|
Hello everyone from Team Liquid! I wrote this guide back in early September, and never really got around to finishing it off. So here it is finally!
Introduction
+ Show Spoiler +
I’m Alek Hrycaiko aka coLPoo from complexity Academy. Welcome to my guide about playing the unorthodox PvP style of Dark Templar (DT) Oracle. The goal of this guide is to highlight the strengths of DT Oracle, its weaknesses and also why HOTS PvP allows for such an unusually gas heavy opener.
The build itself is something I’ve developed during the beta, and has been played by me countless times over mid-top NA GM ladder. My winrate with this build currently on ladder is greater than 50%, but of course like all builds in PvP – it has clear BO hard counters and soft counters. I’ll be discussing those later on in the guide.
Build Order:
+ Show Spoiler +
9 pylon
13 gate
14 gas
15 gas (double gas on 15 is fine as well. Just don’t skip any probes.)
17 pylon
18 core
Chronoboost probes.
21 Mothership core (no cb)
23 twilight council ***
23 warpgate (cb)
23 zealot
Chronoboost warpgate again.
25 pylon
Chronoboost warpgate again.
Probe
When tc finishes start DT shrine.
Next 150 gas purchase a stargate.
Start 2nd gateway and a 3rd potentially. Whether or not you will need both depends on your own intuition scouting.
4th pylon either in-base or somewhere well hidden on the map.
5th pylon closer to opponents natural or in their natural base.
From here the build is quite simple.
- Build only 1 dark Templar.
- You should have a remainder of gas for a single oracle.
- 2nd gas expenditure should be on a 2nd oracle unless you’re playing vs a non-dt based all-in. If you’re playing vs an all-in, start building voidrays.
** *Some modifications to the opener are possible. For example, if you revert the order in which you purchase your twilight council and mothership core you can get a faster DT shrine and a mothership core timed well enough to defend any warpgate pressure. For this to work optimally, your opponent cannot be probe scouting you. Otherwise they’ll have knowledge of your tech path and can likely gain a large advantage against you early on.
** Of course, this opener is still possible if your opponent scouts. However to execute this you’ll need to make a zealot after your core finishes (faking a stalker), make a mothership core, a 3rd pylon and then start your tech. Timing your tech from there is relatively simple.
Strengths, Weaknesses and situations you BO win/BO loss.
+ Show Spoiler +
Soft-counters (Builds that do ‘well’ against this under good execution):
(1) When your opponent plays a safe 1g FE: See Naniwa vs Huk Belshir from MLG Anaheim 2013. Any opponent opening 1g fe where they produce 2 stalkers, a mothership core and 2 sentrys should be capable of defending if their stalkers units are camping their main mineral line. Of course, there are ways around this (such as baiting units out of position with a dt), but when executed perfectly the 1g FE should be ahead.
(2) Your opponent plays a heavy 1b sg build and abuses your lack of gas for anti-air. The oracle dt player can play evenly with significant dmg, but generally you’ll be behind.
(3) Robotics/dark Templar builds tend to get your opponent ahead mostly because of the low cost detection by observers. Oracles cost a lot, and you’ll eventually need at least two to maintain detection versus dark Templars. Your expansion also ends up being much later in general in a macro situation vs robo dark Templar. You play from behind versus this build.
Hard counter builds:
(1) 2-3 gate pressure + a robotics facility. Forcing an early nexus cannon and then following up by defending oracle+dt pressure & doing a larger stalker observer counter attack will outright win versus this build. Too much gas is invested into tech and not units to defend versus this.
(2) Properly executed 10 gate 3 gates. Not a widely known transition with this strategy, is to identify a mothership core defense is being abused to perform heavy tech (such as dt) and blindly make a forge + cannon behind this. By then diving in with your stalkers to trade for probes, the 10 gate player should win with proper execution in this scenario.
(3) Delayed blink observer strategies.
Why the mothership core is a necessity.
+ Show Spoiler +
The mothership core prevents fast warpgate timings from being effective.
The general wg timings (taken from some SC2Johns blog post awhile back for convenience) are
10-gate 3-gate: ~5:20 [ 10 gate still counters our build, check BO counter list for why ]
-4-gate: ~5:45 [dt tech should finish during your nexus cannon time. Warp in a dt aggressively to attack opposing mineral line, and remainder of gas for dts to defend ]
-3-4 gate pressure: ~6:00-6:30
If you’re playing a fast mothership core build, either as your 1st unit or immediately after you tech doing the opening listed above – your nexus cannon should be finishing around 5:30. Because of this, you buy yourself an additional 60s of immunity from any timing that is thrown at you. In this time, your tech should be put into play & allow for it to aid in your defense.
So this build has a lot of counters, why do it?
+ Show Spoiler +
There are a number of reasons to incorporate this strategy occasionally:
1) Because its an unexpected transition that has a lot of potential to do large damage and secure you a winning position early on.
For example: Even if the build is soft-countered, (example: 1g,sg) an oracle can still potentially do anywhere from 4-6 probe dmg with only a 2 stalker defense or 1 phx defense. Or perhaps, your opponent is playing aggressive and your dt does minimal dmg, the oracle again has potential to do high amts of dmg killing a higher range of probes 6+. Unexpected transitions into oracle can be profitable. Oracles in general, even on a suicide run, tend to kill a very high number of probes, which can give you an advantage if controlled properly.
2) It's fun to play with oracles and dts.
3) People identify dt, and remove the possibility of oracle from occurring. The oracle then has a 'surprise' opportunity to do damage. [Example: your opponent opens up robo expand. He kills your DT, or identifies its there. Pushes down the ramp to secure his expansion. Suddenly he has a mineral line free to wreck havoc upon.].
4) Fast DT is actually quite strong vs a number of openers, but adding DT/SG removes BO loss from DT only builds.
Saying this, I would not recommend doing this strategy a lot. It’s something nice to mix in with hopes to catching your opponent off guard. But like all things in PvP, it can be countered.
Mid game transition:
+ Show Spoiler +
I haven't really mentioned anything about a mid-game transition b/c it does vary heavily game to game. You should generally try to secure a very quick fast expansion behind this strategy. Depending on your economy, you should expect to follow up with either blink stalkers or a chargelot archon based composition. Having a good game sense is very key to securing yourself with a solid mid-game.
Other stuff:
+ Show Spoiler +If you have specific questions about the opener just post in the thread and I’ll keep an eye on things + respond when I have time. Currently I’m quite busy in school + midterms and fitting in practice time; so don’t expect a fast reply.
I hate to really write about str vs builds b/c I feel like there are too many variables to account. But, use this as a general rule to whether this build will be a good option:
As a general rule, if a build either has later detection, a lack of detection or favors having units aggressively put on the map - this build will generally excel vs all those scenarios.
Replays:
+ Show Spoiler +Pardon that these replays are not super recent, they’re from July-Sept. I wrote this guide over a month ago and have been smurfing, so I can’t really release current replays. If anyone has good examples of shortcomings or execution of this build vs specific strategies, please post and I’ll update. Three replays in this pack cover different strategies executed vs dt oracle. 1 Successful execution, 1 BO Counter, 1 BO soft counter are covered in the replay pack. All replays are either me punishing dt oracle vs an opponent on ladder performing the strategy. http://www.mediafire.com/?89p25dumz72n14y
Plugs:
www.twitch.tv/stealthypoo
www.twitter.com/604poo
|
This build has blown my mind. OH EM HASUS
|
pooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
|
Because that is not asking for a ban, even though what possessed you to name yourself Poo?
You self admittedly say this build has a lot of counters and then you go on to say it's a coinflip build, you didn't really sell your build very well. Surely if your opponent made one stargate they win?
|
On October 14 2013 05:48 Ephemera wrote:
Because that is not asking for a ban, even though what possessed you to name yourself Poo?
You self admittedly say this build has a lot of counters and then you go on to say it's a coinflip build, you didn't really sell your build very well. Surely if your opponent made one stargate they win?
Every build has a list of soft/hard counters in PvP, regardless of what build it is (basically meaning all builds in PvP are flipping a weighted coin).
This build can cope with Stargate play, but under the condition you're able to accomplish dmg with your oracles/dt. If you accomplish no dmg, of course you'll be behind. The guide isn't about trying to sell my build as a super-safe amazing strategy that will get you to GM, its about outlying the principles to execute the build and understanding where it falls short + where it excels. Understanding the situations in which a build doesn't work in PvP is really essential to having a grasp on the matchup.
Also: If I didn't list a strategy under the BO counter list, likely this dt oracle build excels or can be played evenly against it.If you think about it, that's actually quite a large range of builds this is viable versus, sorry If I didn't make that more clear. Of course, I'm open to more builds being added to that list. I just added the most obvious ones I've commonly come across.
|
As someone who has only done this build until getting to GM (experimenting with robo), yeah, it will get you GM. I had over 70% win in p vs p, with hundreds of games in the mu, until I started insta leaving a lot. Once you start playing people that know what you're doing, you get hard countered by blink + robo builds, so it becomes a challenge.
DT + robo sucks because there's no real reliable way to catch up in probes/eco vs 1 gate expand. With this, though, the percentages are much better. I must have missed it, but it doesn't even look like you talk about what the build is strong against. Not even a mention of 4 gate. Luckily it's your guide. I wouldn't want people to know how to do it, either.
|
United States4883 Posts
This build as a whole makes no sense. It's simply a DT rush with a risky, blind stargate added to the tail end of it. Of course, the DT play does well against a variety of builds including all gateway pressure and blink builds, but you add on the unnecessary risk of skipping two stalkers for a surprise transition that may or may not be good (if, for instance, your opponent does 1-base stargate or a very normal FE), making you much more vulnerable to a lot of things. Like, what if your opponent just does a 3-stalker poke with a normal robo expand behind it? You just lose a bunch of probes for literally nothing.
If you want to use a stargate to exploit certain builds, you should just do a normal DT expand with a robo, and then upon scouting your opponent's tech with the DT, cancel the robo and build a stargate. For example, if your DT walks in and things start shooting it, you know your opponent opened robo and you can reactively go stargate. Doing this type of cancel cuts down on all the stupid coinflips and makes your build a ton more solid.
I actually do have an oracle/DT build that I use, but I get it in reverse order. You open with a proxy stargate (on second pylon) and build an oracle. When the oracle flies in (~5:20), you can get a good idea of what your opponent is planning and decide if you need to cancel your twilight council to defend pressure. Out of the stargate, you go oracle -> 2 phoenixes, then followup with the DTs. If they went for a robo expand, you can activate the oracle's envision ability, snipe the obs, the run in with the DTs. It's a funny little metagame build, but a lot safer in general. the first oracle is generally almost guaranteed 3-4 probe kills with correct micro, while the followup DT pressure can net you even more kills.
On October 14 2013 06:22 playa wrote:
As someone who has only done this build until getting to GM (experimenting with robo), yeah, it will get you GM. I had over 70% win in p vs p, with hundreds of games in the mu, until I started insta leaving a lot. Once you start playing people that know what you're doing, you get hard countered by blink + robo builds, so it becomes a challenge.
DT + robo sucks because there's no real reliable way to catch up in probes/eco vs 1 gate expand. With this, though, the percentages are much better. I must have missed it, but it doesn't even look like you talk about what the build is strong against. Not even a mention of 4 gate. Luckily it's your guide. I wouldn't want people to know how to do it, either.
http://www.teamliquid.net/forum/viewmessage.php?topic_id=431203
DT/immortal so good. You just need to pressure and force stalkers and stuff.
|
On October 14 2013 06:22 playa wrote:
As someone who has only done this build until getting to GM (experimenting with robo), yeah, it will get you GM. I had over 70% win in p vs p, with hundreds of games in the mu, until I started insta leaving a lot. Once you start playing people that know what you're doing, you get hard countered by blink + robo builds, so it becomes a challenge.
DT + robo sucks because there's no real reliable way to catch up in probes/eco vs 1 gate expand. With this, though, the percentages are much better. I must have missed it, but it doesn't even look like you talk about what the build is strong against. Not even a mention of 4 gate. Luckily it's your guide. I wouldn't want people to know how to do it, either.
I mentioned briefly about m-core being defense vs 4gate, probably should edit that dt tech will finish after that.
I'm going to let people fill in the holes where they think its 'strong' vs certain strategies. I hate to make claims about strength vs strategies b/c I feel like there are too many variables in terms of builds/unit placement that can really effect the outcome of this build.
But, I do think its probably fair to edit in a general rule in the op. I'll make it something like the following:
As a general rule, if a build either has later detection, a lack of detection or favors having units aggressively put on the map - this build will generally excel vs all those scenarios.
|
On October 14 2013 06:29 SC2John wrote:
This build as a whole makes no sense. It's simply a DT rush with a risky, blind stargate added to the tail end of it. Of course, the DT play does well against a variety of builds including all gateway pressure and blink builds, but you add on the unnecessary risk of skipping two stalkers for a surprise transition that may or may not be good (if, for instance, your opponent does 1-base stargate or a very normal FE), making you much more vulnerable to a lot of things. Like, what if your opponent just does a 3-stalker poke with a normal robo expand behind it? You just lose a bunch of probes for literally nothing.
If you want to use a stargate to exploit certain builds, you should just do a normal DT expand with a robo, and then upon scouting your opponent's tech with the DT, cancel the robo and build a stargate. For example, if your DT walks in and things start shooting it, you know your opponent opened robo and you can reactively go stargate. Doing this type of cancel cuts down on all the stupid coinflips and makes your build a ton more solid.
I actually do have an oracle/DT build that I use, but I get it in reverse order. You open with a proxy stargate (on second pylon) and build an oracle. When the oracle flies in (~5:20), you can get a good idea of what your opponent is planning and decide if you need to cancel your twilight council to defend pressure. Out of the stargate, you go oracle -> 2 phoenixes, then followup with the DTs. If they went for a robo expand, you can activate the oracle's envision ability, snipe the obs, the run in with the DTs. It's a funny little metagame build, but a lot safer in general. the first oracle is generally almost guaranteed 3-4 probe kills with correct micro, while the followup DT pressure can net you even more kills.
I'd actually argue that doing DT Robo cancels into a DT SG is significantly worse than the build I'm describing. The longer you wait to produce oracles, the less chance you're going to have to remain unscouted and have a mineral line to wreck havoc upon. Dt robo itself, is a very strong build as well but I think its worse in different situations than DT Sg build. As playa mentioned, you're likely going to have a higher % chance of punishing FEs. Also, the idea that the stargate is 'random' is part of the purpose behind this strategy, it's not a typical follow up and it requires a very specific response to prevent damage from occurring. This strategy will never be a 'standard' strategy, but just an option that's viable to mix in to catch opponents off guard.
You're correct to describe that a 3 stalker robo 2g or 3g will do well versus dt oracle. But why are we playing the 'what ifs game', if you name me any build I can also describe a specific scenario where you'd be behind after. I've also outlined in the guide that the dtoracle build is weak vs this.
A similar DT/Oracle build you're describing was used at MLG Anaheim actually by sage with 2nd pylon proxy. I actually like that build as well, I do it a bit differently but I wouldn't argue that its any safer than the strategy I've listed.
|
vouching this build
this guy has a lot of really amazing builds and tactics
one of NAs finest for sure
|
I think naniwa or grubby did a similar build against duckdeok in premier EU season 2 top8, losing to the 3g robo build.
This build is such a gigantic curveball. Of course it can lose to a perfect response from most builds, but I'd say not many players know what that response is. It's quite easy to misread it or have units out of position.
Thanks for the guide.
|
On October 14 2013 03:14 Poo wrote:
Strengths, Weaknesses and situations you BO win/BO loss.
Soft-counters (Builds that do ‘well’ against this under good execution):
(1) When your opponent plays a safe 1g FE: See Naniwa vs Huk Belshir from MLG Anaheim 2013. Any opponent opening 1g fe where they produce 2 stalkers, a mothership core and 2 sentrys should be capable of defending if their stalkers units are camping their main mineral line. Of course, there are ways around this (such as baiting units out of position with a dt), but when executed perfectly the 1g FE should be ahead.
(2) Your opponent plays a heavy 1b sg build and abuses your lack of gas for anti-air. The oracle dt player can play evenly with significant dmg, but generally you’ll be behind.
(3) Robotics/dark Templar builds tend to get your opponent ahead mostly because of the low cost detection by observers. Oracles cost a lot, and you’ll eventually need at least two to maintain detection versus dark Templars. Your expansion also ends up being much later in general in a macro situation vs robo dark Templar. You play from behind versus this build.
Hard counter builds:
(1) 2-3 gate pressure + a robotics facility. Forcing an early nexus cannon and then following up by defending oracle+dt pressure & doing a larger stalker observer counter attack will outright win versus this build. Too much gas is invested into tech and not units to defend versus this.
(2) Properly executed 10 gate 3 gates. Not a widely known transition with this strategy, is to identify a mothership core defense is being abused to perform heavy tech (such as dt) and blindly make a forge + cannon behind this. By then diving in with your stalkers to trade for probes, the 10 gate player should win with proper execution in this scenario.
(3) Delayed blink observer strategies.
So basically you are saying I can use
1g FE, 2g FE, Stargate, Robo, Blink, DTs or Scouting to end up ahead?
Sounds like a good strategy!
|
One of the main concerns when doing the build is my lack of confidence in air play. Example: you scout 2 phoenixes or w/e, before you have an oracle out, so obviously you don't make an oracle. Do you try to catch up in phoenixes? I'd like to make 2 void rays, but if they keep making phoenixes, then it's a waste of money it seems.
So I guess the question is do you play it phoenix vs phoenix, and how does that go for you, or what do you try to do?
|
On October 15 2013 05:48 playa wrote:
One of the main concerns when doing the build is my lack of confidence in air play. Example: you scout 2 phoenixes or w/e, before you have an oracle out, so obviously you don't make an oracle. Do you try to catch up in phoenixes? I'd like to make 2 void rays, but if they keep making phoenixes, then it's a waste of money it seems.
So I guess the question is do you play it phoenix vs phoenix, and how does that go for you, or what do you try to do?
I never attempt to start a phx vs phx war when you're choosing SG as a 2ndary later tech. Skipping an oracle is fine, but I think b/c you're committing to the tech, its preferable you make an oracle and control it away from phx line of sight and try to ninja it into a mineral line when possible. Ofc, if you've already done enough dmg with dts, then maybe this isn't as necessary.
A couple transitions from here:
Expand into a later blink + forge after you get about 4-5 stalkers a mineral line and then play a ground vs air style opener. Note that you really do need to rely on nexus cannons earlier on to minimize dmg b/c your gas count + army value will be lower from double teching.
A second option, that's a possibility only if you're economically ahead, is to fake commitment to the stargate by producing a # of phx, and then swapping off into a 2b blink archon timing attack (this hopes to catch them attempting a phx mirror style). If you're not ahead, don't attempt this.
I find both of those transitions to be pretty effective vs SG play. Personally I don't think I've ever really toyed with making voidrays vs a person with a phx lead. It seems to me that voidrays would be easier to cope with for the phx player than positioning a # of stalkers, later a cannon and using well timed nexus cannons. Early voidrays really aren't that good in my opinion unless they're guaranteed to be preserved and can lead into some sort of timing attack (or they're useful if you really need units to defend an all-in mb).
|
On October 15 2013 06:02 Poo wrote:Show nested quote +On October 15 2013 05:48 playa wrote:
One of the main concerns when doing the build is my lack of confidence in air play. Example: you scout 2 phoenixes or w/e, before you have an oracle out, so obviously you don't make an oracle. Do you try to catch up in phoenixes? I'd like to make 2 void rays, but if they keep making phoenixes, then it's a waste of money it seems.
So I guess the question is do you play it phoenix vs phoenix, and how does that go for you, or what do you try to do? I never attempt to start a phx vs phx war when you're choosing SG as a 2ndary later tech. Skipping an oracle is fine, but I think b/c you're committing to the tech, its preferable you make an oracle and control it away from phx line of sight and try to ninja it into a mineral line when possible. Ofc, if you've already done enough dmg with dts, then maybe this isn't as necessary. A couple transitions from here: Expand into a later blink + forge after you get about 4-5 stalkers a mineral line and then play a ground vs air style opener. Note that you really do need to rely on nexus cannons earlier on to minimize dmg b/c your gas count + army value will be lower from double teching. A second option, that's a possibility only if you're economically ahead, is to fake commitment to the stargate by producing a # of phx, and then swapping off into a 2b blink archon timing attack (this hopes to catch them attempting a phx mirror style). If you're not ahead, don't attempt this. I find both of those transitions to be pretty effective vs SG play. Personally I don't think I've ever really toyed with making voidrays vs a person with a phx lead. It seems to me that voidrays would be easier to cope with for the phx player than positioning a # of stalkers, later a cannon and using well timed nexus cannons. Early voidrays really aren't that good in my opinion unless they're guaranteed to be preserved and can lead into some sort of timing attack (or they're useful if you really need units to defend an all-in mb).
Oh, this is embarrassing. I've actually never used blink in p vs p. No confidence there, either. I could try that. But, honestly, anytime you have to make a forge that early and cannons, it feels like you're giving your advantage away. Especially, if they're still able to pick off probes here and there. It just really feels like you're supposed to make phoenix, too, and play defensively with them, until you catch them off guard/end up with a higher count than them.
It just seems like if someone opens phoenix then they should be expecting blink stalkers and feel pretty confident that the phoenix/immortal combo will prevail. I'd have to see some reps sometime. Kudos to you for making it work.
|
|
|
United States4883 Posts
Lol this build was just used in Challenger League (WCS EU, Adominus vs. BabyKnight). BabyKnight scouts the stargate and goes herp de derp blink stalkers, only to lose instantly to the DT followup lolol.
|
On October 22 2013 01:16 SC2John wrote:
Lol this build was just used in Challenger League (WCS EU, Adominus vs. BabyKnight). BabyKnight scouts the stargate and goes herp de derp blink stalkers, only to lose instantly to the DT followup lolol.
you have excellent deductive abilities.
|
|
|
|
|
|
EPT
