|
On May 21 2011 11:51 Pro]ChoSen- wrote:
I've seen some Grand Master players going 4 gate ---> expand.
If the Z plays greedy / drones hard, then gg Protoss wins instantly. If Z is being aggressive you will be able to cope with it pretty easily. If Z is being mildly aggressive / droning behind it then you will probably out-macro him and be able to pressure him back while you expand. If Z plays super defensive with like 4 spines or whatever then oh well take ur expansion and the money you spent on 4 gating he wasted on spines.
How is 4 gate expand a bad choice ^^
Do you mean defensive 4gate? If I am not mistaken you are talking about a 4gate attack. And if you don't kill him right away, you are WAY behind, cuz you cut probes and don't have money for expo. Can you give some more info on this?
|
I tested some different builds. One of them was to add 2 Gates instead of forge / gate and cut probes for units after nexus is finished. It worked really well because you can defend and win against an all in by having unit advantage or you can push early to do dmg and force units (pretty strong timing). It's kind of the same play compared to ling/roach all in.
|
On May 21 2011 20:19 FeyFey wrote:
mmm the allin abuses the larger space at the expansions, and the fact that sentrys need alot of energy and can't fight on their own. Sound for me like a don't use sentrys ^^. I am not a grandmaster toss though, mostly switch races around to test stuffies and have fun. So the zergs i play aren't top level and might react poorly to different situations. (well from my experience almost everyone reacts poorly to strange strats, never played someone above low grandmaster though).
Well that is probably the reason why they play this allin even if they don't see the mass sentrys.
I generally go for fast immortals and +1 ground attack (I chrono that over the warp gate).
A zerg can run by easily against this army composition, but attack it directly with ling roach, not a chance (its heavy micro for the toss) (i use a few sentrys too, they are a perfect extra buff if you have a bit overgas). Just need to scout the zerg pretty well so you won't be suprised by mutas. (luckily an almost forgotten tech in tvz).
That is terrible advice my friend. Fast immos against this build is a free win for Zerg. How many immos are you going to have at 8 mins? Your composition will simply be overrun.
|
Can we have some of the pro's comments editted into the OP please?
I am seriously struggling against this type of play by Z and would be much appreciative if we could filter the thread and update the OP.
I'm at my wits end here.....TT
Thanks
Edit:
If it helps, I'm finding the trouble to be that even if I hold the first two waves or so, there seems to be no viable way to hold the waves that follow, the units that can be replaced quickly from warp gates seem to be too weak to hold once sentry energy runs lows and sentries feel to expensive to replace, not to mention they can't win the fight solo, uggggh this is driving me bananas.
|
|
|
2 gate FE with cannons is my play against this, using a relatively quick hallucination to scout how many cannons I need.
Builing more gates is only counterproductive imo as you don't really have gas to support much more gas units anyway, early on you cant make more then 2 gates worth of sentries and when he pushes I just switch to stalkers, chronoing the gates non stop and making cannons.
Except a DT build there is no magical way to autobeat this build imo. It does quite good against any other aggresive build because it gets quite an amount of lings.
Trick to beating it is just to FF him out while your cannons finish imo, if you have 2 up your are quite safe.
|
This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks.
Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark.
I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build.
Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber.
http://drop.sc/11053
|
On May 21 2011 23:18 FreddieY wrote:This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks. Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark. I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build. Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber. http://drop.sc/11053
A lot of the VT guys, and some of the other pros I've been practicing with regularly are opting for hallucinate before warpgate.
This lets you see if Z cuts drones and starts saving larva so that you can opt for well-informed cannons, or whatever other changes you might like to make to your unit composition.
Honestly, I think its the best option you have...
|
On May 21 2011 23:43 MrBitter wrote:Show nested quote +On May 21 2011 23:18 FreddieY wrote:This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks. Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark. I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build. Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber. http://drop.sc/11053 A lot of the VT guys, and some of the other pros I've been practicing with regularly are opting for hallucinate before warpgate. This lets you see if Z cuts drones and starts saving larva so that you can opt for well-informed cannons, or whatever other changes you might like to make to your unit composition. Honestly, I think its the best option you have...
I just can't imagine this being safe... Can you really afford to use 2 force fields that early in the game? I just don't see how you can possibly get your nexus down against some standard ling pressure when you are delaying warp gate for so long and using up so much energy.
Also, this is far less of an issue, but if you play against some CatZ style proxy hatch roach/ling shenanigans I think you're completely dead if you wasted 100 gas on hallucinate and your warp gate research is that delayed.
|
On May 21 2011 23:47 PJA wrote:Show nested quote +On May 21 2011 23:43 MrBitter wrote:On May 21 2011 23:18 FreddieY wrote:This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks. Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark. I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build. Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber. http://drop.sc/11053 A lot of the VT guys, and some of the other pros I've been practicing with regularly are opting for hallucinate before warpgate. This lets you see if Z cuts drones and starts saving larva so that you can opt for well-informed cannons, or whatever other changes you might like to make to your unit composition. Honestly, I think its the best option you have... I just can't imagine this being safe... Can you really afford to use 2 force fields that early in the game? I just don't see how you can possibly get your nexus down against some standard ling pressure when you are delaying warp gate for so long and using up so much energy. Also, this is far less of an issue, but if you play against some CatZ style proxy hatch roach/ling shenanigans I think you're completely dead if you wasted 100 gas on hallucinate and your warp gate research is that delayed.
I think that's looking at it a little unfairly...
There is obviously a trade-off when you go hallucinate first, but sentries did get a gateway build time reduction...
And when the trade-off is
"If he sacrifices eco by making some extra lings, I take my natural 1 production cycle later."
I think it makes a lot of sense.
Also, look at your alternatives:
Blindly expand at the normal time and risk losing outright, or slow down your first warp-in, while still producing off of gateways so that you can know what's happening.
I can understand it being "not ideal", but I don't think it's a bad option at all.
edit: And with proper timing, you should really only need a single hallucinate. I couldn't say exactly when, but I would think that at around 6:30 you'd be able to see whether Z is saving larva or not.
|
On May 21 2011 23:56 MrBitter wrote:Show nested quote +On May 21 2011 23:47 PJA wrote:On May 21 2011 23:43 MrBitter wrote:On May 21 2011 23:18 FreddieY wrote:This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks. Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark. I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build. Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber. http://drop.sc/11053 A lot of the VT guys, and some of the other pros I've been practicing with regularly are opting for hallucinate before warpgate. This lets you see if Z cuts drones and starts saving larva so that you can opt for well-informed cannons, or whatever other changes you might like to make to your unit composition. Honestly, I think its the best option you have... I just can't imagine this being safe... Can you really afford to use 2 force fields that early in the game? I just don't see how you can possibly get your nexus down against some standard ling pressure when you are delaying warp gate for so long and using up so much energy. Also, this is far less of an issue, but if you play against some CatZ style proxy hatch roach/ling shenanigans I think you're completely dead if you wasted 100 gas on hallucinate and your warp gate research is that delayed. I think that's looking at it a little unfairly... There is obviously a trade-off when you go hallucinate first, but sentries did get a gateway build time reduction... And when the trade-off is "If he sacrifices eco by making some extra lings, I take my natural 1 production cycle later." I think it makes a lot of sense.
Well, I have not tested it myself, so in that sense I am being a bit unfair.
But my experience leads me to believe that spending 100/100 so early is not cheap, and also that you will barely have 100 energy at that point. So basically if they do any really early aggression after you've hallucinated a phoenix, I would think that you'll have a pretty hard time. I also think delaying warp gate by 60 seconds is going to delay your expansion by a lot more than any hit the zerg is taking by producing a few lings early.
I have also played against a number of non-standard strategies from CatZ, Destiny and Slush where I think delaying warp gate is nearly an auto-loss. For example, if CatZ does a proxy hatch in your main into roaches on close spawns, you're dead if you spent 100/100 on hallucinate that early. Also, if he does early lings into taking the gold base, you won't have warpgate research done early enough to punish before the econ advantage kicks in.
Also, look at your alternatives:
Blindly expand at the normal time and risk losing outright, or slow down your first warp-in, while still producing off of gateways so that you can know what's happening.
I can understand it being "not ideal", but I don't think it's a bad option at all.
edit: And with proper timing, you should really only need a single hallucinate. I couldn't say exactly when, but I would think that at around 6:30 you'd be able to see whether Z is saving larva or not.
Those aren't your alternatives, though. You can expand at the normal time, get a cannon, get hallucinate at the more standard time, and then add 2 cannons if they are going losira build. I think this is how aLeJ is currently playing PvZ, and is having success with it.
You can also do any build other than 3 gate expand and be safe against losira. 2 Gate forge expand into fast void ray holds very well, as does 3 gate DT expand.
|
Also to supplement what mrbitter has just said, I could be wrong about this because I'm not a top level player but if a zerg is making a significant enough amount of lings to make you need to cancel your nexus and wait for a few more sentries, this has caused his drone count to be far too low to do a roach ling all-in anywhere near nestea's timing, as he will have to produce up to the necessary drone count before he could pump the 7 or 8 roaches.
On top of that, the limiting factor in nestea's roach ling timing is how long it takes the roaches to walk to your base, as the build involves a very quick roach warren that finishes just about at 6:30 like mrbitter said. following the roach warren completion is a huge roach ling pump, with the initial lings to bait FF's while trying to not lose any, followed by copious amounts of lings as the lings quickly catch up with the slow unupgraded roaches. So to cut into his drone count so heavily and so early for the initial 8 or 10 lings that you refer to as being a threat to expanding, plus to have to wait again to pool larva and then for the roaches to make their way across the map, I think a massive roach ling follow-up would be the least of your problems.
|
On May 22 2011 00:21 kyarisan wrote:
Also to supplement what mrbitter has just said, I could be wrong about this because I'm not a top level player but if a zerg is making a significant enough amount of lings to make you need to cancel your nexus and wait for a few more sentries, this has caused his drone count to be far too low to do a roach ling all-in anywhere near nestea's timing, as he will have to produce up to the necessary drone count before he could pump the 7 or 8 roaches.
On top of that, the limiting factor in nestea's roach ling timing is how long it takes the roaches to walk to your base, as the build involves a very quick roach warren that finishes just about at 6:30 like mrbitter said. following the roach warren completion is a huge roach ling pump, with the initial lings to bait FF's while trying to not lose any, followed by copious amounts of lings as the lings quickly catch up with the slow unupgraded roaches. So to cut into his drone count so heavily and so early for the initial 8 or 10 lings that you refer to as being a threat to expanding, plus to have to wait again to pool larva and then for the roaches to make their way across the map, I think a massive roach ling follow-up would be the least of your problems.
Yes, if they pressure with 8 to 10 lings or something they won't be going a massive roach/ling follow-up.
But that's not the point. If you are getting hallucinate quickly to scout for a roach/ling all-in, then sure, you'll be fine if they do losira build. But what if they do some other standard build, and delay your expansion by an extra 30 seconds to a minute because you researched hallucinate first. Then you are behind.
It's like if you played some opening move in chess and said, "Well, if my opponent continues to play the king's indian after this move they will be behind a tempo," while completely ignoring the fact that they can transpose into some other variation where you are strictly worse than the main line.
|
Why don't protoss for defense, hallucinate zealots more? IMO they are one of the BEST units to hallucinate because the AI loves to target them. If you make hallucinated stalkers/voids/phoenix/whatever for tanking damage...you really probably won't lower enemy DPS like 2 fake zealots would...no other unit gives you a better chance of absorbing more shots from your real stalkers/sentries/cannons. Plus you get two of them unlike say the immortal or archon.
In fact when doing a gateway push, hallucinated zealots could be so key to tanking damage from spines/lings/drones.
|
If you can pick the terrain and have better quality units, forcefields have way more stamina. I'll agree that hallucinated tanking units can often be good when things get messy or when you're attacking.
|
Perhaps...I just think forcefields are kind of overrated. Spines ignore them...roaches/hydras can actually fire over them for the most part. If you have zealots, you pretty much can't use forcefields without helping your opponent. Vs lings it can be tough to make waterproof walls vs them. Plus they can fake attack, let you expend your energy, then attack for real. That or they are so fast they get inside your forcefields which obviously is never a good thing. Even if you have that perfect army split, if their backline hugs the forcefields chances are good you won't have the energy for a 2nd wall and all they have to do it to be relentless and press for victory. I see too many time zerg players give up after they army had been split by just a few sentries and gateway units, when if they had kept on pressing they would have easily won.
|
I've seen some of the Chinese players like XiaoT going for 1 gas 3 gate expand. Pumping out only Zealots and Stalkers. I feel that this is a very good mix to hold the losira push. Zealots for Lings and Stalkers for Roaches. Tos players are getting to used to the 3 Gate Sentries openings.
Pre Losira push, Protoss players are warping 6-7 Sentries in to save up mana and holding Zergling pressure. However with Roaches currently in the mix, Sentries needs too much proper positioning. Zealots and Stalkers will probably hold the push fine.
With a standard 3 Gate expo, often times you will be lacking in minerals to get probes, cannons, Stalkers/Zealots. I think getting 1 gas is a good choice in the current meta.
The upside is the build looks exactly like a 4 gate. The downside is very low Sentry mana for Force Fields, which may affect the mid game. Who cares about the mid game if you cant even pass the 7:30 mark.
|
Yeah...to cove, I'm not a pro, but have noticed in watching a lot of pro games an inverse correlation in PvZ games between the number of sentries produced and the chances protoss has in winning the game. I see mass sentry and think...protoss is probably going to lose...and it almost always happens! Sentries are just SO expensive. So with 6 sentries that cost you 300m/600g. Even with this number you will probably only get one good army split on say that roach army but this won't make the sentries pay for themselves. By using 600gas that many sentries really delays your tech (they're aren't many ways of winning vs zerg without fast and early colossi with attack upgrades). Plus it takes up valuable warpgate capacity that could have been used on attacking units to dissuade the zerg from expanding.
Only ways to beat zerg are to 4 gate them, do a phoenix/void rush, do a DT rush, do a 2 base timing rush, do a early blink stalker push or macro up a colossi ball with supporting units. Late/late game you are probably dead unless you get stalker/colossi + storm to cut through the defensive upgrades and bulking caused by max zerg armies. Very tricky to do...
|
On May 21 2011 23:43 MrBitter wrote:Show nested quote +On May 21 2011 23:18 FreddieY wrote:This replay is an example of some of the various roach timing attacks that are possible. However, I and probably along with many others are looking for a safe build that stops a perfect copy of the nestea/losira timing attacks. Particularly, i am trying to figure out a build that stops 9 roaches and 30 lings that hit around the 7:50 mark. I hate to be rude, but some of the replays that are being shown are just not that same build. I see roach builds that are hitting at the 8:30 mark with like 5 lings, then the protoss says that "I stopped it easily with perfect forcefields." Just take a look at your build and see if it is even near capable of stopping 9 roaches and 30 lings at that 7:50 mark. That is what we are looking for. I posted a possible solution with a 3 gate expand while taking probes off of one gas, making it possible to add 2 cannons and a forge along with a quick expand. There are downsides, but I was hoping for some discussion and help to improve on the build. Here is a somewhat close example of what we are trying to stop. Sheth in this replay does a roach ling timing, but still it doesn't quite match that of nestea caliber. http://drop.sc/11053 A lot of the VT guys, and some of the other pros I've been practicing with regularly are opting for hallucinate before warpgate. This lets you see if Z cuts drones and starts saving larva so that you can opt for well-informed cannons, or whatever other changes you might like to make to your unit composition. Honestly, I think its the best option you have...
Why bother to go hallucination first though, you can easily go WG and then hallucination and you can just scout in time. You only need the scout to see how many cannons you need so there really is no need to scout THAT fast. The reduction in sentry build time just means you can chrono warpgate more instead of sentries and then afterwards chrono hallucination it still gets out quite fast, ie about 6:30.
As long as hallucination finishes before the nexus finishes you are fine really, any attack that comes before that can simply be dealt with by canceling the nexus and going back up the ramp, there isn't anything that can break you then..
Hallu before warpgate throws away ANY offensive options (thus if zerg see's the super fast fake phoenix they can just take a third...). Super fast hallu also has an awkward early cost, ie you need to have the cybercore or gateway idle for some time and finally super fast hallu sucks in case they are doing some other form of pressure, like super fast 3 roach allin or just some speedling harass.
|
so i've been looking into this alot and i've found a couple things that work. The main thing that makes them work is that you can't base your attacks on your expo at all (standard play is that you only want to attack as you start your nexus, or after u benefit from it to be more optimally, but this restrains you A LOT in the early game). so here are some ideals:
1) 3gate expand, while getting +1 attack before you expand. still get sentries n zealots with some stalkers, and then push out after +1 finishes. if roaches are present, just fall back, make stalkers, get blink, get +2, and push when that all finishes (grab a 3rd here if you can).
2) 1 base pheonix voidray. this build isn't all in. you can do a lot of damage with it, and expand with around 4 cannons once you (if you) scout a lack of roaches/lings. keep you cannon count up with the zergs potential army while making sentries. but you do have to get a lot out of your pheonixs.
|
|
|
|
|
|
EPT
