|
http://www.gomtv.net/2012gsls3/vod/67559/?set=2&lang=
http://www.gomtv.net/2012gsls3/vod/67559/?set=3&lang=
Final Engagement in 2nd example (3rd set) of Symbol vs Parting using Ultras before Broods vs 3+ colossus 3 base deathball tosss push (youtube)
Spoilered so you dont have to leave this page if you want
+ Show Spoiler +
(Sorry, you will need a pass to watch either game)
So in both of these games, Symbol made ultras in ZvP. We all know that ultra/baneling can rape any Toss army, but the problem is cost efficiency at a stage in the game where cost efficiency starts to become really important for Zerg.
In modern ZvP, Zerg uses roach/ling to defend 2 base aggression, and then goes for mutas/infestors after Toss gets his third. You can't make them earlier or you'll die to the 2 base all-in (you can't even start a spire/infestation pit and 5th+6th gas or you just won't have enough), but ideally you want to start infestation pit/infestors asap against macro Toss. Mass roach style is a lot less popular nowadays as maps and Toss just are too strong against it (but is still viable, of course - in the 2nd set, 1st game listed, he goes mass roach aggression in the midgame).
You need to have broodlords against 3+ colossi from 3 base Toss, when that push comes, but the problem is not the cost, but rather the tech time to do this. At the moment, Toss tends to do this 3+ colossus push when Zerg may have hive done, but the greater spire isn't done, et cetera, so Zerg needs to delay Toss from killing them by using either mutas or mass spine+infestor+corruptor or roach aggression, which leaves an uncomfortable timing that does rely on many factors, such as skill, builds, and the flow of the game, but irregardless, there is a timing when Toss is pushing out and Zerg simply isn't ready. It's very uncomfortable.
Symbol solves this with Ultralisks, something we are actually seeing in the other 2 match-ups as well (vZ, vP) - using Ultralisks as a sort of 'tier 2.5' unit to use when lair tech isn't strong enough, but going broodlords is too vulnerable, so Zerg can safely hold 3 base deathball pushes (150+ Terran rine/tank push when only infestors/mutas are not enough, and mass roach/hydra/infestor would kill you).
In these games, Symbol goes for Hive, and then baneling nest and Ultra cavern. He goes for Hive somewhat quicker than most Zergs seem to, and actually does not make many infestors. Using Ultras and slowbanes, you run over forcefields and kill off the sentries and colossus at a time when Toss will not have HT from 3 base (Toss can't afford HT unless he fought off mutas) and has to push before Zerg has broodlords. He defends the 3+ colossus 3 base push using ultras, and then transitions into mass broodlords as most zergs do, when he's safe because Toss does not have an army.
The 'problem' of cost efficiency is solved because you simply need to make sure that 3 base 3+ colossus/sentry push doesn't hit you, because like ultras are very good at, you kill the Toss army, and you are just using Ultras temporarily (like in the other 2 match-ups) so you can go into Broodlords.
He goes Ultras in 2 very, very different ways in both of the games though, which is extremely interesting and I think warrants this thread since it's clear Symbol knows what he's doing here.
1. Symbol made a lot of roaches in this game on Daybreak, to do roach aggression (with drops, actually). He fails miserably with his roach aggression, and this does put him behind - which makes ultras all the better a choice rather than mass muta/infestor into broodlords after his roach aggression failed.
Parting does his 3 base 3+ colossus/sentry push, and Symbol has mass roach/ling. Symbol goes right into Hive as soon as he knows his aggression isn't working and his drops won't work, which is a lot quicker than most Zergs, who would instead mass infestors and spines or mutas instead.
He lets Parting come all the way across the map uncontested, but instead of the "oh shit just give me 2 minutes for my broodlords, I have the money and the GS is morphing I swear!" moment, he has 5 Ultras popping and a ton of banelings morphed from his lings, with no speed (but speed on the way) or chitinous plating (again, on the way).
Symbol very well could have waited until baneling speed finished, it wasn't far off, and Parting wasn't all up in there yet - but Symbol knew he didn't need to. He used the Ultras to smash the forcefields, and the banelings do a ton of damage and kill off the sentries so his mass roach army could push back the stalkers.
From here, Symbol had 2 choices. In this game, he goes with the choice of "Stay on lower tech, be more aggressive" much like Zergs mass roaches instead of infestors/mutas. He ends the game with Ultra/speedbane, and starts adding infestors. He would have eventually gone for Broodlords, but he knew the game was his already, so he went for more aggression.
2. Symbol gets roach speed, +1, lots of roaches on 3 base vs 2 base Toss, all the normal stuff. But what's different here, is that Symbol really found the answer vs a Toss who takes a fast third - Symbol completely cut roach production, went for infestors, and made lots of lings instead of roaches (you still need an army of course). He got NP, and a quick hive, but did not mass spines like most zergs would here. He made very, very few roaches, he literally did not make any more when he saw Parting take a very fast third on Cloud Kingdom.
Now most of Zergs just hate when Toss does this. Wtf are you supposed to do, and there is still going to be that uncomfortable "just give me a sec bro I have broodlords coming!" no matter how fast you tech or how greedy you play.
Symbol does not make too many infestors, but he did make some, unlike the last game. In this game, he got NP, something he did not do last game. I assume that if you don't make lots of roaches, you need NP. You only mass roaches if you fear aggressoin, or want to do aggression, but in this game Symbol decided not to go for roaches due to the fast third of Parting.
Just like in ZvT and ZvZ quick hive, short lair style into ultra, you can only get Ultras quickly if you did not make many roaches.
So Parting does his 3+ Colossus 3 base deathball push. In this game though, he got run over when he moved out as Symbol had Ultra/speedbane/NP ready, as he teched quickly when he saw the fast third, and got this army out while Broodlord/infestor would never have been ready.
He NP'd the 4 colossus and crushed everything. He made NP useful, which was interesting.
He clearly went for a macro game instead this game, and instead of making more Ultras and speedbanes to follow-up and close the game, he had a spire already morphing when he started ultra production, and a greater spire morphing during the fight.
The game ended before he made any corruptors though, but it was obvious what his intent was.
I imagine there's some vulnerabilities to this - if Toss knows you are cutting roaches in favor of Ultras, or going quick hive Ultras after mass roach instead of mutas/infestor, he would probably make HT and take a fourth and be more passive instead of pushing. But fortunately for Zerg, we can rely on Toss' instinct to all-in and push out with his 3+ colossus 3 base deathball for a long time ^^
Seriously though, I imagine the 'answer' for Toss is to sit back and take a fourth and HT tech instead of push out, and get archons, zealots, and storm before pushing, which shouldn't take too long (relatively), and then push out with an army that can really kill the ultra based army really efficiently. He should probably get mothership during this heavier HT/archon/zealot push.
The biggest vulnerability I really see in this style, is that Toss could take that fourth and go archon/HT/zealot before pushing, and because he's on 4 bases, he could much easier transition into Carrier/HT/Archon at the same time you are making those broodlords. But I think it'll be a long time before we see that in a pro game...
TLDR Just like in recent metagame shifts in ZvZ/ZvT, in ZvP, zerg can cut out mass roach in midgame if Toss goes fast third, or cut out infestors/mutas if you went mass roach, and go ultras before broodlords so you can handle the 3 base deathball push of the opponent, so you can safely hold this deathball push when broodlord/infestor would be uncomfortable and lair tech would not be enough.
In all 3 match-ups, Ultras are amazing to use as a sort of 'tier 2.5' unit to use when lair tech isn't enough, and broodlords is too greedy, so you can safely transition into that ideal BL/Infestor army while giving Zerg another chance to possibly end the game with lower tech aggression.
|
This is a pretty incredibly write-up. Thanks for all the insight. As someone who is just starting to switch to Zerg I can't comment too much on this but Symbol's play was truly incredible as it has been for the past month or so. I love that he's bringing new builds into the Zerg metagame and it seems like he has a much higher understanding of the game than a lot of other players out there in Code S right now. I'm looking forward to seeing Symbol as the next GSL champion.
Using ultras has been long overlooked in ZvP and I'm glad Symbol found a way to not only successfully use them, but dominate with them in a fashion we haven't seen in a long time.
|
Good post, although Ultras still got their flaws in ZvP, the 2.5 thingy maybe the answer to our complaints regarding the role of them. I especially like the synergy with MorroWs/Dimas Ling/Bane - Style, which makes the roach looking like an obstacle in your flow to the later stages of the game (At least in any games without a strong and early 2Base All-In).
|
Not another option for zerg!
Stop this!
It's hard enough as is on ladder in P v Z.
Thanks for adding content to TL though, this looks so strong.
|
United States8476 Posts
The counter for this as toss is extremely passive turtle play. A unit composition consisting of some combination of mainly archon/immortal/colossi/templar is the best. Zealots are pretty much worthless except for reinforcements after all banelings die, and you want as few stalkers as possible, as they don't really contribute much to the fight. Leftover sentries should be used to hallucinate archons. Fortify all your bases with spread out cannons. Basically just turtle like no other. If you reach a maxout with this composition, baneling/ultra cannot touch you. Double upgrades are also really bad versus this style cause all the main units(banelings and ultras) don't really care much about armor.
In the 2nd game, Parting got a huge lead, but was just overly aggressive. In the 3rd game, Parting was just caught off guard. He didn't get to use hallucination or spread out his army. Both games are good demonstrations of this style, but in both games I feel like Parting just wasn't used to facing it enough and made huge blunders.
|
WOW this is EXACTLY the post I was looking for... consider myself a fan of yours, herr doktor Belial. Have been wondering for a long time how to make ultras viable zvp
Edit:
On June 20 2012 15:22 NrGmonk wrote:
The counter for this as toss is extremely passive turtle play. A unit composition consisting of some combination of mainly archon/immortal/colossi/templar is the best. Zealots are pretty much worthless except for reinforcements after all banelings die, and you want as few stalkers as possible, as they don't really contribute much to the fight. Leftover sentries should be used to hallucinate archons. Fortify all your bases with spread out cannons. Basically just turtle like no other. If you reach a maxout with this composition, baneling/ultra cannot touch you. Double upgrades are also really bad versus this style cause all the main units(banelings and ultras) don't really care much about armor.
In the 2nd game, Parting got a huge lead, but was just overly aggressive. In the 3rd game, Parting was just caught off guard. He didn't get to use hallucination or spread out his army. Both games are good demonstrations of this style, but in both games I feel like Parting just wasn't used to facing it enough and made huge blunders.
Aww man, you mean this ISN'T another nearly-unstoppable death strategy? xD I wonder what you would recommend in turn to beat the turtling strategy? Brood lords perhaps?
|
The counter for this as toss is extremely passive turtle play. A unit composition consisting of some combination of mainly archon/immortal/colossi/templar is the best. Zealots are pretty much worthless except for reinforcements after all banelings die, and you want as few stalkers as possible, as they don't really contribute much to the fight. Leftover sentries should be used to hallucinate archons. Fortify all your bases with spread out cannons. Basically just turtle like no other. If you reach a maxout with this composition, baneling/ultra cannot touch you. Double upgrades are also really bad versus this style cause all the main units(banelings and ultras) don't really care much about armor.
In the 2nd game, Parting got a huge lead, but was just overly aggressive. In the 3rd game, Parting was just caught off guard. He didn't get to use hallucination or spread out his army. Both games are good demonstrations of this style, but in both games I feel like Parting just wasn't used to facing it enough and made huge blunders.
The problem with turtling is that you have no 4th. As shown in the 2nd game, against Parting on Cloud Kingdom, if you attempt to play passive, you will be forever denied your fourth while Zerg gets his broodlord/infestor army (now, with ultra support...).
I don't think it's as easy as you make it out to be. It's like you are saying 'mutas are easy' and to which I'd say "no, mutas are not easy, but there is a way you are supposed to fight against them and it's still a challenge"
Upgrades are still useful against the roaches or lings the player will still have massed (in game 1, mass roaches were still present, in both games, mass ling) as well as the eventual broodlord/infestor. Armor is also useful on certain units, bane vs archon and immortal for example.
Aww man, you mean this ISN'T another nearly-unstoppable death strategy? xD I wonder what you would recommend in turn to beat the turtling strategy? Brood lords perhaps?
'Standard' play hasn't been made unviable by any means - I still go mutas in ZvT, and many in the GSL do (with more success statistically than infestors I might add) still. You can still do the old mutas/infestor+mass spine/roach aggression play. On daybreak, Symbol did the classic mass roach aggression and then went ultras (it was a great way to catch up from behind though, as his aggression didn't work and if he went broods he clearly would have died) but on game 2 he very much did something novel.
WOW this is EXACTLY the post I was looking for... consider myself a fan of yours, herr doktor Belial. Have been wondering for a long time how to make ultras viable zvp
Thanks. I wrote a guide a while back about ultras in zvp, but it only worked because I was diamond back then (its still a neat guide imo though, well written - just dont follow it, although it does sound like dimaga did something like that guide, i havent seen what games he did that everyone is talking about). Really interesting how ultras are being introduced into all match-ups at the moment.
Makes me laugh at how back in the day people would use ultras for tech switching (like in zvt vs viking/marauder) such as bl to ultra (to lose). It took 2 years for people to finally start learning ultras (here's hoping blizzard doesn't watch these games and not change ultra in hots). ah, we were so terrible 2 months ago.
|
Nice work belial. I know you Hang out in our channel. Maybe we need a coach like you 
|
Wow, awesome post. I would just like to say I am glad to see ultralisks used (and having a role in ZvP), instead of just BLs all the time. Even if it only works in the current metagame briefly, it is refreshing to see them again, as ZvP seems to have gotten a bit stale with the same BL/infestor/mass-spine vs mothership/deathball
|
United States8476 Posts
On June 20 2012 15:44 Belial88 wrote:Show nested quote +
The counter for this as toss is extremely passive turtle play. A unit composition consisting of some combination of mainly archon/immortal/colossi/templar is the best. Zealots are pretty much worthless except for reinforcements after all banelings die, and you want as few stalkers as possible, as they don't really contribute much to the fight. Leftover sentries should be used to hallucinate archons. Fortify all your bases with spread out cannons. Basically just turtle like no other. If you reach a maxout with this composition, baneling/ultra cannot touch you. Double upgrades are also really bad versus this style cause all the main units(banelings and ultras) don't really care much about armor.
In the 2nd game, Parting got a huge lead, but was just overly aggressive. In the 3rd game, Parting was just caught off guard. He didn't get to use hallucination or spread out his army. Both games are good demonstrations of this style, but in both games I feel like Parting just wasn't used to facing it enough and made huge blunders.
The problem with turtling is that you have no 4th. As shown in the 2nd game, against Parting on Cloud Kingdom, if you attempt to play passive, you will be forever denied your fourth while Zerg gets his broodlord/infestor army (now, with ultra support...). I don't think it's as easy as you make it out to be. It's like you are saying 'mutas are easy' and to which I'd say "no, mutas are not easy, but there is a way you are supposed to fight against them and it's still a challenge" Upgrades are still useful against the roaches or lings the player will still have massed (in game 1, mass roaches were still present, in both games, mass ling) as well as the eventual broodlord/infestor. Armor is also useful on certain units, bane vs archon and immortal for example.
Of course, this strategy isn't easy to beat. It's about as difficult as pretty much every other Zerg strategy. I was just saying what steps Protoss has to take to mitigate the potential damage.
Yes, it is difficult to secure a 4th. But Parting would have had an easy time in game 2 if he didn't suicide his entire army by being too aggressive. And Parting just played badly in game 3.
Also, armor upgrades are completely useless as a vast vast majority of the dps from this composition comes from banelings and ultras, not lings and roaches. You can use forcefields and blink to act as armor versus lings and roaches if you have to. Armor helps archons and immortals versus banelings??? The resources required to upgrade would be much better spent on additional cannons or archons.
|
^ Ah okay. yea i can't tell how well he played, thanks for the insight.
Nice work belial. I know you Hang out in our channel. Maybe we need a coach like you
what channel. thanks.
Thanks Yoshi. ZvP will still be stale, it's just the staleness will happen later now on.
|
Symbol is such a beast lately 
I watched the games live and i was just like "Holy [censored]"
He made Parting looking like a noob.
Ultras are great since release the problem with them is their ridicilous bad path-finding. You need amazing Unit-Control to make them efficient. And i like the idea to use them as a gapcloser to get your BL/Festor out more safe.
PvZ felt a bit random lately,since P was pushing out when his obs scouted that Hive is starting (win/lose)
or pushed when BL were morphing (if you hit the timing win / if you don't like 70% lose becauseyou are in an awful fight position)
|
Does he go for double evo before Lair? And if so, Does he upgrade ranged or melee?
|
Honestly it worked because Parting fucked that up
He had no archons, didnt pressured the 4th, let the creep spread, and above all kept his colossus in the front where they got neuraled/ultrarolled
I dont think it have any chance against a protoss that just know what to do
|
One important part of the succes that had symbol with this strat pointed out by Artosis is that even in late game, Symbol still perfectly hit his inject, which made him able to remax several time on Speedling.
|
Is the ultralisk cavern build time significantly shorter than the GS morphing time? Never realized that ultra cavern was quicker
|
Actually the funniest thing about the second game is:
In the main engagement at the end where he NPed the collossi, he actually wasted like 30 banes into one archon, and it didn't matter. Watch and behold.
|
after watching it over and over, i still think this was mostly due to Parting not having any storms and the mass baneling cleaning the vast majority of the protoss army (with fungal to prevent blink). At that point about any unit that can attack ground should have been able to clean that up, even queens
Sure, ultras were nice to prevent forcefields, but i don't think they did as much damage as they get credit for.
To sum it up, perfect engagement for Symbol as he cought Parting's army off guard (not even in a choke, it was already on open terrain, all clumped up), landed good fungals, banelings roll in for the kill, ultras have easy cleanup time. No storm, no feedback, no air units from protoss whatsoever.
Why hasn't this been done earlier by zergs? Because if Protoss has Storm, does not walk up a ramp into the open blindly and storms the ramp twice when zerg attacks, there are suddenly no banelings left and those ultras do literally nothing at all except for the largest burning animation in game
So this thread should be about the banelings in ZvP in my opinion, and just mentioning ultras as forcefield-braker (which are not even necessary. If I'm not mistaken Dimaga did a ton of baneling stuff in Dreamhack against Protoss with great success, cleaning up armies just with banelings, surpassing force field walls with good flanks, sometimes even use drop play). Plus the obvious weakness to this strategy, Storm and any kind of air play (or perfect forcefields in an attack before ultras are out)
On June 20 2012 16:46 SolidMustard wrote:
Is the ultralisk cavern build time significantly shorter than the GS morphing time? Never realized that ultra cavern was quicker
not really, and if you factor in the upgrade for ultras they need more time to be ready for fight than brood lords.
BUT: if you don't have a spire when you hit hive, and if you don't have corruptors when the greater spire is done morphing, brood lords take a ton longer to prepare for. Thus if you want to get them in time you need to prepare early and invest a lot in them (double spire when starting hive, air upgrades, building corruptors while greater spire morphs). This is a huge investment, and often you die before they are finally ready (often just because you didnt have any of the prerequisites ready and thus delayed it). Whereas going ultras, you place that ultra cavern as soon as you get hive, you don't need any air upgrades, and if you get attacked before you can get them out you did not already waste 70% of your money into useless corruptors, buildings and upgrades. Instead you can just go for a ton of lings or roaches or whatever to hold off the attack.
So on paper, BL would be faster, but it requires huge investments to do so, opening timings for you to die (and if you do not play high risk style you will have to delay stuff as building second spire, double upgades, getting corruptors out, thus your BL come way later than ultras would). Going ultras is much safer that way and in most cases faster
|
I enjoy your posts a lot Belial and next time I reference anything related to this information I will source you Thanks for a great read.
|
ive always liked ultras but honestly..i think the reason why symbol is that good and why he won is his strong mechanics, terran might be the most micro intensive race but with overall great mechanics you benefit most from them as a zerg player...
i mean..symbol sure is a smart player - its impossible to do that well against the top players without beeing - but i never was inspired by his decisionmaking like i was by nestea's. i think his great mechanics will get him very far tho i look forward to watch him play again
|
|
|
|
|
|
EPT
