage. Undoubtedly, this leads to a "deathball syndrome" of sorts, which only superior pathing, high ground advantage, and unit versatility tweaks can manage (we won't get into this), but the greater problem that no one wants to address is the underlying flaw of "free damage" in SC2. From the ghost snipe to the WoL infestor to tempests and SHs, the ever present problem is there, but we still keep running into walls with it without actually addressing it. While mass Tempest, Raven and a few other units exhibit some of the signs of Swarm Host syndrome, in my view only the Swarm Host eliminates risk to such an extent that it is TOO favorable. The inherent issue with the Swarm Host vs. those other units is that it can do damage without ever exposing itself. Ravens can get feedbacked, Templar can get EMPd. Tempests get abducted. Swarm Hosts live practically forever. My view is that to fix Swarm Hosts (and address the issues that I brought up, that in a theoretically even game the Protoss player must ouplay the Zerg to win) Blizzard should add more risk to using Swarm Hosts. 25 second Locust spawn times and 25 second Locus lifetime means there are always Locusts. You could shorten the lifespan of the Locusts so they can't travel as far BUT also shorten the respawn time of the Locusts. Essentially the only difference would be that the Swarm Hosts take more risk by being closer to the opponent. This would force the Zerg to reposition them more often (exposing them to enemy fire) and demanding more of the Zerg player's attention (higher APM requirement). It also adds player risk (the Zerg player could fuck up and lose a lot of his Swarm Hosts if he isn't careful). In my view this would alleviate the issue of Protoss having to outplay a Swarm Hosting player (again, given a theoretically even game up to that point).
I think you're missing the point. With a shorter range, swarm host + static D + infestor/viper/corruptor is still the most cost effective thing a Zerg can do. Sure, it gives Protoss more options, and perhaps the map can be split better. But in the end, assuming that it doesn't somehow make SHs completely obsolete, the game will still come down to Tempests vs SHs. Giving a unit with a flawed design a bigger weakness or a harder counter does not fix the design flaw. Your argument in the OP is that "free damage" doesn't belong in an atmosphere that is inherently designed on finite resources (which is all games). It's true that SHs deal "more" "free damage" than tempests do, but why pretend that only SHs fall under this category when you can easily think of several situations in which massing a unit is the best strategy?
|
|
|
|
|
|
EPT![[image loading]](http://smg.photobucket.com/user/daftboi/media/nefnclanlogotrans2_zpsc2f3f3fd.png.html)
