|
Here's the original thread for those who are interested:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=118754
You can show logically that 9ol is the best without needing too many calculations.
On April 08 2010 22:04 Klive5ive wrote:
Ok let's sort this out then.
Here's my measurements from replays, feel free to check this:
With 9 drones it takes 14 tics to mine the extra 100 minerals needed to get an overlord on 9.
Drones mine at 5 minerals every 7 tics
Extractor trick costs effectively 10 minerals (it's actually 6 but you can't spend the extra 4 unless you cancel another building).
10o
This build wastes larvae time, setting all your future drones back. The result is less minerals in the long run.
However at the 10 larvae stage this build has the most minerals. Therefore it will get a 10 pool down faster.
If you plan a fast pool or something crazy like a 12hatch this would be the correct build.
10o+e Vs 9o
Neither of these builds waste larvae time.
9o gets it's 10th drone 100 minerals later. Effectively that's about 14 tics. 1 drone would mine 10 minerals in that time.
The 10o+e loses an effective 10 minerals on the extrick so it's immediately equal. Factor in the lost mining time on the extrick drone and the later 12th drone and you can see this build is worse.
Conclusion:
9o is the strongest economic opening for Zerg. This should always be done before the popular 15pool 16hatch.
|
Other people concluded in that thread that 10OL is still good for overpool.
I just want to make it abundantly clear that while overpool might have been a viable option in BW, it is NEVER a good idea in this game, because extractor tricking for an 11 pool gets the pool *faster* than overpool, is more economic, and I'm fairly sure it wastes fewer larvae too.
Which, of course, means that 10OL is never a good option under any circumstance. If you happen to forget to build your OL on 9, go ahead and build it on 10 and then extractor trick. That will put you as close to 9OL as you can be.
|
Thanks for all this. I assumed that 10 OL/ET was done to optimize the larva usage (when I do 9 OL I find I have 3 larvae for about a second or so, but it seems that's because I'm not quite as quick as the computer AI). Regardless, great guideline for the various build orders based on when the spawning pool comes out, I've been looking for this kind of breakdown!
|
I have tested my builds on the slowest speed (so i can manualy do everything almost as fast as the computer can). I vastly perfer the 10-T-OL build, the reason is that I like to scout with my 11th or my 13th drone, depending on if its vr terran 11 drone, or protoss/zerg 13 drone.
Terran is an 11 drone, so I can do more damage to the terran eco by forcing him to pull 2 more scvs early to stop my drone from killing the scv building the rax, and I feel that 2 scvs pulled off minerals is more worth my one drone being a little earlier. As well as having a quick gas steal (before he can build 2 refineries) after harrassing the rax is over.
Zerg and protoss dont change all that much till the middle game usualy, so i can send a slightly latter scout, but still early enough to catch any silly 6/8/11 pools, as well as getting a scout to protoss quick enough to tell me if I can safely FE at 16 or so (1 gateway in far positions is clear indication that a FE is good before pool, 1 gateway in close positions and you can do it very quickly after the pool, otherwise you need to stay on one base for a while).
So all in all given that a 10-T-OL might be 1 mineral worse on avr if you dont scout early (and my testing showed it EXACTLY the same not 1 mineral worse then a 9 OL). I will still choose the 10-T-OL VASTLY over a 9 OL if your doing a early scout.
Clearly this isnt as good an idea on something like scrap station or LT if you scout him in the close air distance location. The early drone scout isnt needed as much (esp against zerg or protoss), and the 9OL is probably better. Although an early drone scout can still harrass vr terran which can be good enough to still send it.
|
9 OL seems like you have less drones stock up larva early on, and 10 lord with extractor trick uses the larva the most efficiently. the mineral difference between 9 and it are negligible so you shouldn't use 9 then.
Did op even take larva into consideration?
|
With 9ol you get the 12th and 13th drones faster than the 10ol et, and 9ol doesn't have any lost larva like I think your suggesting CHarlieMurphy.
|
This study needs to take into account the worker scout.
I would imagine that having an early scout makes a much larger economic impact on 9 OL as opposed to 10 OL trick as you lose 1 of 10 drones with 9 OL and you lose 1 of 11 drones with 10 OL.
|
On July 21 2010 13:43 obsid wrote:
So all in all given that a 10-T-OL might be 1 mineral worse on avr if you dont scout early (and my testing showed it EXACTLY the same not 1 mineral worse then a 9 OL). I will still choose the 10-T-OL VASTLY over a 9 OL if your doing a early scout.
1 mineral isn't 1 mineral, it's 5. You can't spend 4 minerals at any point in the game. Also, how many times did you test on slow? Because I tested far many more times than I'd care to count, both letting the AI do the work, and doing it myself on normal speed (slow was too slow since I was letting them run longer, and I could easily do everything I needed to, including a perfect 3/3 split, on normal), and 10-T-OL was very consistently 10 minerals behind.
But even if it's just 1 mineral, behind is behind is behind, so unless you're doing something very specific with that 11th drone, then why pay a price to do a more risky build (misclicks can cost you a LOT here)?
As for scouting, if you insist on scouting at 11, and that exact timing is useful and meaningful, then sure, extractor trick. But 9OL will get the 13th drone almost 10 seconds before 10-T-OL, so if you still choose it vastly for that reason, you're doing something wrong IMO.
|
On July 21 2010 14:06 CharlieMurphy wrote:
9 OL seems like you have less drones stock up larva early on, and 10 lord with extractor trick uses the larva the most efficiently. the mineral difference between 9 and it are negligible so you shouldn't use 9 then.
Did op even take larva into consideration?
A single extractor trick doesn't waste any larvae spawn time. 9OL wastes less than a second of larvae spawn time, while you're waiting for minerals after building your overlord. Every other build I've tried, including the double-extractor trick overlord that TLO claims to do for larvae management, wastes quite a bit more than that.
Is less than a second worth of larvae worth 10 minerals, which represents about 15 seconds worth of worker mining time?
|
So... I have a fairly unique BO and am wondering how the best way to go about testing the economic impact would be.
I do:
Extracto-drone 11
Spawning pool
OL
extracto-drone 11
12 & 13 when OL pops
14 when larva pops
Queen/ling (lings do my scouting around the 3:30 mark)
Ling
Extractor
Drone
OL
Spawn larva when queen pops
2nd queen when OL pops
Power drone to 26
Yes, a lot of drawn out steps, but very cheese resistant, allows for some early aggression, and getting spawn larva so early let's me catch up in drone numbers to more standard builds.
I know I have lost minerals doing this build, but how many?
|
this is very helpful, thank you for the information.
I'm going to start using 9ol, and I know I'm going to mess it up and go for 10 ol-extractor trick and then try to switch back and mess myself up big time with my opener >.>
But that will only last for the first few days, or if I'm really drunk.
Thanks for the research, much appreciated.
Repped
|
I have to agree that your testing proves conclusively that 9ol is superior to 10ol purely in terms of minerals mined for the rest of the game. It doesn't however take into account when you get those minerals and the fact that at certain points in the early stages zerg really can't spend all their minerals even if they have them.
For example, from my own testing doing double ET, 12 ol, 12 pool you will be able to do a 12 pool without wasting larvae, which to my knowledge is the fastest possible pool without permanent economic damage(basically, as opposed to other races zerg only loses cumulative income if they are larvae blocked, cutting drones will not permanently delay other drones unless it wastes larvae).
This means that your queen spawns roughly 10-15 seconds earlier than other "early" pool timings that do not waste larvae(13-14). This can help tremendously against certain ultra fast rushes, such as 6 rax proxy reaper or 6 pool. Not that these builds are hard to stop regardless, but it makes stopping them a lot less of a hassle. You can from there either get a really quick ling force, or power drones like any of the "economic" builds thanks to that early larvae pop.
So while your testing is very much appreciated, it doesn't paint the entire picture as well as you make it sound like(although for most builds, it's very informative). Especially since you say you can't get the AI to properly double extractor trick.
Edit: Posted before fine tuning my post due to stress, so some things may have changed if you read this post just as I originally posted it. Sorry about that.
|
I checked BO's on last days of phase2 from several best known zergs and all seemed to be doing 9OL now. I did think until then that gastrick (after OL ofc) was slightly ahead, but have started switching to 9OL now. Last few days I tried to remember this, but often noticing my brains autopiloted myself to always 10/10 before OL  Oh well, training training required.
|
On July 22 2010 21:37 Alsn wrote:
I have to agree that your testing proves conclusively that 9ol is superior to 10ol purely in terms of minerals mined for the rest of the game. It doesn't however take into account when you get those minerals and the fact that at certain points in the early stages zerg really can't spend all their minerals even if they have them.
For example, from my own testing doing double ET, 12 ol, 12 pool you will be able to do a 12 pool without wasting larvae, which to my knowledge is the fastest possible pool without permanent economic damage(basically, as opposed to other races zerg only loses cumulative income if they are larvae blocked, cutting drones will not permanently delay other drones unless it wastes larvae).
Check out the spawning pool timings on the second page. I tested various ways of pooling at different supply amounts (6 pool all the way through 14 pool), all manually because I couldn't get the AI to double-extractor efficiently, but it was all done on normal speed instead of faster so that there were fewer opportunities to screw up, and doing each build numerous times.
I also did separate testing on how much larvae time was wasted by each build. As I already mentioned, 9OL wastes a very small amount of time, where single extractor tricks don't waste any. But a double extractor trick into overlord then pool does waste more larvae time than 9OL by a few seconds.
In addition, 12 pooling with 9OL gets the pool slightly *faster* than double-trick-OL (and by slightly, I mean *very* slightly, because it's only by about 3/4 of a second on average), and is better economically.
So if you're 12 pooling but building an overlord first, you should *never* double extractor trick.
The only reason you should ever double extractor trick IMO is if you're going to build the pool first then the overlord. Otherwise, you're sacrificing minerals for absolutely zero benefit.
|
On July 22 2010 21:45 Ouga wrote:I checked BO's on last days of phase2 from several best known zergs and all seemed to be doing 9OL now. I did think until then that gastrick (after OL ofc) was slightly ahead, but have started switching to 9OL now. Last few days I tried to remember this, but often noticing my brains autopiloted myself to always 10/10 before OL Oh well, training training required.
Heh, yeah, I still find myself forgetting to OL at 9 sometimes. The good news is that if you build the overlord on 10, and then extractor trick, you only end up very slightly behind 9OL, so its a good fallback if you forget the OL. 
|
You should always do 9OL unless your BO requires you to do extractor tricks (very fast exp/pool)
|
Hi, you got the replays/maps? That would be cool to see firsthand.
|
So I've always built my OL when I'm at 10worker count, done the ET, and when the OL pops I make my 2 drones right away, then another when the larva is out.
So if I'm reading this correctly, the superior build to this is actually to get to 9 supply, save to 100minerals for the OL,(while still at 9 supply), THEN get my 10th drone?is there time to do the ET before th OL pops, still giving you 11/10? then the OL pops, and you'll have 100 minerals ready to instantly make another 2 drones?
so...forgive my for being baked and unable to comprehend, and dumb this shit down for me
I go
9 OL
drone for 10
ET
OL pops
drone,drone
then shortly after there's a third larva for another drone, but the third larva is quicker than my 10 OL, ET, double drone when OL pops, and another drone when the 3rd larva appears?
I have this correct?
9OL is superior build?
sorry i'm retarded right now, I'm just making sure I'm understanding this
|
nm I just now saw the post with the link and quote at the top of page three
DERP
thanks ^^ lol
|
"Computer players do a pretty dang near-perfect 3/3 split (a lot better than any 3/3 split I've ever seen in a replay)"
Have you ever looked at really early APM for the computer? I've seen them hit 400 before, it's insane.
|
|
|
|
|
|
EPT
