EPTOn December 03 2010 11:27 TheNomad wrote:
Ok make overlords take up 1 or 2 supply.
Ok make overlords take up 1 or 2 supply.
.. What?
[D/H] The end of Supply Depots? - Page 11
| Forum Index > StarCraft 2 Strategy |
|
Deadlift
United States358 Posts
.. What? | ||
|
Yoshi Kirishima
United States10292 Posts
And LOL did NOT know that CCs give you 11 food! Wow o.o and haha i saw that 4 OC 200/200 13 minute push build, gonna try it out when I can xD | ||
|
Xapti
Canada2473 Posts
On December 03 2010 11:12 30to1 wrote: The mining rate of an SCV depends on saturation. You're right that if each SCV has its own patch, a mule is only worth 4scvs and change, if you are comparing saturated SCVs - it ends up being closer to between 5 and 6. The simple proof being that 3 SCVs saturate a patch - and 8 patches give you ~800 m income. An OC can generate very close to 200m (you have ~5 seconds with 2 mules). Meaning that each OC is actually almost exactly equal to 6 SCVs - each mule individually is worth like ~5.7 scvs saturated or ~4.5 scvs unsaturated. It's possible that what you're saying can be considered right, but in my opinion it is flawed logic. The way I look at it is that MULEs aren't getting worth more SCVs when there's more saturation, the SCVs are just loosing value once they don't optimally mine. There will be periods where more than optimal numbers of SCVs accumulate, but generally this is for better mining upon expansion. I am just taking 3rd party numbers here, but assuming mule averages 170 mn/min (as opposed to 160 or 180), and that SCVs mine 42 mn/min then that makes a MULE equal to 4.04 SCVs (until 16 SCVs) not 4.5. Past 16 SCVs, at 170 average mining rate (which is best comparison because you can't stack MULEs on one field when others are nearby, and even if you could it would get mined out eventually), that is 4.8 times less than the 816 minerals per minute 24 workers would mine. 24/4.8 is 5 SCVs, so MULE would never be worth more than 5 SCVs unless one counts extra SCVs that are useless for mining. 2 miners per patch is 80% max mining rate (some people call it efficiency but that is quite mistaken), 100% cost efficiency, 3 miners per patch is 100% max mining rate, 80% efficiency. (this is not true efficiency though, since it's comparing costs with rates, and costs no matter how large that generate rates of growth are always efficient, just take longer to be beneficial) | ||
|
JBrown08
Canada306 Posts
| ||
|
DiracMonopole
United States1555 Posts
| ||
|
michaelhasanalias
Korea (South)1231 Posts
On December 03 2010 10:39 TheOracle wrote: Just wow. I can't believe this is viable. Have to try this sometime lol. On a serious note, i'm curious as to how fast this mines out your expo and main. It's fine to have a 200/200 army earlier than them, but if you are almost mined out you can't follow through. I'd want to know how fast you can remax afterwards to continue to throw stuff at them. (Both assuming fresh bases and without fresh bases). If you can get a second or even third wave going off your 2 bases, this could be incredible. I've been averaging between 10 minutes and never on my main. It depends how quickly you can secure a muling station. If you can hold any expansion for 90 seconds, you can pick it clean late game. I've mostly been playing this in 4's, but the ability to just do things like deny expansions... have an island cleaned out before the 15 minute mark of a game, etc... it's just ridiculous. And you'd be surprised how fast a 170 stimmed 3/3 CS ball of marines can annihilate a late-game base if the enemy's army pushes out in a different direction. 170 marine deaths? No problem, use your 30 barracks to make a new clone army in about 2 minutes... And I wouldn't even consider myself "good" at utilizing this build yet. The people who speak of just adding this on top of regular builds when the resources inevitably stack up are the ones to really listen to. And when you attack, you don't need to expand, but you definitely should make another OC that will have paid for itself by the time the next big fight happens. | ||
|
dahorns
21 Posts
On December 03 2010 11:43 Xapti wrote: It's possibly what you're saying can be considered right, but in my opinion it is flawed logic. The way I look at it is that MULEs aren't getting worth more SCVs when there's more saturation, the SCVs are just loosing value once they don't optimally mine. There will be periods where more than optimal numbers of SCVs accumulate, but generally this is for better mining upon expansion. I am just taking 3rd party numbers here, but assuming mule averages 170 mn/min (as opposed to 160 or 180), and that SCVs mine 42 mn/min then that makes a MULE equal to 4.04 SCVs (until 16 SCVs) not 4.5. Past 16 SCVs, at 170 average mining rate (which is best comparison because you can't stack MULEs on one field when others are nearby, and even if you could it would get mined out eventually), that is 4.8 times less than the 816 minerals per minute 24 workers would mine. 24/4.8 is 5 SCVs, so MULE would never be worth more than 5 SCVs unless one counts extra SCVs that are useless for mining. If the numbers from haploids production calculator are accurate, 1 mule mines as quickly as 4 scvs for the first 16 scvs mining in a standard 8 patch. After 16 scvs, the mule mines as quickly as the next 8 scvs (assuming again a standard 8 patch). | ||
|
eth3n
718 Posts
rofl, clearly terran involves more than just spamming mules and MMM ball, they truly know their foes in and out | ||
|
30to1
105 Posts
On December 03 2010 11:43 Xapti wrote: It's possibly what you're saying can be considered right, but in my opinion it is flawed logic. what i'm saying is right. I stand corrected though with ~4.5 unsaturated. My tests indicate scv rates to be roughly 40/35/25 in terms of decreasing returns on scv saturation (these were not exhaustive, but I trust them more than your numbers). At 16 workers then SCVs would generate 75m/ ms per patch or 37.5 per worker. So according to this, you're right in that at unsaturated levels a mule isn't worth ~4.5 scvs --- its actually worth ~4.8. edit: cleaned up last sentence. | ||
|
prakkus
13 Posts
The build in this thread: http://www.teamliquid.net/forum/viewmessage.php?topic_id=173703 looks like it'd be a blast. | ||
|
Xapti
Canada2473 Posts
On December 03 2010 12:09 30to1 wrote: My tests indicate scv rates to be roughly 40/35/25 in terms of decreasing returns on scv saturation (these were not exhaustive, but I trust them more than your numbers). At 16 workers then SCVs would generate 75m/ ms per patch or 37.5 per worker. You're telling me you're getting reduced mining rates when going form 8 miners to 16 miners per patch? Doesn't make any sense to me, because workers do not have to wait for each other when there's only 2 so they should not be any slower. I can't say those stats aren't wrong though. Values I heard was 42 mn/min up to 16 miners on liquipedia, but it is possible that information is not accurate. Secondly I don't know if you're just not explaining yourself, or if you misunderstood what I was saying about false logic. When I said false logic, I didn't mean your math/stats is wrong, that was a separate issue. What I was saying is that you shouldn't valuate a MULE compared to at the worst efficiency number of SCVs. In certain situations MULEs will be more valuable to a player than the standard base valuation, but that doesn't mean it should be considered that for all cases. I don't really know why I said that though, since you never truely valuated a MULE at only the higher value, just range between the lower an higher, which is acceptable. | ||
|
30to1
105 Posts
from my tests (again, it was only a few runs) - the second set of workers per patch harvest less unless you extensively micro them - it depends on exact build times - but occasionally they'll go from patch to patch looking for one thats free - and this costs you minerals. | ||
|
michaelhasanalias
Korea (South)1231 Posts
On December 03 2010 12:22 Xapti wrote: You're telling me you're getting reduced mining rates when going form 8 miners to 16 miners per patch? Doesn't make any sense to me, because workers do not have to wait for each other when there's only 2 so they should not be any slower. I can't say those stats aren't wrong though. Values I heard was 42 mn/min up to 16 miners on liquipedia, but it is possible that information is not accurate. Mining rates have been analyzed and overanalyzed and the numbers have been pretty clear since release. I suggest referring to one of the several other threads or the Liquipedia page on mining efficiency. Let's return to discussing the utility of mass OCs! | ||
|
tetramaster
Canada253 Posts
On December 03 2010 12:26 30to1 wrote: Xapti, from my tests (again, it was only a few runs) - the second set of workers per patch harvest less unless you extensively micro them - it depends on exact build times - but occasionally they'll go from patch to patch looking for one thats free - and this costs you minerals. Ah, anecdotal evidence here, but I have noticed that once in a while I attribute it to the "uneven" placement of mineral patches. It happens more on some maps than others, and on some bases vs others. EDIT: On the discussion of OCs... Good lord, this thing frightens me as a protoss player. Then again, my default build is a 2-gate 5 stalker early pressure build. Lemme cross reference the times and figure out when it hits in conjunction with this build. EDIT: According to the BO test run posted on page 10, 5 stalkers hitting at around 5 minutes would encounter 6 or so marines. Now, if you were to cut out 2 marines and drop a bunker, that would likely hold long enough to get more marine reinforcements (assuming you bring SCVs to repair) | ||
|
revenir
Malaysia7 Posts
If this build works out, it appears to be capable of accelerated starvation once you have enough minerals by just scanning and dropping loads of mules on your enemy's minerals. | ||
|
MichaelJLowell
United States610 Posts
And here we are. | ||
|
30to1
105 Posts
The reality as Shlowpoke pointed out is that by trying to abuse 4+ OCs very early in the game, you're essentially turning this into risky cheese - and the reality is that OC Farming and mule mechanics are much more important that this. I've started thinking that as a guideline for how to utilize this to its maximum would basically be starting out with a very FE as terran if possible, and slowly using OC Farming as a general tool throughout the game. The idea would be to generally just keep a constant economic advantage over your opponent - instead of a cheesy/high risk build. Basically, you always keep an economic advantage by adding 1 OC for every base your opponent has - so that you always have more income than he does (while generally keeping pace with actual expansions). So in a nutshell, 1 additional OC per base. | ||
|
Conrose
437 Posts
| ||
|
Xapti
Canada2473 Posts
On December 03 2010 12:28 mlbrandow wrote: Assuming you read my post, it said that I got my info from liquipedia.Mining rates have been analyzed and overanalyzed and the numbers have been pretty clear since release. I suggest referring to one of the several other threads or the Liquipedia page on mining efficiency. Liquipedia said 42 mineral per minute for 2 workers per patch on one page. and about 40 minerals per minute for 2 workers per patch on another page. On a third page it gets even more specific and says 39-45 with 1 per patch, 78-90 with 2 per patch, which also averages out to 42 per minute. If you have additional conflicting information you'd like to share, then do so instead of saying what you said, because I found your statement to be very disrespectful and ignorant. On December 03 2010 12:09 30to1 wrote: what i'm saying is right. My tests indicate scv rates to be roughly 40/35/25 in terms of decreasing returns on scv saturation (these were not exhaustive, but I trust them more than your numbers). Well, I'd rather trust this guy http://www.teamliquid.net/forum/viewpost.php?post_id=5864468 And his experiments, Than you and your experiments. You can say you're right all you want, but it doesn't make it so. | ||
|
30to1
105 Posts
On December 03 2010 12:40 Conrose wrote: Okay, from preleminary tests vs the Very Hard AI: I can say with great certainty that you'll want your first Rax to pump out Marauders rather than Marines, 5 RR and 7RR do alot of damage and can occasionally snipe a marine. What build are you using to test? I've just run about 15 different builds through various optimizers, I'm about to do some in game tests as well. EDIT: would you mind posting a replay? Even if its crappy, having more games to look at always help. If its too embarrassing just PM me link  | ||
| ||
