|
I'm not neglecting the time value of money, (I'm actually doing a Finance MBA so its at the front of my mind)
But, that's why I'm asking, has anyone quantified that value? Like an expo 30 seconds faster doesn't seem like it would actually make you back the money from 4 SCVs...you'd need it probably, 68 seconds faster (4 scv build times) right?
There's a lot of intuitive stuff here, but anyone who has actual econ benefits from the faster minerals calculated out for instance?
I'm not strongly taking the 'SCV first' side, but that seems where it points if you expect a longer game.
|
Technically if the game will go on long enough that an SCV will harvest more than the MULE will in the amount of time it has remaining, it would be better to target an SCV.
At least I think so, could be I'm ignoring some variable.
|
There is at least one case where targeting a mule is always perfered, and that is when he is on 1-2 mining bases and fully saturated or oversaturated (assuming you cant kill enough to make him less then full saturation). Then its a case of strictly making him get resources slower.
Also if the mule has 1 or 0 trips left before it blows up, then clearly you should be targeting scvs.
So, in the end it depends?
|
Do the minerals mined after killing 4 SCV's take into account mules that are active as well? Because if you compared killing 4 mules from a batch of SCV's to killing SCV's from a group of SCV's only, it would make a difference?
I generally go with any mules that are relatively high on lifetime left and the workers on gas. Those seem to be the most vital. However, if the game has reached the point where the terran player can drop 4 mules, it shouldnt really matter if you kill a mule or SCV, as long as you keep up the harass because you can assume at that point, the Terran economy is already rolling.
|
Mules earlygame, SCV's lategame.
|
Well, no, in the example I gave, it doesn't make a difference. Whether there's 4 more SCVs mining or not, unless its enough to go beyond saturation.
Should make a difference if you're killing 4 from a group of 8, or 4 from a group of 16, it's still only the SCV compared to the mule that matters.
The idea is just to compare all benefits gained from one versus the other. Basically what's the benefit of 135 NOW vs 45/min perpetually? (or 270 if you just saw the Mule Land)
A few are saying Mule early game, SCV late...wondering why? The math alone would suggest the exact opposite, the closer you are to the game end, the more you should target Mules imo. But one SCV killed in like minute 5, has missed out on like 1000 minerals by minute 25
|
On February 16 2011 16:00 EnderSword wrote: Well, no, in the example I gave, it doesn't make a difference. Whether there's 4 more SCVs mining or not, unless its enough to go beyond saturation.
Should make a difference if you're killing 4 from a group of 8, or 4 from a group of 16, it's still only the SCV compared to the mule that matters.
The idea is just to compare all benefits gained from one versus the other. Basically what's the benefit of 135 NOW vs 45/min perpetually? (or 270 if you just saw the Mule Land)
A few are saying Mule early game, SCV late...wondering why? The math alone would suggest the exact opposite, the closer you are to the game end, the more you should target Mules imo. But one SCV killed in like minute 5, has missed out on like 1000 minerals by minute 25
Mules are a resource that is shared with Scans. By taking out a mule early game you deny the Terran player that burst of minerals that he might have needed to construct or build more forces. Not to mention mules are the only eco boost the Terran has. Protoss can always chronoboost more Probes and you can't do anything about it. Zerg can infect.
Terran has to decide between a scan and a mule believe me nothing demoralises a Terran player more then having his fresh mule abducted by space aliens on a drunken driveby. The theorycraft might suggest that SCV's are the better target but really think about the psychological factor here : You have 50 energy, you can use it to scout the enemy base and see what's up or you can get a big boost for those marines / structures. You pick the mule to get more marines since you are going bio only to have it driveby'd 2 returns later.
Tell me, would a Terran player feel more pissed about losing some SCV's (this happens all the time) or would he get more pissed about losing a very important economical linchpin that cost him a scan?
|
He'd be more pissed off by a MULE, But is that always good?
If it means he scans more, and catches you going DT, or he gets frustrated and pushes when oyu didn't want him to etc...
I personally find the most effective harassment is small stuff that they don't bother to address, and they just end up 6-8 workers down all game.
|
It depends on what you want to achieve with the harass.
Reasons to go for mules: -base is oversaturated, so the scv brings no income anyway -you want to slow down the terran for the next minute, as the immediate effect is bigger -you want to bait another mule drop so that no scan is available
Reasons to go for scvs: -long term effect is much bigger, a mule dies anyway -base is not oversaturated
In the end, in general it is more efficient to go for scvs, but there are scenarios where mules are preferred targets.
|
Um. This isn't rocket science.
Always hit the MULES. Because you always want to immediately damage their economy.
Yes, you kill an scv, and you potentially killed off an infinite amount of minerals.... so what. Are you planning on playing for an infinite amount of time? =/
Seriously. I'm pretty stupid, and even I can figure this out...
Side bonus, they usually re-MULE after you pick off MULES, so it leaves their OC's with less energy for scans. hehehe.
|
I think you got this a little wrong. 1 mule gets 270 minerals in 90 sec. 4 Mules get 4 times as much, not just 270. I think you need to look over your numbers once more
Also simple economics. Money now is worth more than money later.
|
On February 16 2011 17:41 mytent wrote: Um. This isn't rocket science.
Always hit the MULES. Because you always want to immediately damage their economy.
Why do you always want to hurt their economy Immediately?
It doesn't matter if you don't play an infinite amount of time when it becomes beneficial after 3 minutes...that's a little shy of infinite
It's not rocket science if you refuse to actually give it any thought. But it has a lot of nuance to it once you consider the different scenarios
|
On February 16 2011 17:47 Neivler wrote:I think you got this a little wrong. 1 mule gets 270 minerals in 90 sec. 4 Mules get 4 times as much, not just 270. I think you need to look over your numbers once more
where is this done wrong?
|
think of it this way, with good mule useage there will ALWAYS be 1 mule per OC period, unless they saved energy for a scan or somthing but even then they can still ALWAYS have 1 mule up and running per CC during their energy regen time. so if i take the mule out that means they have to wait for their energy to regen in order to cast another mule.. and SCV may only cost 50 minerals, kill 4 of my SCV's its ok because i have a mule giving me the same resources... as those 4 , also i don't plan to stop SCV production just because you killed a few, choosing the mule makes it so they mine minerals slower which means they are forced to produce units at a slower pace... so i wonder why kill 4 scv's if i can force them to use their energy to drop another mule(this removes options like scans/supply depot drop) because i killed the last one they dropped. well honestly there isnt any reason to kill anything but the mules if u have 3 seconds to do dmg or lose units, kill the mule it is going to give more minerals in less time than the SCV's thus making their build slower. also if u have a choice between 1 mule and killing off 4-5 GAS workers imho i would just kill the gas workers as many many players DO NOT re saturate for a long time( from my reps about 5-45 seconds **game time**)
the non-Logical choice would choose mules due to the fact you SLOW their mineral gain which slows units/buildings being built.
The Logical choice would choose mineral gathering workers. Long term damage(only useful if the game lasts for a long period)(20+ mins)
The smart choice- Targeting gas mining workers. not only does this kill their workers but slows production of ALL units that require gas.
Mules only mine minerals so you have to smash 4 or more workers that mine minerals in a single harassment attack to make it worth not killing the mule. killing the gas workers causes them to use actions to replace the workers; actions that could be spent building units or controlling your army(pros normally don't forget this so don't expect this to be your savior vs one)
|
Whatever you can kill the most of with the least risk to your own units. This is a really good place where your own judgment can come into play.
|
well....if u do kil l mules...because the mine shit so fast, isnt it more logical? they mine mins FASTER which means in that space of 10mins terran would have an extra rax or so bekuz of that temporary increase of mineral income...
|
It's better to hit scv's imo because the mule you kill might have only had 3 seconds left till it timed out in which case the damage you have done to their economy is zero.
|
MULEs are actually considerably harder to kill than SCVs. It's not just the 15 extra hitpoints - you usually have to walk further to get to them, and you don't benefit at all from injuring them and coming back later. And it's pretty clear that SCVs are worth more in terms of long-run minerals mined.
Only time you'd want to target MULEs is when they're oversaturated or when you want to mess with their timings. And even then, only if you're pretty sure it isn't about to time out by itself.
|
Ok, lets look at this mathematically.
You do not know at which point of a mules lifetime it is, unless you saw it drop, or you have insane knowledge of the timings of terran. So lets assume you do not know it, which means that on average, you kill 50% of a mule if you kill this mule instead of an scv. This would be 135 minerals over the next 45 seconds.
Assuming the base at this point only has two SCVs on each mineral patch, one scv will collect about 30 minerals over these 45 seconds. So 45 seconds after this action, the terran has either lost 100 minerals, or a worker. A worker costs only 50 minerals, however, worker production is also limited by the amount of CCs you have. It will take the worker another 1:15 to collect those 50 missing minerals, or 2:30 to collect the whole 100 minerals if we are only interested in how much minerals the terran has to spend on an army.
You could use those extra minerals to expand earlier. Assuming you would want to build a CC exactly 45 seconds after the harrass, and you accumulated money to do so at a rate of about 100 mpm, since you would also need to produce units throughout this time, this could ideally delay the next CC by another minute. Which obviously would be worth it, however this is also the very unlikely best possible option.
If the terran has a better saturation than 2* mineral patches, but is still constantly producing workers, the effects of killing workers instead of mules become drastically less. Assuming the third worker on a patch produces 20 mpm, 45 seconds after the harrassment, it would have collected 15 minerals, meaning that the mule is 120 minerals more of a loss at that point in time. To collect the 70 additional minerals to the costs of the worker, this worker would need 3:30, at which point both harrasses would have been equally effective. At any point before that, killing the mule was more beneficial. As for the total minerals to be spend on army, it would take this pure worker a full 6 minutes to accumulate those 120 minerals, and only thereafter it would have been better to have killed this worker instead of the mule.
If we are that late in the game that the terran is not constantly producing workers anymore, killing a worker effectively costs the terran 50 minerals + what that worker mines in 17 seconds, which could be anything from 5-15 minerals depending on saturation and the position of the worker between mineral patch and OC when you kill it. However, even with 65 minerals lost it would still have been smarter to target the Mule, which costs him 135 minerals.
If you decide to target SCVs instead of Mules, you should always take those that are mining gas instead of those that are mining minerals. Even with instantaneous transfer of different workers into the gas, this costs the terran some gas, and most terran builds are gas-limited. Furthermore, the better your enemy is, the more likely it is that they will mine the exact amount of gas at the exact time they need it to do what they want to do, so by reducing their gas-income, you directly delay whatever they are doing. Also, it takes away some of your opponent attention by having to put workers back into the gas, and there is also a slight possibility of your enemy not noticing that a worker on gas is missing, which would be the best possible result.
As a conclusion, i would target worker over mules if the base we are talking about is at 2*patches or lower saturation, but always focus the ones on gas. If the base is on higher saturation, or the terran has reached his limit of scvs in the game, i would always target mules, or maybe single scvs on gas in the hope that he does not notice it it gone.
It should also be noted that a Mule has more HP than an scv, and takes longer to kill as a result, at least if you have a low phoenix count.
|
On February 16 2011 17:48 EnderSword wrote:Show nested quote +On February 16 2011 17:41 mytent wrote: Um. This isn't rocket science.
Always hit the MULES. Because you always want to immediately damage their economy. Why do you always want to hurt their economy Immediately?
Because you just spent a bunch of resources getting a unit to harass, then used a bunch of energy on those harass units to render their effectiveness in real combat even less.
You want to keep him pinned down with those phoenixes, the best way to do that is to delay his army creation/turrent creation as long as possible, both so you can harass more efficiently and have enough time to turn that harass into a significant long-term lead, rather than a minor annoyance that leaves you at a disadvantage when he pushes out. You can't boil down complex systems down into a simple dollar sign that easily.
|
|
|
|