|
On October 19 2010 04:06 the p00n wrote:Show nested quote +On October 19 2010 03:57 Amber[LighT] wrote:On October 19 2010 03:53 the p00n wrote:On October 19 2010 03:49 Amber[LighT] wrote:
so the roach warren sits there collecting dust while 150 minerals could have went to anything more important?
Just saying your way is easily scoutable. The old way is better because you can at least chase a scout out first before shooting yourself in the face with this build. On what would you spend those 150 minerals, when you have no larvae? And you either need creep or speedlings to chase a scout away. I don't know how you scout a Z, but I don't leave my scout in his base when I have seen gas and pool go down. I retreat a bit and scout a min later or so, and there is no way you are going to deny me vision of the roach warren without delaying your roaches A LOT. Then that's your own fault for poor scouting. That has nothing to do with the build order. And drones would be first. You have the larvae for at least 2 extra. There is a definite economic advantage to waiting on the roach warren (extra drones = extra income). A roach warren that sits there and doesn't even produce an efficient amount of roaches is useless. No, you do not have the extra larvae and you cannot spend the 150 minerals on anything else (minus maybe a spine crawler which you don't need?). This build order is better. Show nested quote +On October 19 2010 03:57 Logo wrote:
You can/should delay the warren until the queen is 70% done and then the warren will pop RIGHT as the vomit larva is coming off the hatchery. There's no reason to make it any earlier. Use the extra minerals to power out whatever else you want.
Also being broke generally is good. Of course if a build is sub-optimal and you could have more stuff at the same exact time then that's a problem, but you want to be broke as it means you are making use of your entire economy.
If it's a supply thing then just extractor trick at 17 to get the queen building. The extra drone time will make up for the cost of the extractor trick.
You can just bank the extra money and use it to expand sooner once the roaches are out. I do not actually want/need criticism, I am just throwing this BO down in order to help some people and if you have any questions I will answer them, but I do not need any 'you shud delay dis lolol'-replies that are false.
even flash talks about the fact that there are STILL things he needs to work on in his build timing's and he's God Tier PLUS the game's been out for 10 years
there are always ways to improve, i'm 100% positive your way isn't perfect and even if it was, we got 2 more expansions to go
|
On October 20 2010 04:41 mOnion wrote:Show nested quote +On October 19 2010 04:06 the p00n wrote:On October 19 2010 03:57 Amber[LighT] wrote:On October 19 2010 03:53 the p00n wrote:On October 19 2010 03:49 Amber[LighT] wrote:
so the roach warren sits there collecting dust while 150 minerals could have went to anything more important?
Just saying your way is easily scoutable. The old way is better because you can at least chase a scout out first before shooting yourself in the face with this build. On what would you spend those 150 minerals, when you have no larvae? And you either need creep or speedlings to chase a scout away. I don't know how you scout a Z, but I don't leave my scout in his base when I have seen gas and pool go down. I retreat a bit and scout a min later or so, and there is no way you are going to deny me vision of the roach warren without delaying your roaches A LOT. Then that's your own fault for poor scouting. That has nothing to do with the build order. And drones would be first. You have the larvae for at least 2 extra. There is a definite economic advantage to waiting on the roach warren (extra drones = extra income). A roach warren that sits there and doesn't even produce an efficient amount of roaches is useless. No, you do not have the extra larvae and you cannot spend the 150 minerals on anything else (minus maybe a spine crawler which you don't need?). This build order is better. On October 19 2010 03:57 Logo wrote:
You can/should delay the warren until the queen is 70% done and then the warren will pop RIGHT as the vomit larva is coming off the hatchery. There's no reason to make it any earlier. Use the extra minerals to power out whatever else you want.
Also being broke generally is good. Of course if a build is sub-optimal and you could have more stuff at the same exact time then that's a problem, but you want to be broke as it means you are making use of your entire economy.
If it's a supply thing then just extractor trick at 17 to get the queen building. The extra drone time will make up for the cost of the extractor trick.
You can just bank the extra money and use it to expand sooner once the roaches are out. I do not actually want/need criticism, I am just throwing this BO down in order to help some people and if you have any questions I will answer them, but I do not need any 'you shud delay dis lolol'-replies that are false. even flash talks about the fact that there are STILL things he needs to work on in his build timing's and he's God Tier PLUS the game's been out for 10 years there are always ways to improve, i'm 100% positive your way isn't perfect and even if it was, we got 2 more expansions to go
He's not saying the BO is the best possible order, he's saying that it's better than the other 7RR. Some posts point out the difference between this BO and the old 7RR build and tell him it's less efficient. His response is something like.. "look, the timings don't lie, I've explained why it's better".
|
On October 19 2010 07:46 ltortoise wrote:
I don't really understand all this talk of denying scouting.
Is it really that important? A solid build can do some damage and pay for itself even if it's scouted and your opponent knows for sure it's coming. If you lose all ability to do damage just because they scout the Roach Warren, my only conclusion would be that the whole strategy just isn't very good.
I mean sure, denying scouting always improves the build... But is it so important that you want to do your build less efficiently?
A strong timing should be strong even if they know it's coming.
-_-
not when its something like a roach rush. Youd have a valid point in BW with mutas since they are strong regardless but......come on you have to be able to see denying scouting is important for a situation like this
|
On October 19 2010 04:30 sylverfyre wrote:
Extractor trick costs 25 minerals? More like 6. 1 for cancel penalty and about 5 for lost mining time. No more than 5.
If 1 build has the same stuff at time X and 150 MORE MINERALS than another, I think it's safe and totally factual that it's 100% superior - it's just optimized better. Deal, it's a provable fact. Provide some replays or calculated BO timing or whatever if you want to disprove. If the build has less drones or something (it doesn't) or whatever, then ok, fine, it has a disadvantage.
Its not "totally factual" and its not "superior" its been tested and proven your wrong. At most there is a 16 mineral difference between the styles.
http://www.teamliquid.net/forum/viewmessage.php?topic_id=133917
|
On October 20 2010 03:59 valckrie wrote:
10 ovie
10 pool
15 extractor
15 queen
18 warren
18 ovie
18 ovie
injecting when queen pops, and making roaches as soon as minerals allow
will get 7 roaches out by 4;40 optimal time, 4:50 real time give or take a few seconds
also, less time wasted with the roach warren idling. this gets you the fastest roaches afaik, and you might be wondering, 10 pool so early? but its not, it allows you to get the queen earlier so and times it so that when your round of injection lava pops you can make most of the roaches
After a little testing, just as I suspected, 11OL/11Pool gets roaches a few seconds faster than this, has more available larvae after building the roaches, and has slightly more resources as well.
Overpool finishes the 7th roach at just under 5 minutes (repeated tests showed times varying from 4:58-5:00), and it had 275/53 resources available at that time.
11Pool finishes the 7th roach at 4:55, had 259/65 at that time, but 12 workers on minerals will bring in about 8 minerals per second, so it would have slightly more resources 3 seconds later. It also had 2 additional larvae compared to the overpool.
Overpool is bad in this game. Period. End of story. You should never be doing it.
But using a 9OL start into a 14pool and extractor tricking to 19/18 for the queen, I was able to squeeze in 2 extra drones and only delay the first 7 roaches until 5:05, at which point I had 354/93 resources.
Anyway, you might want to consider not going overpool if your goal is to get the queen faster for larvae. Because you're wasting enough extra early larvae that it's simply not worth it. If you really want those extra seconds, go for 11Pool instead.
|
On October 20 2010 05:00 CowFu wrote:Show nested quote +On October 19 2010 04:30 sylverfyre wrote:
Extractor trick costs 25 minerals? More like 6. 1 for cancel penalty and about 5 for lost mining time. No more than 5.
If 1 build has the same stuff at time X and 150 MORE MINERALS than another, I think it's safe and totally factual that it's 100% superior - it's just optimized better. Deal, it's a provable fact. Provide some replays or calculated BO timing or whatever if you want to disprove. If the build has less drones or something (it doesn't) or whatever, then ok, fine, it has a disadvantage. Its not "totally factual" and its not "superior" its been tested and proven your wrong. At most there is a 16 mineral difference between the styles. http://www.teamliquid.net/forum/viewmessage.php?topic_id=133917
Actually, if one build has the same stuff at time X and 150 more minerals, then it is totally factual that it is superior. That's just common sense. The rest of this isn't targeted at you in any way, just suggesting you re-read the statement you're attempting to refute there. 
As for the rest of the thread:
The fact of the matter is that none of the builds people are claiming are "optimal" in this thread are actually optimal at what they're trying to do.
The original poster was posting a build that's more economic than what he typically sees, while only sacrificing a few seconds in timing, but I was able to improve on the timings of his build by simply doing 9OL instead of an extractor trick, and using an extractor trick later to go to 19/18 rather than requiring the roach warren to be put down early, making it both more economic by squeezing out 2 extra drones, *and* faster, because the 7th roach pops sooner. (because the pool gets built faster, so the queen gets built faster, so the first inject is sooner)
And that one comes from somebody who has tried to reject any criticism in the thread at all because "he knows best", even though he clearly doesn't.
And I was able to improve on the overpool build by simply not doing an overpool, getting the 7th roach out *faster* by 11pooling instead of 10pooling, and having 2 extra larvae to play with to boot. I wasn't paying attention to the times when larvae were available during that test, but it's very possible you could squeeze 2 extra drones out of that one too. Or have 2 extra larvae available for more roaches, or whatever.
Seriously, there are a lot of people in this thread claiming they know the best way to do something, but haven't even spent time optimizing their own builds. Every time I see somebody put an overpool or an extractor trick in a build, it makes me cringe a little, because I know exactly what situations you should do those in (never, and extremely rarely, respectively).
Am I claiming the changes I've made make them the best possible builds? Absolutely not. I just improved on what was posted, given the things that I know about early zerg build orders. There may very well be a more optimal way of getting 7 roaches out ASAP. But if you're willing to sacrifice economy for the fastest possible roaches, it's pretty likely that 11pooling won't be beat by anything else, and given that the gap is already so small, I'd guess that the 14pool is probably as good a tradeoff as any, giving up some time for some economy.
|
For what it's worth, I've tried the 12pool build from the other thread a few times (after changing 10OL to 9OL, which is *always* better), and can't get it to beat the (improved) 14pool by more than 2-3 seconds.
But both 12pool and 14pool will be 2 larvae ahead of the overpool version, as well as being more economically sound.
|
Sorry for sounding dumb but ...
Is Overpool going 10 OL / 12 Pool?
Or extractor-trick 11/10 Pooling?
And is the consensus extractor trick is the way to go or 9 OL or 10 OL?
|
Overpool is drone to 10, build an overlord, and build the pool on 10 before building any more drones. You really don't need to know what to call it though, because it's pretty much always bad.
9OL is the best way to go unless you're going to 11 or 12 pool *before* building the overlord. 12pool off a 9OL start gets the pool faster than a double extractor trick to 12, followed by an overlord, followed by a pool.
|
12pool off a 9OL start gets the pool faster than a double extractor trick to 12, followed by an overlord, followed by a pool.
I was under the impression that if you did a double extractor trick to 12, you would pool first then overlord? Making an overlord first then a pool at 11 is kind of silly, and as you said a 9ovie into 12p is faster than that, but its not faster than pooling first at 12/10
Overpool finishes the 7th roach at just under 5 minutes (repeated tests showed times varying from 4:58-5:00), and it had 275/53 resources available at that time.
Are you sure about this? From my experiences it comes out faster than 4;58
11Pool finishes the 7th roach at 4:55, had 259/65 at that time, but 12 workers on minerals will bring in about 8 minerals per second, so it would have slightly more resources 3 seconds later. It also had 2 additional larvae compared to the overpool.
When you say 11ovie 11pool do you mean extractor trick to 11, overlord, then extractor trick to 11 again, then pool? Also why is it 12 workers on minerals, is it not 13 as you have 18 supply [2 for queen, no units, so rest 16 drones]; unless you built a ling.
Another not totally relevant point but if you build a ling when a 14pool finishes and you want to put down roach warren at 18, most of the time if the enemy is a good scout his worker will still be in your base and will see the roach warren if you put it down asap... most of the time if you dont want him to see you'll have to delay it by a few seconds. so sometimes, imo, a ling isnt worth making
I'll give the 11Ol 11p a try sometime since you say its better than overpooling
|
On October 20 2010 23:53 valckrie wrote:Show nested quote +12pool off a 9OL start gets the pool faster than a double extractor trick to 12, followed by an overlord, followed by a pool. I was under the impression that if you did a double extractor trick to 12, you would pool first then overlord? Making an overlord first then a pool at 11 is kind of silly, and as you said a 9ovie into 12p is faster than that, but its not faster than pooling first at 12/10
That's exactly what an overpool is, making the overlord first then making the pool. And yes, 12pool building the pool first will get the pool faster, but you waste a *lot* of larvae spawn time that way, and it turns out that at least for getting 7 roaches out asap, it doesn't work as well as many of the other starts, because it leaves you short on money.
On October 20 2010 23:53 valckrie wrote:Show nested quote +Overpool finishes the 7th roach at just under 5 minutes (repeated tests showed times varying from 4:58-5:00), and it had 275/53 resources available at that time. Are you sure about this? From my experiences it comes out faster than 4;58
Quite sure. Exact timings might be different on different maps (I was using YABOT Blistering Sands), but I tested *many* times with each build. Note that was the finish time for the 7th roach though, and every build I tried except for a 14pool squeezing in extra drones had to wait on resources to build the last few roaches, not larvae. Which, by the way, means that if 7 roaches is the target, you can delay the initial parts of the build to get a better economy, which is why the 14pool timings are so much closer to all the other timings than you'd expect. You just have to choose what you're optimizing, the first roach or the seventh. If you're not going to move out until the 7th one is done, then it makes more sense to optimize for that.
On October 20 2010 23:53 valckrie wrote:Show nested quote +11Pool finishes the 7th roach at 4:55, had 259/65 at that time, but 12 workers on minerals will bring in about 8 minerals per second, so it would have slightly more resources 3 seconds later. It also had 2 additional larvae compared to the overpool. When you say 11ovie 11pool do you mean extractor trick to 11, overlord, then extractor trick to 11 again, then pool? Also why is it 12 workers on minerals, is it not 13 as you have 18 supply [2 for queen, no units, so rest 16 drones]; unless you built a ling.
There's no "extractor trick to 11 again" since the drone is still there? But yes, extractor trick to 11, overlord, then pool. It is pretty much *always* going to be better to do that than to overpool. As for worker count, I must have counted wrong, or missed one somehow. You're right, it should be 13 workers.
On October 20 2010 23:53 valckrie wrote:
Another not totally relevant point but if you build a ling when a 14pool finishes and you want to put down roach warren at 18, most of the time if the enemy is a good scout his worker will still be in your base and will see the roach warren if you put it down asap... most of the time if you dont want him to see you'll have to delay it by a few seconds. so sometimes, imo, a ling isnt worth making
I agree that the ling timing is cutting it *very* close, and that unless you chase the worker off *immediately* after the lings spawn, you're going to end up either delaying the warren or placing it in sight of the scout, but that's consistent no matter how you go about it, because the time you need to place the warren depends completely on the queen timing (if you don't have the warren down by the time the queen is 60-70% complete, the warren won't finish by the time your first inject is done), and obviously since the queen and the lings both depend on the pool, there's no way to get the lings faster. Still, though, it does help to not be scouted, and the scout will often leave immediately upon seeing the first two lings.
|
Skrag, thanks for all the in-depth testing! When you say you are extractor-tricking your queen on 17, does that mean you are still getting your first extractor at 15 and tricking the 2nd extractor? And then warren on 18?
|
On October 21 2010 03:02 BadassHamster wrote:
Skrag, thanks for all the in-depth testing! When you say you are extractor-tricking your queen on 17, does that mean you are still getting your first extractor at 15 and tricking the 2nd extractor? And then warren on 18?
Yes. Well, technically you're putting the warren down on 19, and because you've tricked to 19/18 rather than relying on building the warren to drop to 17 so you can build a drone back to 18, you can delay the warren as long as you feel you need to, in order to avoid being scouted. If it doesn't go down by the time the queen is about 70% done though, the warren won't be finished in time for the first inject, so you can't delay too long.
However, when I was testing 14pool, I built drones up to 17, then tricked to 19 for the queen. Doing this allows you to squeeze in an extra drone before the queen (because you're not building the warren with it), and one more drone after your next overlord finishes, which leaves you with enough minerals to expand right about the same time your 7th roach pops out. If you do want to get a set of lings to chase off the scout, you have to cut one drone, building up to 16 rather than 17, then you can extractor trick to 19/18 for the lings. That still would be better than placing the warren early, because the worker that would place the warren is still mining, and you don't give away your plan any earlier than you have to.
The full build order would be this:
Without a set of lings:
9OL
14 pool
15 extractor
Drone to 17
Build the second extractor just before the pool finishes so you have enough supply to start the queen, then cancel the extractor
19 overlord
19 overlord
Drone as soon as overlord pops (you can make this 2 drones if your timing is *perfect*, but if you're not perfect, you risk delaying the 7th roach while waiting for a larvae, putting the 7th roach a full 15 seconds behind the other 6. I only managed to squeeze that drone out without delaying the 7th roach about a third of the time)
Roach warren when queen is about 70% done
Warren and inject should both finish about the same time your hatch has 3 larvae. Those 3 + the 4 from the inject = 7 roaches.
To get a set of lings to try to chase off the scout, you'd need to stop droning at 16, still build an extractor before the pool finishes, and build a queen+lings before canceling the extractor. You do have a little bit of time between when the lings spawn and when you have to place the warren.
|
I did this BO, but the opponent fought me off with two stalkers, two zealots, and an immortal by the time it arrived. It feels totally all-in because you're on one base after the attack dies.
|
So after all the debate.. what is the consensus on the most optimal build for the 7RR?
|
On November 08 2010 23:00 DeCoup wrote:
So after all the debate.. what is the consensus on the most optimal build for the 7RR?
It depends on how much economy you want to sacrifice to get the 7th roach out slightly faster, and whether you want to have a chance to chase off an enemy scout before placing the warren.
There isn't really an "optimal" build for anything. There are always tradeoffs, and you have to decide what your primary goal is.
|
The optimal build is not doing gimmicky all ins and learning how to play so when you face better people you stand a chance.
|
On November 09 2010 03:18 MorsCerta wrote:
The optimal build is not doing gimmicky all ins and learning how to play so when you face better people you stand a chance.
"Optimal" depends upon the situation, as said previously. A macro build is sub-optimal to early pressure, a turtle build is sub-optimal to a macro build, an aggressive build is sub-optimal to a turtle build, etc. Yes, the optimal build is catered to the player and the situation but that doesn't make this build any less optimal than a 14 hatch 15 pool. It is all situational.
I have won and lost because I chose the 7RR, but that is no different than any of the other builds I've done including the "standard" builds. Even "gimmicky" builds, as you call them, have their place for noobs and pros alike. I think we all understand that a build's viability depends on the skill of the players and the state of the game. What works today may not work tomorrow and what works for me may not work for you.
|
On October 19 2010 03:53 the p00n wrote:Show nested quote +On October 19 2010 03:49 Amber[LighT] wrote:
so the roach warren sits there collecting dust while 150 minerals could have went to anything more important?
Just saying your way is easily scoutable. The old way is better because you can at least chase a scout out first before shooting yourself in the face with this build. On what would you spend those 150 minerals, when you have no larvae? And you either need creep or speedlings to chase a scout away. I don't know how you scout a Z, but I don't leave my scout in his base when I have seen gas and pool go down. I retreat a bit and scout a min later or so, and there is no way you are going to deny me vision of the roach warren without delaying your roaches A LOT.
I completely agree it just isnt going to happan i can scout you 30 later and if it isnt down they im not worried about your 7 roaches shambling across the map at an incorrect push timing.
|
The title of this thread is like nails on a chalkboard to me, as is the OP's attitude.
|
|
|
|
|
|
EPT
