|
If anything, this thread shows that almost any opening can be an "economics" build. The differences are so meaningless that it becomes purely a matter of preference/safety/practicability.
The initial debate was driven by the "hatch first cheese" thread and zerg claiming that hatch first and pool first don't even compete in the same leagues as far as economics goes. This has been proven false. I'm not going to buy it even if IdrA says that the extra 50mineral you get from a hatch first opening over 6minutes is the determining factor to say who ggs/ragequits first.
The only argument that remains is safety: early pool gives you more flexibility to stop really quick cheese, early hatch gives you a better opportunity at having a chance to lay down spines in time to stop a later bunker+scv rush at your nat. It comes down to determining which build is safer, on which map, against which strategy!
|
jdseemoreglass could you please take another look at 14 pool 16 hatch? after seeing your new graphs I went ahead and played on until 7 min too and this is from my first try (so might be some small mistakes) I did NOT send my drones to individual mineral patches, just hatch waypoints.
![[image loading]](http://www.gamereplays.org/community/uploads/repimgs/repimg-33-170522.jpg)
BO:
+ Show Spoiler +
9 overlord
14 pool
16 hatch
16 queen (inject and send to natural)
18 overlord
extractor trick
22 queen
24 overlord
32 overlord
35 overlord
44 overlord
50 overlord
Results:
+ Show Spoiler +
Drones:
- 4:30 -- 5:00 -- 5:30 -- 6:00 -- 6:30 -- 7:00
21.29 - 28.53 - 35.76 - 41.7 - 52.59 - 61.88
Resources:
4:30 -- 5:00 -- 5:30 -- 6:00 -- 6:30 -- 7:00
3029 - 3469 - 4049 - 4709 - 5454 - 6269
basically it's behind all other BO's at 4:30, but at 5:00 it's ahead in drones and stays ahead with the exception of 6:00's 11 pool.
at 5:00 it's ahead of 11 pool in resources and at 6:00 it overtakes 16 hatch 15 pool and is in the end only SLIGHTLY behind 14 hatch 15 pool in resources which is so minimal it can be luckbased...
|
On December 02 2010 01:34 Dakaru wrote:jdseemoreglass could you please take another look at 14 pool 16 hatch? after seeing your new graphs I went ahead and played on until 7 min too and this is from my first try (so might be some small mistakes) I did NOT send my drones to individual mineral patches, just hatch waypoints. BO: + Show Spoiler +
9 overlord
14 pool
16 hatch
16 queen (inject and send to natural)
18 overlord
extractor trick
22 queen
24 overlord
32 overlord
35 overlord
44 overlord
50 overlord
Results: + Show Spoiler +
Drones:
- 4:30 -- 5:00 -- 5:30 -- 6:00 -- 6:30 -- 7:00
21.29 - 28.53 - 35.76 - 41.7 - 52.59 - 61.88
Resources:
4:30 -- 5:00 -- 5:30 -- 6:00 -- 6:30 -- 7:00
3029 - 3469 - 4049 - 4709 - 5454 - 6269
basically it's behind all other BO's at 4:30, but at 5:00 it's ahead in drones and stays ahead with the exception of 6:00's 11 pool. at 5:00 it's ahead of 11 pool in resources and at 6:00 it overtakes 16 hatch 15 pool and is in the end only SLIGHTLY behind 14 hatch 15 pool in resources which is so minimal it can be luckbased...
One problem I have with this build is that you skipped the final overlord which all other builds produced. This leaves you supply capped at 68/68 at the end of the replay with no overlord in sight. This can account for an increase in at least a drone.
I find it unlikely this build will surpass the others, but I will certainly test it again and post the new results later today.
|
yeah I forgot that but really it matters only 1 drone because everything is saturated already and the limiting factor is the larva... also I had 2 larva spawns remaining at 7 min mark because I was supply capped so that would increase my drone count slightly too
either way even with a whole drone less the build is still ahead in drones of all the others by nearly another drone
|
jdseemoreglass, I really think you should include the corrected partial drone counting once queens have appeared. See earlier posts by myself and Hyrkul. If you disagree we'd obviously appreciate the reasoning.
|
On December 02 2010 02:25 Dragar wrote:
jdseemoreglass, I really think you should include the corrected partial drone counting once queens have appeared. See earlier posts by myself and Hyrkul. If you disagree we'd obviously appreciate the reasoning.
No one disputes that different builds will have queen spawn larvae at different times. The point is that all the builds will necessarily produce them at the same RATE beyond a certain time. Showing the data over time provides a visual representation of where each build is in the spawn larvae cycle, but the discrepancy overall in the drone count will not change in the long run, only in short time frames.
|
Try 14hatch/14pool. It does get larvae-blocked for a short time while waiting for minerals for the pool, but might make that up later with faster queens. If there's enough of a larva advantage, it might be worth a slight reduction in total mining.
|
I played two games today on diamond ladder (2200+) using the 11pool and won each of them easily.
The first game illustrates how flexible this build is in reverting to an all-in attack. I scout my opponent opening hatch first, so that is exactly what I go for. Going hatch first or 14 pool would prevent this sort of pressure, and would potentially negate what is for me an easy victory.
![[image loading]](http://www.gamereplays.org/community/uploads/repimgs/repimg-33-170544.jpg)
Game two illustrates how well this build can hold up against a 4gate. This felt like the easiest 4gate I have ever held off because I had many units out quickly. At first I assumed my opponent was making a mistake, or was possibly going 3gate-expand, but the replay shows he did fairly well in keeping his resources low and drone count high.
After holding off two attacks I prepare for a mid-game timing attack before his late expansion can reap dividends.
![[image loading]](http://www.gamereplays.org/community/uploads/repimgs/repimg-33-170545.jpg)
|
Thx for the rep, the 4 gat is reaped so badly
|
I've been running evo chamber for 9 hours on a quad core and it couldn't find a build that gives 46 drones at the 6 minute mark. Builds that give 45 drones at the 6 minute mark are easy to produce. The program stagnated on
10 OL
Etrick
11 pool
16 queen
18 hatch
18 OL
20 OL
20 queen
All I really wanted to say is that the builds listed here may be the best possible builds and they are likely limited by this "45 drone ceiling" at the 6 minute mark. I'll publish a graph soon to help understand how the time limit influences this experiment.
|
On December 02 2010 03:40 ubershmekel wrote:
I've been running evo chamber for 9 hours on a quad core and it couldn't find a build that gives 46 drones at the 6 minute mark. Builds that give 45 drones at the 6 minute mark are easy to produce. The program stagnated on
10 OL
Etrick
11 pool
16 queen
18 hatch
18 OL
20 OL
20 queen
All I really wanted to say is that the builds listed here may be the best possible builds and they are likely limited by this "45 drone ceiling" at the 6 minute mark. I'll publish a graph soon to help understand how the time limit influences this experiment.
Has anyone tried silly openings like double expand yet?
9 Overlord
16 Hatchery #2
15 Spawning Pool
17 Overlord
18 Hatchery #3
20 Queen
24 Queen
26 Overlord
28 Queen
30 Overlord
36 Overlord
40 Overlord
To minimize travel time:
- transfer 1 drone and rally main hatch to hatch#2 until both have 16 then rally both to hatch#3
- hatch#3 not rallied
If there's one build that allows you to make more than 45 drones at 6:00, it should be something like this.
|
Ok, so my testing does seem to indicate that 14hatch/14pool does have a larva advantage of around 2-3 over 14h/15p, which I think would put it as the clear larva leader overall, and that larva advantage comes at the cost of a very small resource loss when compared to 14h/15, somewhere around 30-40 minerals worth.
The actual order is pretty much identical to the posted order for 14h/15p, and in some cases the overlord timings line up much more nicely anyway. For example, the overlord on 17 feels pretty early, and I *think* you could get away with waiting til 18 to build the overlord with 14h/15p, getting an earlier drone, but OL on 17 is necessary for 14h/14p.
I'm measuring things differently than the OP though (I still don't understand where the "resources mined" number is coming from, and I'm also taking the queen spawn larvae progress into account for drones), so having me post numbers isn't going to be particularly useful.
|
Personally I'm loving the 11 overpool 18 hatch build, played some games with it last nite and it worked out well. Always wanted a pool that early but assumed I was at a massive eco disadvantage. Even if its slightly behind the other front runners I'll take it.
|
agree with eladmir.
whilst there may be the other build that's slightly higher on econ, there's really no price tag on how much safer the 11pool is, and how much more variable you can be with it.
I've been loving it :D.
In the end it will just be a case of looking at the highest econ builds, and working down the list untill you find a sensible game viable one. So far thats the 11overpool :D
|
On December 02 2010 05:56 Scrimpton wrote:
agree with eladmir.
whilst there may be the other build that's slightly higher on econ, there's really no price tag on how much safer the 11pool is, and how much more variable you can be with it.
I've been loving it :D.
In the end it will just be a case of looking at the highest econ builds, and working down the list untill you find a sensible game viable one. So far thats the 11overpool :D
I agree with everyone on this. In fact, I think I may start a new thread on the potential new zerg standard in all match-ups. I really can't see any reason or justification for doing any other build yet, but more testing in-game is clearly needed.
One thing that should be analyzed is the "current resources" of this build compared to the other. I suspect this build may be more fragile to variations in unit production with regard to delaying the ordering.
|
On December 02 2010 03:40 ubershmekel wrote:
I've been running evo chamber for 9 hours on a quad core and it couldn't find a build that gives 46 drones at the 6 minute mark. Builds that give 45 drones at the 6 minute mark are easy to produce. The program stagnated on
10 OL
Etrick
11 pool
16 queen
18 hatch
18 OL
20 OL
20 queen
All I really wanted to say is that the builds listed here may be the best possible builds and they are likely limited by this "45 drone ceiling" at the 6 minute mark. I'll publish a graph soon to help understand how the time limit influences this experiment.
If you are talking theoretical maximum (from evo chamber) I got 48 drones in under 6 minutes using
10 Extractor Trick
11 Overlord
11 Hatchery
12 Spawning Pool
14 Overlord
15 Queen, then Constant Spawn Larvae
17 Queen, then Constant Spawn Larvae
19 Overlord
27 Extractor Trick
29 Overlord
34 Extractor Trick
36 Extractor Trick
38 Overlord
51 Drone
Though I did use the GUI version as it runs much faster - my understanding is there shouldn't be a time diff between the two. Not sure on how many minerals you'd have at this point but it would probably be close to the others. However, I don't think evo chamber takes into account travel times and such so no one would actually be able to realize this directly. Not very useful but it does pass the 46 drone limit you were talking about.
|
On December 02 2010 06:14 jdseemoreglass wrote:Show nested quote +On December 02 2010 05:56 Scrimpton wrote:
agree with eladmir.
whilst there may be the other build that's slightly higher on econ, there's really no price tag on how much safer the 11pool is, and how much more variable you can be with it.
I've been loving it :D.
In the end it will just be a case of looking at the highest econ builds, and working down the list untill you find a sensible game viable one. So far thats the 11overpool :D
I agree with everyone on this. In fact, I think I may start a new thread on the potential new zerg standard in all match-ups. I really can't see any reason or justification for doing any other build yet, but more testing in-game is clearly needed. One thing that should be analyzed is the "current resources" of this build compared to the other. I suspect this build may be more fragile to variations in unit production with regard to delaying the ordering.
Despite the fact that it probably seems I'm fighting to have 14hatch still be considered the "best" build, I agree as well. I think it's totally awesome that 11overpool/18hatch is close enough economically to the best true economic builds that you can use it in all but the absolute safest of situations. I'll probably hatch-first on shakuras, jungle basin, and maybe metal (although even there it's nice to know that 14h/15p or 14h/14p are both better than 15h/14p which is what I've been doing), but everywhere else it's almost certainly going to be 11overpool for me from now on.
|
|
|
|
|
I counted each mineral node at the end of the 11 pool 18 hatch replay and came up with 20065. Subtracted from the 24000 minerals from 16 nodes at 2 bases, that comes out to 3935 minerals mined. Where does 4620 come from? Same for the 14 hatch 16 pool Dakaru posted. Why are the results off by 600-700 minerals if you have a replay.
|
|
|
|
|
|
EPT
