The Art of Cannon Rushing - Page 2
Forum Index > StarCraft 2 Strategy |
Fat_Elephant
Angola54 Posts
| ||
AkashSky
United States257 Posts
For a future project, could you perhaps calculate mineral mining efficiency on various maps? I notice that I get a lot more money on overgrowth than on other maps in the early game. I think this is because of the mineral placement, but are there any mathematical limitations on this? The Art of cannon rushing was a really good read. Keep up the good work! | ||
![]()
Whitewing
United States7483 Posts
| ||
printf
United States13 Posts
However, as a top masters cannon rusher (I haven't quite gotten GM yet though), there are a few things I would like to note. PvP + Show Spoiler + For PvP, I think you discounted 4 pylon wall offs far too quickly. I understand and agree with your assertion that pvz they're far too expensive, but in PvP you're all in anyway, and a 4 pylon wall off with often allow you to execute a direct, rather than an indirect, cannon rush. The reason why this works requires analyzing a variable that you exempted entirely from all of your calculations, however. You did a good job of accounting for the time it takes to kill the pylons (although there's another issue with how you did this that I'll address in a different spoiler tag), but you didn't account for the ability to reinforce pylons. When one pylon is killed, it is possible to build a backup pylon directly behind it, with a 1 hex overlap, such that the attacking probes cannot break through the wall. With 3 pylon walls this is usually impossible, which is why I don't ever recommend a 3 pylon wall off in PvP unless it's indirect, but with a 4 pylon wall it's generally very possible. By utilizing backup pylons, you can virtually guarantee that a cannon finishes (although 10 gate fast zealot can screw with this). The immediate question that should be asked now is whether or not the 4 pylon wall can be blocked before it gets placed. After all, it is being built directly behind the mineral line. The answer to this, of course, varies from map to map and wall off to wall off, however the best 4 pylon walls cannot actually be blocked, even with a 0 second response time (this one for example) (although I'll admit in the video I did a poor job of proving this, I actually timed it out once and it does indeed take more time to block it than to create it). Next are the ideas that have been presented already in this thread, of either making a ninja expansion or proxying void rays. A ninja CAN work if you also ninja your cybernetics core. If you fail to do this second step, you'll lose all your tech and die to a 4 gate. Even if you do ninja your cyber however, you still are behind, because you lose a significant amount of mining time. Proxy voids shouldn't work unless the cannon rusher screws up. It's a very simple procedure to proxy a gateway and cybernetics core at the other guy's base, get a couple assimilators back home, and then pump out stalkers. In 99.9% of cases you won't be able to afford a second void ray, and it only takes 3 stalkers to kill one void. The only time you can get a second void is if the cannon rush is indirect, you don't do anything to stop it, and you let all your probes die mining every last mineral possible. Even then it's not guaranteed. Void rays are damn expensive. EDIT: it should be mentioned that the responses to ninja's and proxy voids only work if you see it coming before it actually happens. Proxy void is telegraphed very early by continued gas mining without commitment to a stalker or mothership core in defense. The gas has to be going somewhere, and the only real option is proxied void rays. As far as a ninja goes, it's pretty much the same situation, except gas mining may be cut, and a mothership core may be built for counter harass (which is shut down by a single cannon in the mineral line). EDIT 2: I just noticed the OP also mentions a counter-attacking with stalkers and a mothership core. This can work, and is actually highly effective, but only in certain scenarios. If the cannon rush is indirect and originates at the ramp, then it's a simple matter to trap the stalkers, and a mothership core by itself is no threat. If, however, you can stall the cannon rush until you have 3-4 stalkers, then you can take those and counterattack, and which point it becomes very difficult to hold. A wall off at the ramp which emergency cannons is the best response, and can actually hold this counter-attack with a probe pull, but if this counter-attack is accompanied by a ninja base, then it might do enough damage to get back in the game, and possibly even get ahead. If more than 1 stalker gets into the main base, especially with a mothership core as support, it's very likely GG Pylon weakness and cannon weakness + Show Spoiler + Ok, so on the topic of pylon weakness. I appreciate the math you did, regarding how long it takes to kill a pylon, and this is very nice, and very relevant. The only issue is some of the conclusions you reaches based on this math. Take the overgrowth triple pylon as an example. If 12 drones are pulled, 6 on the top left pylon and 6 on the bottom left pylon, then according to your math it will take 34 seconds to kill those 2 pylons. Since a cannon takes 45 seconds, on top of the 25 from a pylon, this leaves 36 seconds to kill the cannon, which can now be attacked from something like 260 degrees. (it's also possible I misread this and that the 14 seconds weren't accounted for by the 34 rating, but this still leaves 20 seconds of 6-8 drones attacking the cannon, which should be enough) Basically, while the pylons will finish, allowing the cannon to be planted, the pylons can be killed in time to expose the cannon to be killed. Pulling 12 drones as in the example I gave is, of course, a significant commitment from the zerg, and requires the drones to be micro'd to get maximum surface area on the pylons, but it will shut down the cannon rush (theoretically) (I use an entirely different cannon rushing style myself, so I can't speak from experience. But this entire post is working off theory and math, so I think that this is still relevant). Building extraneous pylons + Show Spoiler + The mark of a skilled cannon rusher (in my eyes) is the minimalist quality of building as few structures as absolutely necessary, so I'm going to have to disagree with you regarding "For this reason, make sure that when you build your first Pylon, you build the other two immediately afterwards". If at all possible, don't build those additional 2 pylons. If you are forced to build them by a drone pull, and then the zerg retreats the drones after you place them, then make sure you cancel those pylons. Letting them finish is an absolute waste of 200 minerals, especially considering how likely the zerg is to cancel the hatchery. As long as you're paying attention, you shouldn't be caught off guard by a drone pull, and having the placement blocked shouldn't be a legitimate issue. This goes for PvP as well when executing an indirect cannon rush, as on Overgrowth with the 3 pylon wall directly next to the ramp. If at all possible, avoid building the extra 2 pylons to finish the wall, it's possible the other player won't even notice it happening, and you would have just thrown away an entire cannon worth of minerals for no reason. There are also a bunch of tricks and techniques that can be used for cannon rushing that aren't covered here, which I don't feel like writing down so I'll just link to a video I made for them (technically this is a playlist, the first 2 videos in the playlist are to what I am referring) (there are couple more that I forgot to put in the videos, such as using 2 pylons at the bottom of a ramp to wall in stalkers, rather than needing a pylon and a gate. I'll make another video at some point with all the stuff I forgot). Anyway, despite how my tone may have come across, I really do think that this write-up was well done and well researched (although I do resent the assertion that cannon rushing doesn't require skill). Also, on the topic of PvT + Show Spoiler + Cannon rushing PvT I don't think should work, but yet even at top masters, and frequently playing GM's, I manage to pull a 40% win rate in the matchup. The reason it shouldn't work is that marines are ranged, and therefore very good at killing walled-in cannons, killing probes, and denying cannon creep. The reason it works is that a lot of people are uneducated about cannon rushes, and, like many people in this thread, think that flying away will solve all your problems. If you fly away, you die. Even if the scv's aren't all trapped, the lost mining time and, more importantly, tactically disadvantageous position of the new base, mean that it's easy to win with proxied zealots/stalkers or a 4 gate. I rarely lose to a terran flying away. Also, if the cannon rusher includes gateways in the rush, it is possible, although very difficult and requiring a lot of apm/multitasking from both players, to harass one side of the base with zealots, while building cannons on the other side of the base. Because marines can't be in 2 places at once, this can eventually reach the mineral line and win the game. If you are ever playing against somebody you know will cannon rush you, just go 11/11 double rax at home, and pump a billion marines for a free win. If you are opening standard and get cannon rushed, then know that reapers are your friend if you can send them to the protoss base (this forces out 2 defensive cannons), or alternatively a well-placed bunker with a bunch of marines is very strong. Be sure to pull scv's though to force the protoss to invest into the full wall-off (because, as I mentioned earlier, an adept cannon rusher will not waste 200 minerals if at all avoidable) | ||
printf
United States13 Posts
![]() | ||
Ovid
United Kingdom948 Posts
On March 23 2015 03:04 printf wrote: Hello, firstly I would like to compliment you on this very thorough and mathematical guide to cannon rushing. However, as a top masters cannon rusher (I haven't quite gotten GM yet though), there are a few things I would like to note. PvP + Show Spoiler + For PvP, I think you discounted 4 pylon wall offs far too quickly. I understand and agree with your assertion that pvz they're far too expensive, but in PvP you're all in anyway, and a 4 pylon wall off with often allow you to execute a direct, rather than an indirect, cannon rush. The reason why this works requires analyzing a variable that you exempted entirely from all of your calculations, however. You did a good job of accounting for the time it takes to kill the pylons (although there's another issue with how you did this that I'll address in a different spoiler tag), but you didn't account for the ability to reinforce pylons. When one pylon is killed, it is possible to build a backup pylon directly behind it, with a 1 hex overlap, such that the attacking probes cannot break through the wall. With 3 pylon walls this is usually impossible, which is why I don't ever recommend a 3 pylon wall off in PvP unless it's indirect, but with a 4 pylon wall it's generally very possible. By utilizing backup pylons, you can virtually guarantee that a cannon finishes (although 10 gate fast zealot can screw with this). The immediate question that should be asked now is whether or not the 4 pylon wall can be blocked before it gets placed. After all, it is being built directly behind the mineral line. The answer to this, of course, varies from map to map and wall off to wall off, however the best 4 pylon walls cannot actually be blocked, even with a 0 second response time (this one for example) (although I'll admit in the video I did a poor job of proving this, I actually timed it out once and it does indeed take more time to block it than to create it). Next are the ideas that have been presented already in this thread, of either making a ninja expansion or proxying void rays. A ninja CAN work if you also ninja your cybernetics core. If you fail to do this second step, you'll lose all your tech and die to a 4 gate. Even if you do ninja your cyber however, you still are behind, because you lose a significant amount of mining time. Proxy voids shouldn't work unless the cannon rusher screws up. It's a very simple procedure to proxy a gateway and cybernetics core at the other guy's base, get a couple assimilators back home, and then pump out stalkers. In 99.9% of cases you won't be able to afford a second void ray, and it only takes 3 stalkers to kill one void. The only time you can get a second void is if the cannon rush is indirect, you don't do anything to stop it, and you let all your probes die mining every last mineral possible. Even then it's not guaranteed. Void rays are damn expensive. EDIT: it should be mentioned that the responses to ninja's and proxy voids only work if you see it coming before it actually happens. Proxy void is telegraphed very early by continued gas mining without commitment to a stalker or mothership core in defense. The gas has to be going somewhere, and the only real option is proxied void rays. As far as a ninja goes, it's pretty much the same situation, except gas mining may be cut, and a mothership core may be built for counter harass (which is shut down by a single cannon in the mineral line). EDIT 2: I just noticed the OP also mentions a counter-attacking with stalkers and a mothership core. This can work, and is actually highly effective, but only in certain scenarios. If the cannon rush is indirect and originates at the ramp, then it's a simple matter to trap the stalkers, and a mothership core by itself is no threat. If, however, you can stall the cannon rush until you have 3-4 stalkers, then you can take those and counterattack, and which point it becomes very difficult to hold. A wall off at the ramp which emergency cannons is the best response, and can actually hold this counter-attack with a probe pull, but if this counter-attack is accompanied by a ninja base, then it might do enough damage to get back in the game, and possibly even get ahead. If more than 1 stalker gets into the main base, especially with a mothership core as support, it's very likely GG Pylon weakness and cannon weakness + Show Spoiler + Ok, so on the topic of pylon weakness. I appreciate the math you did, regarding how long it takes to kill a pylon, and this is very nice, and very relevant. The only issue is some of the conclusions you reaches based on this math. Take the overgrowth triple pylon as an example. If 12 drones are pulled, 6 on the top left pylon and 6 on the bottom left pylon, then according to your math it will take 34 seconds to kill those 2 pylons. Since a cannon takes 45 seconds, on top of the 25 from a pylon, this leaves 36 seconds to kill the cannon, which can now be attacked from something like 260 degrees. (it's also possible I misread this and that the 14 seconds weren't accounted for by the 34 rating, but this still leaves 20 seconds of 6-8 drones attacking the cannon, which should be enough) Basically, while the pylons will finish, allowing the cannon to be planted, the pylons can be killed in time to expose the cannon to be killed. Pulling 12 drones as in the example I gave is, of course, a significant commitment from the zerg, and requires the drones to be micro'd to get maximum surface area on the pylons, but it will shut down the cannon rush (theoretically) (I use an entirely different cannon rushing style myself, so I can't speak from experience. But this entire post is working off theory and math, so I think that this is still relevant). Building extraneous pylons + Show Spoiler + The mark of a skilled cannon rusher (in my eyes) is the minimalist quality of building as few structures as absolutely necessary, so I'm going to have to disagree with you regarding "For this reason, make sure that when you build your first Pylon, you build the other two immediately afterwards". If at all possible, don't build those additional 2 pylons. If you are forced to build them by a drone pull, and then the zerg retreats the drones after you place them, then make sure you cancel those pylons. Letting them finish is an absolute waste of 200 minerals, especially considering how likely the zerg is to cancel the hatchery. As long as you're paying attention, you shouldn't be caught off guard by a drone pull, and having the placement blocked shouldn't be a legitimate issue. This goes for PvP as well when executing an indirect cannon rush, as on Overgrowth with the 3 pylon wall directly next to the ramp. If at all possible, avoid building the extra 2 pylons to finish the wall, it's possible the other player won't even notice it happening, and you would have just thrown away an entire cannon worth of minerals for no reason. There are also a bunch of tricks and techniques that can be used for cannon rushing that aren't covered here, which I don't feel like writing down so I'll just link to a video I made for them (technically this is a playlist, the first 2 videos in the playlist are to what I am referring) (there are couple more that I forgot to put in the videos, such as using 2 pylons at the bottom of a ramp to wall in stalkers, rather than needing a pylon and a gate. I'll make another video at some point with all the stuff I forgot). Anyway, despite how my tone may have come across, I really do think that this write-up was well done and well researched (although I do resent the assertion that cannon rushing doesn't require skill). Also, on the topic of PvT + Show Spoiler + Cannon rushing PvT I don't think should work, but yet even at top masters, and frequently playing GM's, I manage to pull a 40% win rate in the matchup. The reason it shouldn't work is that marines are ranged, and therefore very good at killing walled-in cannons, killing probes, and denying cannon creep. The reason it works is that a lot of people are uneducated about cannon rushes, and, like many people in this thread, think that flying away will solve all your problems. If you fly away, you die. Even if the scv's aren't all trapped, the lost mining time and, more importantly, tactically disadvantageous position of the new base, mean that it's easy to win with proxied zealots/stalkers or a 4 gate. I rarely lose to a terran flying away. Also, if the cannon rusher includes gateways in the rush, it is possible, although very difficult and requiring a lot of apm/multitasking from both players, to harass one side of the base with zealots, while building cannons on the other side of the base. Because marines can't be in 2 places at once, this can eventually reach the mineral line and win the game. If you are ever playing against somebody you know will cannon rush you, just go 11/11 double rax at home, and pump a billion marines for a free win. If you are opening standard and get cannon rushed, then know that reapers are your friend if you can send them to the protoss base (this forces out 2 defensive cannons), or alternatively a well-placed bunker with a bunch of marines is very strong. Be sure to pull scv's though to force the protoss to invest into the full wall-off (because, as I mentioned earlier, an adept cannon rusher will not waste 200 minerals if at all avoidable) Print the maximum area on one Pylon is 4 if at the top of the ramp. | ||
TelecoM
United States10646 Posts
| ||
printf
United States13 Posts
On March 23 2015 06:29 Ovid wrote: Print the maximum area on one Pylon is 4 if at the top of the ramp. I was referring specifically to the values listed in the OP for the wall on overgrowth, which listed 6 as the weakness value for all 3 pylons. I may be mistaken, but I don't believe I ever mentioned walls at the top of the ramp | ||
Florencio
1 Post
User was banned for this post. | ||
WombaT
Northern Ireland23759 Posts
Truly an art form | ||
General_Winter
United States719 Posts
| ||
ZigguratOfUr
Iraq16955 Posts
On May 22 2019 10:41 General_Winter wrote: Was that a fake account or the real Florencio? The real one is semi famous as a commonly featured master level cheesier on Pig’s you tube channel. Real I think, but creating an account and bumping old threads to post links to your own Youtube channel still falls under advertising. If he wants to post his own content it's supposed to go in the External Content forum and have more content than just a link to a video. | ||
Jamesrr332
2 Posts
| ||
| ||