Technics' Write Up on 5 Pool
+ Show Spoiler +
I love this topic. I guess I can add some colour to it as I extensively practiced 4/5pooling zvp exactly 5 years ago for my ISL Ro32 matchup vs GoTuNk. My rationale after studying his replays was that GoTuNk delays his forge a lot and this should be punishable by 5pool on Jade to start with a nice lead in the bo5 series. The game I prepared for starts immediately in this VOD:
+ Show Spoiler +
A game I used to learn from is the Bisu vs ZerO 5pool example on Othello back in 2008, which reminds me how important distances and openness of the protoss natural dictates how P should react, which maps zerg can choose to 4/5pool on and how much damage zerg should look to inflict and be okay with it:
+ Show Spoiler +
Luckily I still remember some of the core differences when doing 4 and 5pool when playing ZvP.
Talking 4 player maps, protoss FFE: for me, 4pool is an all-in, 5pool is an eco build.
With 4pool, yes, the zerg is having earlier lings and more time to inflict more damage but it absolutely must inflict enormous amount of damage to make up for the low drone count and for the "lost" minerals of the scouting drone. I won't go into 4pool specifics as I did not practice this build as extensivly as 5pool and cannot expand on them.
With 5pool, we have a relatively healthy eco going and a nice rush. However, if the protoss is dilligent in placing his forge and cannon early on FS for example (and cross positions too), we have a really tiny window to inflict damage and usually 5pooling this kind of opponent is not worth the risk.
Some 4/5pool specifics and commentary to build onto was already mentioned:
What we should always check on any map is the distance between us and the opponent. We have to know 2 things. 1st is when lings arrive from our spawning location to the naturals of the other three locations. 2nd is when our lings arrive into their mains. This will give us a good idea where we will have more time to inflict damage against the probe/cannon defences.
One really important thing to lookout for is the protoss scouting probe and what we do about it.
In general, as a zerg I never ever try to spend more than half a second to kill the scouting probe. Only if lings and probe paths might intersect almost fully, it is then I go and try to kill the probe. When the probe moves away, I immediately continue to the natural of the protoss.
There are two options for the protoss to gives us a headache with the scouting probe:
Go around with the probe and continue in our main and harass our workers OR go back home alongside with the incoming zergling attack, intersect the zergling path and slow our lings down the road as the probe leads the group. It is critical for the zerg to react as best as possible in BOTH these scenarios.
If protoss decided to harrass the drones: If we have 2-3 seconds before our lings will be reaching the natural, i will just get a drone to chase and attack the probe if p is microing against my drones. We must be focused on the lings when they arrive at the natural because the protoss might be setting up the probe/cannon defence there.
If protoss decided to go back home we must watch if our lings are slowed down by running right after the probe's ass. In a scenario like that we just click diagonally to the left or to the right to "leave the path" of the probe and continue outrunning the probe towards the natural of the protoss.
The 5pool economy:
What I recall is that if there's no probe to harass us, it is significantly better for the zerg economy to gastrick an additional drone and at 10/9 to proceed and place a hatchery on the natural BEFORE making an overlord. If the zerg considers his drone vs probe at natural micro good enough, he might as well try to place that expo hatch even in the presence of a scouting probe as well. However, there's a big chance that placing this hatchery overlaps with the seconds where we will be engaging the protoss defences so perhaps just gastrick drone overlord will be the easier and safer thing to do.
The Battle:
The most important thing to pay attention to of course is when the 6 lings reach the protoss probe/cannon defences.
2 scenarios: 1st - the defences (probes + unfinished cannons) are at the natural; 2nd - the defences are in the main.
3 goals: kill 2+ probes, delay protoss mining time as much as possible, save as many lings as possible in the process.
Scenario 1 - The possibilities here vary a lot depending on many factors: spawn positions, defence probe count, defence probe positioning, openness of the natural, how much time left before cannons finish. Usually killing 3+ probes with 2-3 lings surviving to deny later probe scouts OR a succesful 3-4 ling run by (with a nice micro inside his base afterwards) I would consider a fine enough damage to go into the later stages of the game ahead.
Scenario 2 - The main goal is to trick protoss into killing as many of his probes as possible. Comitting for a pylon kill only is usually not a good idea. The possibilities here vary depending on these factors: whether protoss built the pylon towards us or "behind" the nexus, whether there's a gas station on our way or it is "behind" the nexus and how much time there is left before the cannon finish.
If protoss built the pylon towards us, it is easier for the protoss to kill lings in this tighter space so carefulness is in order. In this case I go for the pylon as I care about killing it, and then look to quickly switch and target probes that came to defend the pylon. In the rare case of pylon towards us + a gas station blocking our way (for example protoss on 5 o clock on FS), we might want to run around the pylon/gas wall and try to snipe probes coming from below.
If protoss built the pylon "inside" his base, then going straight for probe kills is better.
In both scenarios we look to kill at least 2 probes and save at least 3 lings to kill the pylon/forge left behind on the natural expo afterwards. We need at least 3 lings as 2 lings kill them way too slowly and he might even make zealots soon enough to save the forge. When going for the pylon/forge leftover, we do it with all lings left and we kill the pylon first for sure and then the forge. Then we send at least 1 ling into the main back to check whether P is going for 2 gates zealots or 1 gategas into quick sairs while the others can finally return back to their base to fend off that protoss scouting probe if still alive lol.
+ Show Spoiler +
A game I used to learn from is the Bisu vs ZerO 5pool example on Othello back in 2008, which reminds me how important distances and openness of the protoss natural dictates how P should react, which maps zerg can choose to 4/5pool on and how much damage zerg should look to inflict and be okay with it:
+ Show Spoiler +
Luckily I still remember some of the core differences when doing 4 and 5pool when playing ZvP.
Talking 4 player maps, protoss FFE: for me, 4pool is an all-in, 5pool is an eco build.
With 4pool, yes, the zerg is having earlier lings and more time to inflict more damage but it absolutely must inflict enormous amount of damage to make up for the low drone count and for the "lost" minerals of the scouting drone. I won't go into 4pool specifics as I did not practice this build as extensivly as 5pool and cannot expand on them.
With 5pool, we have a relatively healthy eco going and a nice rush. However, if the protoss is dilligent in placing his forge and cannon early on FS for example (and cross positions too), we have a really tiny window to inflict damage and usually 5pooling this kind of opponent is not worth the risk.
Some 4/5pool specifics and commentary to build onto was already mentioned:
What we should always check on any map is the distance between us and the opponent. We have to know 2 things. 1st is when lings arrive from our spawning location to the naturals of the other three locations. 2nd is when our lings arrive into their mains. This will give us a good idea where we will have more time to inflict damage against the probe/cannon defences.
One really important thing to lookout for is the protoss scouting probe and what we do about it.
In general, as a zerg I never ever try to spend more than half a second to kill the scouting probe. Only if lings and probe paths might intersect almost fully, it is then I go and try to kill the probe. When the probe moves away, I immediately continue to the natural of the protoss.
There are two options for the protoss to gives us a headache with the scouting probe:
Go around with the probe and continue in our main and harass our workers OR go back home alongside with the incoming zergling attack, intersect the zergling path and slow our lings down the road as the probe leads the group. It is critical for the zerg to react as best as possible in BOTH these scenarios.
If protoss decided to harrass the drones: If we have 2-3 seconds before our lings will be reaching the natural, i will just get a drone to chase and attack the probe if p is microing against my drones. We must be focused on the lings when they arrive at the natural because the protoss might be setting up the probe/cannon defence there.
If protoss decided to go back home we must watch if our lings are slowed down by running right after the probe's ass. In a scenario like that we just click diagonally to the left or to the right to "leave the path" of the probe and continue outrunning the probe towards the natural of the protoss.
The 5pool economy:
What I recall is that if there's no probe to harass us, it is significantly better for the zerg economy to gastrick an additional drone and at 10/9 to proceed and place a hatchery on the natural BEFORE making an overlord. If the zerg considers his drone vs probe at natural micro good enough, he might as well try to place that expo hatch even in the presence of a scouting probe as well. However, there's a big chance that placing this hatchery overlaps with the seconds where we will be engaging the protoss defences so perhaps just gastrick drone overlord will be the easier and safer thing to do.
The Battle:
The most important thing to pay attention to of course is when the 6 lings reach the protoss probe/cannon defences.
2 scenarios: 1st - the defences (probes + unfinished cannons) are at the natural; 2nd - the defences are in the main.
3 goals: kill 2+ probes, delay protoss mining time as much as possible, save as many lings as possible in the process.
Scenario 1 - The possibilities here vary a lot depending on many factors: spawn positions, defence probe count, defence probe positioning, openness of the natural, how much time left before cannons finish. Usually killing 3+ probes with 2-3 lings surviving to deny later probe scouts OR a succesful 3-4 ling run by (with a nice micro inside his base afterwards) I would consider a fine enough damage to go into the later stages of the game ahead.
Scenario 2 - The main goal is to trick protoss into killing as many of his probes as possible. Comitting for a pylon kill only is usually not a good idea. The possibilities here vary depending on these factors: whether protoss built the pylon towards us or "behind" the nexus, whether there's a gas station on our way or it is "behind" the nexus and how much time there is left before the cannon finish.
If protoss built the pylon towards us, it is easier for the protoss to kill lings in this tighter space so carefulness is in order. In this case I go for the pylon as I care about killing it, and then look to quickly switch and target probes that came to defend the pylon. In the rare case of pylon towards us + a gas station blocking our way (for example protoss on 5 o clock on FS), we might want to run around the pylon/gas wall and try to snipe probes coming from below.
If protoss built the pylon "inside" his base, then going straight for probe kills is better.
In both scenarios we look to kill at least 2 probes and save at least 3 lings to kill the pylon/forge left behind on the natural expo afterwards. We need at least 3 lings as 2 lings kill them way too slowly and he might even make zealots soon enough to save the forge. When going for the pylon/forge leftover, we do it with all lings left and we kill the pylon first for sure and then the forge. Then we send at least 1 ling into the main back to check whether P is going for 2 gates zealots or 1 gategas into quick sairs while the others can finally return back to their base to fend off that protoss scouting probe if still alive lol.
2018 October Update
I am still 5 pooling to this day. I'd say I've 5 pooled nearly half of my games vs. Protoss this ladder season of remastered. Here I want to provide some more tips and developments in today's scene.
+ Show Spoiler +
Trends/Patterns Today
When 1 gate fast expand became the new norm half a year ago, but mostly confined to Koreans and high level players, 5 pool became increasingly risky. Still, it is possible to defeat a worse opponent with 5 pool by maneuvering around the zealots, killing probes, and getting up their ramp. Nowadays though players do less 1 gate fast expand so it has become more doable.
Also, in this post Technics writes about the benefits of doing the gas trick and making an expansion hatchery before overlord. Contrarily, I've mostly been feeling comfortable just making the overlord and not gas tricking.
Most intermediate/high level protosses come to automatically expect an overpool, and have at least a forge timing (around 12 supply) ready for a 9 pool trying to maximize probe mining time. That means when 5 pooling half the time it results in the protoss automatically cannoning their main and abandoning the natural due to how late the forge is. Of course, rush distance, sim city, and other such factors can make situations differ.
In the case a protoss' two cannons are built at the natural before your lings can do anything, if they've pulled 6-9 probes, which some overreactive tosses do, you're actually ahead/even despite not having killed anything.
Versus really autocannon and protosses who lack subtlety in their play that get a really fast forge, this build obviously is bad. Surprisingly though even the low level tosses go for the 12 supply forge.
Every single game I drone scout. If you want to play it safe, scout at 100ish, but I generally delay it to 138, or even more. That is to squeeze out every last bit of mining while still knowing in time where toss is.
Optimizing the Economy and the Followup
The lower the supply and worker count, the more valuable each individual unit and resource becomes. Some examples are like how the 20th worker is not as valuable as the 5th worker or 100 minerals means far more 2 minutes into the game than 15 minutes. Due to 5 pool being such a microcosm of an early game build, optimizing mining in the early game is important. It's why I put incredible emphasis on queueing the two drones immediately after the overlord finishes and sending them to mine right away. Often the drones will complete just as my lings are arriving at my opponent's natural, but I take special care I take the 1 second spared to me to send the drones to mine. Also, getting the second hatch on time is important. After all, nearly all games are a matter of running away w/ the economic lead.
Most, if not all of my games always have a point going beyond the 5 minute mark and I've played numerous macro late games opening 5 pool. Sometimes I will follow up 3 hatch hydra, and even fake it out and drone off the back of it to disorient the opponent.
Taking advantage of noob building placement
If a person has a cannon on one side of their nexus, and a gateway sticking out on the opposite side of their nexus, it's in range of lings.
Example pic: + Show Spoiler +
Here the gateway side furthest from the cannon (right side of gateway) is out of range of the cannon. Protoss noobs will do this because they have little experience of being in this situation, so they aren't privy to such small minute details of building placement.
Mineral Boosting
I've also incorporated mineral boosting into my play. Here's the guide I first learned from. + Show Spoiler +
https://www.teamliquid.net/forum/bw-strategy/530231-mineral-boosting
I've found this essential to the subtlety of the 5 pool. I've mapped out and understand certain patterns of minerals on specific maps to know which ones have better mining than others. Generally minerals where workers have to travel up/down rather than left/right are inefficient. Minerals that start to curve inwards more, generally the edge ones, are like this. Also, some mineral patches (second from top, second from bottom), etc. tend to be the ones where the mineral boosting trick can be done. Then there are generally just funky mineral patches and ones you know where the drone takes a straight forward path as opposed to a longer path which you need an eye of detail for. Some maps, especially the older ones, have a lot of this funkiness across many mineral patches. The few seconds earned from this can easily mean killing more probes, or running by a protoss who tries to hold their natural.
5 Pool vs. Terran
5 pool vs. Terran is tricky and far more riskier. If they scout you first, it's like a protoss' second try scout, so the SCV arrives when the lings are out at your natural. Sometimes you can even dodge the SCV scout depending on the map. A drone scout is also essential. Generally you want to be very aggressive and keep the momentum going, so you want to continue making lings, upwards of 12, 14, before droning. Doing the gas trick to get the 7th and 8th ling is important for momentum. You can even attempt the double gas trick for the 9th and 10th ling at 10/9 supply (supply goes to 11/9).
When you're at the main of Terran things get really dicey. You want to kill SCVs only when they haven't surrounded you and you've got a good concave. Also, at times you want to avoid the SCVs and focus down the marine if you can. It's highly situational. Prevent the marine from getting into the bunker if you can.
If you picked off a supply depot outside of the range of the bunker and Terran is limited to 1 marine, it's a good opportunity to just dive in and kill as many SCVs as you can. In other cases you can perform an awesome ling surround Jaedong vs. Fantasy style:
+ Show Spoiler +
https://www.youtube.com/watch?v=kwkYpTTkGcc
That's cool but most times it just loses.
In the case they've gone factory opener, you immediately want to search for the factory. They may have hidden it in their main base. Killing the building SCV and forcing them to pull workers makes it a contention point where you can set up good engages for a concave. If the factory finishes, you'll have to do a good job of surrounding the bottom left of the factory to trap the vulture that spawns. You can do this w/ a minimum 3 lings by having one lined up at the midway point, and one aligned to the leftmost point of the factory, and a third ling down below.
Similarly, if they 14 CC'd, go for the building barracks and attempt to setup a concave, or just avoid SCVs as you hunt for the building SCV. SCV surrounds are no joke.
It's pretty good vs. terrans who wall-in at their natural, assuming they don't get an ultra fast 2nd supply depot to complete the wall-in. If you're lucky that the drone scouts first, use the drone scout to block the supply depot.
Also, make sure the annoying scout SCV doesn't kill all of your workers while busy microing lings.
I want to give a special thanks to SCC-Faust for recommending that I do an infograph as a convenient way for people to absorb information. This was my first foray into anything graphic design related, and it came out a lot wordier than I would have liked. However, I'm somewhat satisfied with it. I was really surprised to see no guides on something as basic as this and I really see the 4/5 Pool being underutilized, so I thought I'd write a guide explaining its various nuances. I also wanted to explain why 5 pool is better than 4 pool. Below I have some examples of this being done with some analysis, some games of my own, and the reference games I used for this infographic. The infograph was getting too crowded so I put down resource differences below.
Resource Differences
4 Pool 1010 minerals by 3:19
5 Pool 1218 minerals by 3:19
For the resource games I made only 6 lings and did not drone scout.
Here are the reference games I used for the timings and some replays: https://www.dropbox.com/sh/i4c6dwtk275eomo/AABZpSLzGNcILmVLG-PE_dYua?dl=0 . Unfortunately my helper did not do the standard timing which is to build the forge with the 10th probe railled to the natural, so I had to adjust a bit. I think this is a good example though of a safer protoss timing that you might see. Notice in the Double Scout 1st Try replay I do a bad split which heavily slows down my timing and throws off identifying the probe scout slightly. With a 4 Pool 1st Try I think you can force a protoss to cannon his main.
The game vs. Nexus1 is a failed 4 pool as he went 1 Gate FE.
The game vs. Dr. Dentist I saw no forge at either adjacent natural so assumed he was cross positions. Luckily he was going center proxy 2 gate.
The game vs. Calcifier I won against all odds. The opponent was cross positions, went 1 Gate FE, but because I miraculously got the surround on the 1st zealot with my initial 6 lings, I won the game.
www.youtube.com
July vs. Best. In this game July successfully mind games his opponent, starting the series unexpectedly with a 5 pool. Best purposely delayed his forge, opting to make a nexus first. By the time July's zerglings hatch, Best has only then started his forge. Notice how because Best's probe arrives by the time July's zerglings are at the ramp, July knows Best has spawned cross positions without needing a drone scout.
http://bwreplays.com/players/LocDog
Pick the Camilia vs. LocDog 9 minute replay on Match Point. I thought LocDog was a pretty high level player, so he'll play greedy, and he'd probably scout really late on this 2 player map.
www.youtube.com
Here Kwanro is fortunate because Goozila scouts him in the last possible location. His 4 pool however is unsuccessful due to losing 3 of his zerglings due to carelessly letting them get drilled by probes. He also made 2 more additional pairs of lings, the last pair coming far too late before the cannon is up. His mining is significantly slowed down due to investing in so many lings and his unsuccessful harass.