|
United States8476 Posts
I've been seeing a lot of zergs asking questions about how to stop 4 gates and I've seen even more people give terrible advice about how to stop it. I also know that even many top level zergs don't know the proper way to stop this. I hope I can help some people by sharing what I know about this strategy. Feedback is greatly appreciated.
The main problem with the general advice that people on teamliquid give about stopping 4 gates is that they're usually about the unit composition to get vs a 4 gate. This is wrong for 2 reasons. First of all, there are many types of 4 gates that come with different timings and different compositions and each has a somewhat different counter. Second of all, unit composition is, in my opinion, the least important aspect of stopping a 4 gate. Most 4 gates, especially the later 4 gates can be stopped with a variety of tactics. In order to stop a 4 gate, the two most important aspects are:
1. Scout the 4 gate
2. Cut drones at the correct timing
Everything else is secondary. Of course, you want spine crawlers and a good unit composition, but these things are by far not the reason most people lose to 4 gates.
Scouting the 4 gate:
First, to stop the 4 gate, you must know that it's coming. This can be done in various ways. Here are some techniques that I've come up with.
1. A drone scout can determine if the gateway was built on 10 or 13 and it can check the nexus for the number of chornoboosts the toss has saved.
2. Overlords can be sacked to check the number of gates and the unit composition of the protoss. 3. Finding out the toss's 2nd unit out of his first gateway can give you a big clue as to what strategy he's doing. A stalker most likely indicates a tech build/4gate. A sentry rules out the zealot/stalker 4 gate, so free feel to drone to 30 drones.
4. Killing the initial scouting probe lets you deny proxy plyons that will make his push much faster. 5. You can run lings up the protoss's ramp every 20 seconds to check how many chornoboosts he uses on his warp gate tech.
6. If you learn warpgate timings, you can tell how many chronos the toss has used on his warpgate tech. This can give you a hint on the mindset of a good protoss. 3 chornos on the warpgate could possibly mean a 4 gate, 4 definately means a 4 gate, and 0-2 probably means either no 4 gate or a very delayed 4 gate. These timings off a 13 gate are approximately:
5 chornoboosts: 5:20
4 chornoboosts: 5:30
3 chornoboosts: 5:40 etc...
0 chornoboosts: 6:10
These timings indicate when the warpgate tech finishes and gateways start morphing. Units can be warped 10 seconds later and will finish warping in 15 seconds. For example, if you use 4 chornoboosts on warpgates, warpgates will finish at 5:30, units will start warping in at 5:40, and units will have finished warping at 5:45.
Types of 4 gates
By learning the types of possible 4 gates, you will be able to recognize them in a game and respond accordingly.
Pure zealot 4 gate
This is by far the easiest 4 gate to stop as long as you scout it. The protoss can stop mining gas after 50 gas and focus only on zealots. If he doesn't make even 1 stalker, he won't be able to deny scouting. Throw down a roach warren and you should be safe.
Zealot/stalker 4 gate
This version of 4 gate comes out very quick and is probably the hardest to stop. Toss cuts probes at 20 probes and in order to stop this variation, you need to cut drones around 18 drones on certain maps(20 drones on most maps). As a reference, this build can have 7 toss units around 5:35. The zerg's 2nd hatch will just have finished by that time off of a 14 gas 14 pool build and therefore there is no way you can build spines at your natural to hold it off in time. On maps with narrow chokes, it's possible to preemptively build a spine crawler in your map and then walk it down to your natural. On maps with wide chokes, you absolutely need both speedlings and roaches to defend against this type of 4 gate. Make more roaches if he's zealot heavy and more lings if he's stalker heavy. Surround the stalkers with your lings and kite the zealots with roaches. Even then, if you do everything perfectly, it's still a micro war with the zerg having a slight advantage.
Zealot/sentry 4 gate
This comes slightly later because toss must get 2 gas in order to execute this version. Toss usually cuts probes around 23 in this version and the initial push is generally pure sentry/zealot. This version might look like a normal 3 gate sentry expand, but you won't be fooled if you have constant scouting of his expansion via either a ling or an overlord. Also, you can sometimes tell by the sheer number of units he has. If the toss spots roaches, he will add a round of stalkers. Spines are decently effective against this 4 gate, because this army generally has a lower dps. A good amount of roach/ling is also effective, because the toss will simply get overwhelmed. Mass lings are not effective at all vs this. As long as you don't overextend yourself, you should hold this off. One key thing to be careful of vs this version is not to let toss up your main ramp or it's pretty much over for the zerg as he can endlessly forcefield your ramp.
Delayed 4 gate
These are by far the most varied and are only hard to stop when you can't scout it or your macro is poor. By 6:10, if the toss has not expanded yet or moved out, you definitely should suspect that something is up. If you have no idea what the toss is doing, then you're not in a very good position. The reason 4 gate is problematic in this case is that you have to prepare not only for 4 gate but also a vast myriad of other threats such as voidrays, phoenix, dts, blink stalker, immortals, and warp prisms. In any case, by 6:10, vs a 1 base toss, you should start your lair to prepare for air/dts/blink stalker and pump mass drones until 30 drones and then cut drones entirely. This is because all the threats i mentioned before actually all come at the same time, around 7:10 to 7:30 and if you cut at 30 drones you should be able to properly respond to all of them. Also, with 30 drones and 2 bases, you'll definitely be able to outmine a toss who has only 1 base anyways if he says on 1 base for too long. In any case, with good macro and a mixture of lings and roaches, you'll hold.
Varying your defense based on the map
Next, you have to vary your 4 gate defense techniques based on the map being played. All of these factors contribute to determining whether spines should be a major component of the 4 gate defense and how much time you have to react.
Maps with long rush distance and tight chokes(lost temple, shakrus plateau)
Lots of spine crawlers are the best answer for defense as you support those spines with additional units. These maps are the worst maps for 4 gate, but it's still possible to lose to a 4 gate if you're not prepared or you don't abuse the choke.
Maps with wide chokes(xelnaga, metal and delta)
These naturals are not defended as efficiently with spine crawlers. Although spines are definitely still useful in these situations be wary that with improper spine placement, the protoss can simply walk around your spines.
Maps with short rush distances(steppes, close distance metal, close distance lost temple, delta)
These maps force you to make preemptive spines because you won't have enough time to make reactionary spines after he moves out of his base.
Maps with ramps to the main that are easy to get up(xelnaga, delta quadrant)
Pay special attention to your ramp on these maps if your opponent has sentries. If the toss gets up the zerg's ramp with sentires, the game is pretty much already over.
Maps with backdoors(blistering, scrap station, and jungle basin)
These generally make spines somewhat less effective. Don't overly invest in them(~3). However it takes forever to kill the rocks, so if you scouted and droned correctly, the time it takes for him to kill the rocks should translate into a bigger army advantage for you.
Unit choices
Unit choices are the least important part of defending a 4 gate. For most of the 4 gates(the later ones), unit choice isn't really that important. Before you start your game, you should have a build in mind and this build should be flexible enough to handle any of these 4 gates. If you scout a 4 gate, you don't have to abandon your build entirely. Instead, just adapt by cutting drones and making whatever unit you were going to make in the first place. Ling roach, ling bling, +1 speedlings, and hydra are all somewhat decent choices to stop some of the later 4 gates. Muta, however, is not.
|
The drone timings are huge and I'd love to hear more specifics if possible.
I lost for a month straight against 4-gating Protoss until I finally started to understand that the biggest problem that no one is giving advice on is not unit composition, but macro timings.
I've found that the mindstate of droning to 30 has turned the tide on most 4 gate builds for me (only cutting drones a little earlier if you see he's really pumping for that very fast 1-gas zealot heavy push). The entire art of beating a 4 gate revolves entirely around drone timing and when to cut drones. I think the unit compositions are really secondary. There is no best answer because there are so many ways the protoss can mix up his composition balance, and also different players are better with different units.
Would you say even probe counting is helpful? How many drones does a Z need to be up on a P to counter any of the pushes would you say... I feel that Zerg really needs the stronger econ in these scenarios by a few drones if possible.
|
Very nice post and as far as I can tell really valuable information. I can't comment on the timing details because I don't pay attention to the clock timer in-game, but everything seems to agree with my sense of the timings.
On December 21 2010 01:24 4kmonk wrote:
The first and imo the strongest type is the zealot stalker 4gate. This 4 gate comes out the quickest and is probably the hardest to stop. Toss cuts probes at 20 probes and in order to stop this variation, you need to cut drones around 18 drones on certain maps. As a reference, this build can have 7 toss units around 5:35. The zerg's 2nd hatch will just have finished by that time off of a 14 gas 14 pool build and therefore there is no way you can build spines at your natural to hold it off in time. On maps with narrow chokes, it's possible to preemptively build a spine crawler in your map and then walk it down to your natural. On maps with wide chokes, you absolutely need both speedlings and roaches to defend against this type of 4 gate. Make more roaches if he's zealot heavy and more lings if he's stalker heavy. Surround the stalkers with your lings and kite the zealots with roaches. Even then, if you do everything perfectly, it's still a micro war with the zerg having a slight advantage.
My understanding of defending this rush was that you're best off sticking to pure lings. 18 (or 20, which was my point of cutting drones) drones is a very meager economy to go roaches off and the rush comes so quickly that you'll be hard-pressed to get any decent number of roaches out. For reference, check out Leenock vs GuineaPig on Blistering Sands in GSL in which he overwhelms with sheer ling numbers and Check vs Chojja on DQ where Check gets out roachling but still gets overwhelmed. Of course the games were quite different and DQ is extremely hard ZvP, but it might be better to go pure lings.
There's also the pure Zealot 4 gate, where P goes 10 gate, saves up all his cb for warpgate and stops mining gas after 50. This is way cheesier than the other 4 gates in that as long as it's scouted, Z shouldn't have any problems holding it off (all you need is a very early roach warren), but it's something to watch out for.
|
United States8476 Posts
On December 21 2010 01:40 gunsakimbo wrote:
The drone timings are huge and I'd love to hear more specifics if possible.
I lost for a month straight against 4-gating Protoss until I finally started to understand that the biggest problem that no one is giving advice on is not unit composition, but macro timings.
I've found that the mindstate of droning to 30 has turned the tide on most 4 gate builds for me (only cutting drones a little earlier if you see he's really pumping for that very fast 1-gas zealot heavy push). The entire art of beating a 4 gate revolves entirely around drone timing and when to cut drones. I think the unit compositions are really secondary. There is no best answer because there are so many ways the protoss can mix up his composition balance, and also different players are better with different units.
Would you say even probe counting is helpful? How many drones does a Z need to be up on a P to counter any of the pushes would you say... I feel that Zerg really needs the stronger econ in these scenarios by a few drones if possible.
I totally agree with you. The most important parts about defending 4 gate is scouting it and drone cutting. I don't necessarily agree that zerg has to be ahead on many drones to defend these 4 gates. Zergs units are actually just better than gateway units on equal cost as long as you have a decent composition. Also, in some cases, it's just impossible to be really ahead of the toss in worker count, especially with the faster pushes. For example, the zealot/stalker push on delta only lets the zerg make around 18 drones imo.
About probe counting. I feel that it would be too hard to count individual probes, but you can definitely look at the nexus and see if it's still making probes when you scout the toss base.
|
On December 21 2010 01:42 Orome wrote:
18 (or 20, which was my point of cutting drones) drones is a very meager economy to go roaches off and the rush comes so quickly that you'll be hard-pressed to get any decent number of roaches out.
the pull at 18 to 20 advice is what I had been following for a while in trying to stop the 4-gate, but for me it never worked out too well. I think it may be b/c it works for pro players, but the timings translate a little differently on ladder, even against alot of diamond players. You can hold the first wave from a 4-gate push on such a crap economy, but you can and will not hold the 2nd or 3rd (unless you get the pylon and that sometimes can be extremely difficult). this is why I feel that it is essential to try and eek out a few more drones.
It does all rely on scouting though and understanding how spot on your opponent is with his timing, when he's cut his probes, etc. Good advice above too on just watching for when the nexus isn't making probes anymore if possible. There are so many variants of how people do things on ladder, and my experience so far is that most of them are just slightly delayed in their pushes and are not the extreme of the lean econ versions.
|
United States8476 Posts
On December 21 2010 01:42 Orome wrote:Very nice post and as far as I can tell really valuable information. I can't comment on the timing details because I don't pay attention to the clock timer in-game, but everything seems to agree with my sense of the timings. Show nested quote +On December 21 2010 01:24 4kmonk wrote:
The first and imo the strongest type is the zealot stalker 4gate. This 4 gate comes out the quickest and is probably the hardest to stop. Toss cuts probes at 20 probes and in order to stop this variation, you need to cut drones around 18 drones on certain maps. As a reference, this build can have 7 toss units around 5:35. The zerg's 2nd hatch will just have finished by that time off of a 14 gas 14 pool build and therefore there is no way you can build spines at your natural to hold it off in time. On maps with narrow chokes, it's possible to preemptively build a spine crawler in your map and then walk it down to your natural. On maps with wide chokes, you absolutely need both speedlings and roaches to defend against this type of 4 gate. Make more roaches if he's zealot heavy and more lings if he's stalker heavy. Surround the stalkers with your lings and kite the zealots with roaches. Even then, if you do everything perfectly, it's still a micro war with the zerg having a slight advantage. My understanding of defending this rush was that you're best off sticking to pure lings. 18 (or 20, which was my point of cutting drones) drones is a very meager economy to go roaches off and the rush comes so quickly that you'll be hard-pressed to get any decent number of roaches out. For reference, check out Leenock vs GuineaPig on Blistering Sands in GSL in which he overwhelms with sheer ling numbers and Check vs Chojja on DQ where Check gets out roachling but still gets overwhelmed. Of course the games were quite different and DQ is extremely hard ZvP, but it might be better to go pure lings. There's also the pure Zealot 4 gate, where P goes 10 gate, saves up all his cb for warpgate and stops mining gas after 50. This is way cheesier than the other 4 gates in that as long as it's scouted, Z shouldn't have any problems holding it off (all you need is a very early roach warren), but it's something to watch out for.
I disagree with you about the lings. First of all, I agree with getting 20 drones vs this push on almost every map except the extremely short distance ones such as steppes, delta, and close pos metal. Then, I find that 18 is probably better. It's a very fine distinction tbh and i'll talk about it latter.
About pure lings: In certain circumstances, if you make pure lings vs this strategy, protoss can just switch to mass zealot and roll over you and your natural imo. I think the 2 games you mentioned perfectly demonstrate what I mean by this.
Leenock vs GuineaPig on blistering
Leenock stops at 19 drones and then saves larva. To stop the rush, he makes a spine and masses zerglings. First of all, blistering is a relatively easy map to stop this rush imo. The rush distance is very long and your choke is decently narrow. This makes spines a much better option than normal. Also, guineapig did a rather slow version of this build, not using many chornoboosts on his warpgates. Then, you have to account for the many many mistakes that guineapig made in his execution of this build. He messed up his wall and lings runby killing 2 probes and stopping mining for while. He loses trades his 2 stalkers for 4 lings, he wastes 3 forcefields, then he procedes to delay the toss even more by hitting his back rocks. Also, it takes forever for the toss to get up a proxy pylon to reinforce. The final fight is actually still really close. Imagine if this happened on a different map and guineapig had +2 stalkers, +3 forcefields, and 1 or 3 extra units that would have come with the lost probes and mining time. Also, imagine if guineapig hadn't been delayed so much with the ling runby and he put his proxy pylon down faster. Pure ling definitely isn't the answer but maybe lings + spines into blings on certain maps would work. I don't have enough experience to answer that though.
choya vs Check on delta
This is by far the hardest map to hold this rush on. Although check did cut drones, and make lings and roaches, I felt as if he made a few key mistakes. His biggest was cutting drones at 16 instead of 18, which is definitely not enough to hold this push. He also tries to make a spine with a drone, bringing it down to 15 drones. This was made worse because choya actually delayed his push for 3 rounds of units, further widening the econ gap. On a map like delta, lings and and spines would never hold this push, because of the short rush distance and the wide choke to the natural in addition to the fact that the ramp to the main can't be covered with spines.
Also, it's especially difficult on this map because if toss moves out with his first 7 units, you won't hold if you're close position and make 20 drones. However, if you cut drones at 18 or 20 as check did, you'll be much worse off vs a toss that waits for more waves. Notice how tight these builds are as we're talking about 16 vs 18 vs 20 drones.
|
On December 21 2010 02:27 4kmonk wrote:
I disagree with your first point.
Dimaga style +1 speedling works well in my opinion. Of course.. always being prepped to tech switch yourself is very important.
I think if you go pure ling against this, throwing an evo down and getting the +1 attack is the only way to ensure you can't be run over by pure zeal.
|
Interesting, and it seems you have way more knowledge and experience with this than I do. You might be right, although I believe Artosis has advocated getting banelings instead of roaches as well.
What level do you play at yourself?
|
United States8476 Posts
On December 21 2010 02:38 gunsakimbo wrote:Dimaga style +1 speedling works well in my opinion. Of course.. always being prepped to tech switch yourself is very important. I think if you go pure ling against this, throwing an evo down and getting the +1 attack is the only way to ensure you can't be run over by pure zeal.
If you read the whole reply, you'll notice that I didn't completely disagree with opening pure lings. I just disagreed with opening pure ling on certain maps and staying pure ling against this. Banelings are a much better alternative to +1 lings against zealots. The +1 for lings just doesn't finish fast enough to stop the zealot/stalker 4 gate.
|
mass lings with +1 armor
User was warned for this post
|
United States8476 Posts
On December 21 2010 02:41 Orome wrote:
Interesting, and it seems you have way more knowledge and experience with this than I do. You might be right, although I believe Artosis has advocated getting banelings instead of roaches as well.
What level do you play at yourself?
I actually play protoss and I regularly play with top 200 and top 100 US zergs. Again, I don't have much experience with banelings, but it most probably works on large maps with narrow chokes.
|
On December 21 2010 02:45 4kmonk wrote:Show nested quote +On December 21 2010 02:41 Orome wrote:
Interesting, and it seems you have way more knowledge and experience with this than I do. You might be right, although I believe Artosis has advocated getting banelings instead of roaches as well.
What level do you play at yourself? I actually play protoss and I regularly play with top 200 and top 100 US zergs. Again, I don't have much experience with banelings, but it most probably works on large maps with narrow chokes.
Banelings are good too... but you need to be on the ball if you get them. It's the whole "make the minimum amount and make sure to not let any go to waste" kind of deal. If you have good ling/bling micro I say it's awesome. If you're ehh with micro and are more comfortable just getting a good position and a moving a bit... then I think roach style is better. It's a matter of play style.. but also a matter of how tight the game is... blings don't transition too well out of the early game unless you know how to use them well.
|
When I used to 4 gate (as Toss), the two things that would stop my 4 gate cold is pure +1 lings, but also 4 well-placed crawlers, especially if they used the evo chamber to block pathing to the crawlers (and ramp). Fast roaches and banelings have worked too, but mass speedlings is what I feared most because there was absolutely no way to recover. I find 4 gate very weak because it doesn't provide enough gas for adequate sentries for the push. And if the push fails, it's pretty much gg.
FYI: 4 gate is scouted if toss is putting down multiple gates off only one gas. As Zerg, I would always scout the gas to see what is coming (I see too few Zerg scout gas, instead scout the front of my main). If only one gas is taken, you can be pretty sure 4 gate is coming. If you scout two gas, then you have to try and scout further to see if there are multiple gates (likely meaning 3 gate expo), a robo (w/ 1-2 gate = likely expo, w/ 3 gate = likely one-base timing push), and 1-2 gates with nothing else = likely void ray.
That's all the help for Zerg I feel like giving for this week 
|
The more I think about it the more I like your OP. Good solid advice and the times you placed are very helpful.
|
United States8476 Posts
On December 21 2010 02:44 Allscorpion wrote:
mass lings with +1 armor
User was warned for this post
Ignore this. +1 armor does nothing vs a 4 gate.
|
United States8476 Posts
I just did a few updates.
|
very nice post. insightful and very helpful!! big thanks here and keep the work up 
p/s: would be great if you can throw some replay up!
|
One thing that makes 4 gate even harder to stop is when toss pokes with 1 zealot+1 stalker or 2 stalkers in the beginning. That makes it even harder to squeeze in drones.
|
Thank you very much for this post!
|
Shouldn't hatch first make it possible to get spinecrawlers up rly fast?
|
|
|
|
|
|
EPT
