EPT
Background:
Recently, Stephano's 11-12 minute roach max has been the central focus of PvZ and every protoss build has to be built around how to deal with this style. 2 of the best zergs in the world, Stephano and DRG, almost exclusive employ one form or another of this build and it is largely because of this standardized build that they hold much more ZvP success than any other Zerg in the world. They are also only 2 zergs who say they are confident in ZvP in interviews, with Stephano even going to claim that ZvP is imbalanced for zerg and that the only way for Protoss to win is to be both tricky and lucky. Even Nestea, the 3rd contender of the best zerg uses this style heavily. The build is basically 2 parts:
- Have a solid 3 base macro build that can react to and defend any protoss all-in and come ahead. Every player's macro build is slightly different, but they're all similar and achieve the same infrastructure. This is not "Stephano Style". It is just standard 3 base macro ZvP.
- Max on Roach heavy/ling light between 11 and 12 minutes with 4 hatches and 4 gas. Proceed to overrun any attempts by the Protoss to take a 3rd. The timing of the max largely depends on pressure, but for reference, a lower level masters player can max by 12:00 without much pressure while Stephano's record is a 10:50 max. This part is the true "Stephano Style Roach Max", not any build that gets 5 or 6 gas or a spire in the middle of maxing.
Theory:
Similarly to the PvZ situation now, until about a year ago, Terrans too were relying on 2 base timings to deal with zerg. As zergs began learning how to deal with all of the various and numerous timings, Terrans' success with those timings began to drop. However, eventually Terrans learned how to play "greedy" but safe macro builds that allow the Terran to keep up with econ. Two examples of these builds include:
- Reactor hellion->cc->cc->Make 4 hellions->tanks
- 1 rax cc->double gas->reactor hellions->banshee->cc
PvZ, on the other hand, has been lagging behind the TvZ metagame for quite a bit as 2 base all-ins are still by far the most common strategy. Now, I could blame this on a variety of things such as:
- The inflexibility of FFE and how Zerg can get a free 3rd.
- The fact that Protoss units work better in balls and thus are less effective non all-in pressure. For example, marines, hellions and medivacs work much better individually than sentires/stalkers/immortals.
- Nexii can't float like CCs.
The problem with current macro PvZ isn't that stalker/colossi can't handle Zerg as well as marine/tank. It's more that Protoss doesn't currently have well-accepted standard solid macro builds that allow them to get enough units to defend a Zerg max. The way Protoss are playing it now, they end up behind if they try to take it to the macro game every time. This is why I suggested faster thirds in order to play a more effective macro game in this thread.
I see the 3 base roach aggression in PvZ as I do the 2 base roach baneling semi all-in in TvZ. It is the main aggression 3 base macro builds have to deal with when they are in their weakest state. Terrans, when adapting to this 2 base aggression, have been forced to tweak their builds to be safe, usually adding either siege tanks or banshees; most other builds are considered risky. The is similar to how the 2 base roach ling aggression from Zerg once limited what Protoss could do off of 3 gate expand and forced a lot of innovations(For example, you could no longer go 3 gate into tech while skimping on units, 1 gate expand into tech into 3 gates was developed as a reaction, and Protoss got much better at controlling versus this push.) In the same token, I believe the 3 base roach aggression will do the same to Protoss. All "safe" macro builds will have to be more refined and very specific; not all types of random Protoss builds into 3rd will work. This will effectively limit Protoss macro options, just as the roach baneling aggression limited safe Terran macro options and the 2 base roach ling aggression limited safe transitions from gateway expands.
General Tips:
Anyways, as I have said before, I think the keys to defending a 3 base build against this specific roach aggression include: scouting, sim city, cannons, forcefields, macro, unit composition, and army splitting. Anytime I see a Protoss fall to roach spam, he's always missing a few of these points.
- Scouting: A lot of people, even pros, neglect this key point. The thing is pros can, because they can sometimes assume from their opponent what they're going to do. For example, in JYP's recent series versus Stephano, JYP just assumed Stephano would do the same build each game and he was right.
- Sim city: When you can put gateways in sim city, you should; they work just as well, if not better than depots as blockers. In many cases, you can make gateways just to sim city even if you don't need them for protection as they can be more useful in defense than cannons or stalkers for the same price. Tails made this mistake of just putting gateways in his main vs Nestea at IPL and just lost when he had to waste all his forcefields to hold roaches at bay.
In addition, I think it's ridiculous that every single Protoss has different sim cities on the same map. I wouldn't have a problem with 2, 3, or maybe even 4 different versions, but literally almost every Protoss does something slightly different. People need to figure out what's best, agree to it, and do it.
+ Show Spoiler +Something I came up by watching Puzzle and Huk: Instead of trying to sim city your third on certain maps, instead sim city the ramp leading to your natural. Maps this is applicable on include Cloud Kingdom, Daybreak, and Atlantis Spaceship(when you take the rich 3rd). On each of these maps, there are problems with sim citying your 3rd for different reasons. Cloud Kingdom's 3rd has 2 entrances to sim. Daybreak's 3rd is relatively wide to fully sim and it doesn't have a ramp so roaches can easily pick off gateways if you try to do it that way. Altantis Spaceship just has an easily accessible rich 3rd you can abuse if you sim the ramp near your natural. So what you do is sim your natural and position your army at your 3rd unless you see a major attack coming from to your natural. Also, you can rally immortals to your natural to act as cannons. This game is a good example of the type of sim I'm talking about. - Cannons: Especially behind sim city, roaches cannot focus fire cannons well. Cannons+ sentries is also a more efficient use of minerals/gas for defense than stalkers
- Forcefields: This is just something people have to work on. Every Protoss besides MC I've seen use more forcefields than they have to in a given situation. Protoss in general just need to get better with this basic Protoss skill. Sim city also lessens the amount of forcefields you have to use. You also can devote forcefields into trapping and killing roaches instead of blocking roaches out of your base. One innovation I've seen mostly in PvT, but also PvZ, is getting faster but fewer sentries. Because sentries hardly do damage, getting 6 sentries earlier can equate to the same number of forcefields as 10 sentries later.
- Macro: Your opponent has an efficient macro build, that with decent macro, maxes between 11:30 and 12 minutes; you should have good macro build as well, backed by sufficient macro. A good macro build gets your 3rd as fast as possible so that when the roaches do hit, you have the economy to back up your defense. Spending all your money, especially under roach pressure, is just mechanics.
- Unit composition: Sentries are the most important unit by far. Voidrays can work, especially when backed up by sim city/forcefieds to allow them to charge; modern Protosses who rely on voidrays to hold off attacks usually get 2. Immortals also obviously work and are the best unit you can make vs roaches. Blink stalkers, on the other hand, aren't as helpful if you're relying on primarily sim city/forcefields/cannons to defend. This doesn't mean they're bad, as you usually need them to defend possible mutas if you didn't open stargate.
- Army splitting: As you're defending 2 positions, you will need to split your army into 2 parts. There are a few things you can do to more efficiently and correctly place your troops. First is scouting. An obs/voidray/phoenix/hallucination in the area outside your base between your natural/3rd. Your sim city also allows you to delay the roaches, allowing you to reposition units.
Actual Builds:
Now for some successful builds: The following are the most successful and solid macro PvZ builds I’ve seen in the last few months gathered from every broadcast PvZ. They’re also the macro builds that I use. The builds are from my personal notes, so they might be a bit messy. I've also included benchmarks, because I've become really into those recently.
- Puzzle’s fast robo into 9:30 3rd
A simple robo expand with fast observers that feigns an immortal all-in but expands behind it. In contrast to a fast third, this build gets more immortals and fewer but faster sentries for an extremely efficient roach defense. Puzzle gets 2 obs, one to scout zerg's tech and one positioned outside his bases to correctly split his army when the roach max comes. Susceptible to being blind countered by people who go blind 3 base muta without accounting for the all-in.
Build:+ Show Spoiler +3:18 gateway
3:55, 3:58 gas-20 supply 2nd game 4:05
5:12 +1
5:37 wg, sentry, sentry, sentry, sentry
6:15 2 gas
6:20 robo-2 obs, rest immortal
7:20 3 gates
7:55 2 gates
8:25 wgs finish
up to 6 sentry
9:30 3rd
10:20 twilight
10:30 3 gates in sim
+ Show Spoiler +Benchmarks:
5:00: 27 supply
6:00; 37 supply
7:00: 47 supply
8:00: 58 supply
9:00: 77 supply
10:00: 97 supply
11 minutes: 111 supply, 64 probes, 2 obs, 2 zealot, 2 immortal, 8 stalker, 6 sentry
12 minutes: 127 supply, 68 probes, 2 obs, 2 zealot, 3 immortal, 11 stalker, 6 sentry
much faster blink, +2
Game 2:
11 minutes: 111 supply, 63 probes, 1 obs, 3 immortal, 8 stalker, 6 sentry
12 minutes: 132 supply, 74 probes, 2 obs, 4 immortal, 8 stalker, 6 sentry
Game 1: Puzzle vs True: Shows a game versus a traditional roach max and how easy it is to win after you’ve defended the push efficiently.
Game 2: Puzzle vs True: Versus a roach/hydra push on a big map. Shows a colossi transition versus hydra. - Puzzle/Squirtle’s 4 gate pressure into robo into 10:10 3rd
Continuous pressure build that puts up multiple threats before expanding to force units. Pressure consists of 2 zealot poke, 4 gate +1 threat, and a gateway all-in threat.
Build:+ Show Spoiler +3:15 gate
4:20 core
4:50 2 zealot(2 cb), sentry, sentry
5:30 +1, +1 armor after
6:30 3 gates
7:30 2 gas
7:45 robo-1 obs rest immortal
4 sentry, 2 zealot/2 stalker, rest all stalker
9:20 2 gates
9:35 twilight, +2, blink after
10:10 3rd
11:20 2 gates in sim
+ Show Spoiler +Benchmarks:
5:00: 28 food
6:00: 35 food, 29 probes, 2 zealots
7:00: 46 food, 36 probes, 2 zealots, 1 sentry
8:00: 57 food, 41 probes, 2 zealots, 5 sentry
9:00: 74 food, 49 probes, 3 zealots, 2 stalker, 6 sentry
10:00: 85 food, 54 probes, 1 obs, 3 zealots, 5 stalkers, 6 sentry
11 minutes: 100 supply, 57 probes, 1 obs, 3 zealot, 1 immortal, 9 stalker, 6 sentry
12 minutes: 120 supply: 59 probes, 1 obs, 4 zealot, 2 immortal, 13 stalker, 6 sentry
Game 1: Puzzle vs Lucky: This game is versus a weaker 6 gas roach max into mutas. It shows the most likely transition into templar.
Game 2: Squirtle vs Nestea: Versus a traditional roach max. One of the most interesting things about this game was Squirtle's use of his one observer. He uses it to accomplish 2 things I've mentioned previously. After confirming the roach max style, he doesn't leave the observer in Nestea's base like most protoss would. Instead, he immediately sends it back to just right outside his base for an effective scout that lets him correctly split his army. This is the hardest working and most purposeful observer I've ever seen. - Hero’s stargate into fast robo into 10:20 3rd
This is a stargate opening that is extremely focused on defending the roach max. Gets 2 voids on smaller maps and rushes an extremely fast robo for a higher immortal count. The 2 voids, phoenix, high immortal count, and low stalker count all go very far into stopping a roach max. Also gets lots of phoenix to keep you safe versus a muta transition. However, it’s slightly weaker versus hydra based pushes on smaller maps and doesn’t allow for as strong a 170ish food push. It also more forces you to head for a colossi transition instead of templar transition versus the other builds, which are more flexible.
Build:+ Show Spoiler +gateway-zealot, stalker, zealot, sentry
5:30 stargate-voidray or 2, 7:30-5 pheonix
6:45 3rd gas
7:10 robo-1 obs, rest immortal
7:50 +1
7:50 4th gas
8:25 3 gates
10:20 3rd ideally
10:50 2 gates
+ Show Spoiler +Benchmarks:
5:00: 28 supply
6:00: 36 supply
7:00: 51 supply
8:00: 61 supply
9:00: 77 supply
10:00: 85 supply
lost: 1 pheonix, 1 sentry, 1 zealot
11 minutes: 106(112) supply, 4 zealot, 1 probe, 1 obs, 2 immortal, 1 stalker, 1 sentry, 1 voidray, 5 pheonix
lost: 5 zealot, 2 sentry, 1 immortal
12 minutes:108(132) supply
Game 1: Hero vs Stephano: Gets 1 voidray, because the map is huge. Shows how careful you have to be when expanding early with this build.
Game 2: Hero vs Stephano: Gets 2 voidrays, because of the short rush distance. Unfortunately, Hero forgets warpgates in this game. =(
