In Starcraft, there is no clear 'optimal' number of miners per mineral patch. More miners gets you more minerals, even if there is a diminishing returns. However, you can theoretically get the maximum possible minerals out of just two workers per mineral patch. The time it takes to mine a mineral patch is generally greater than or equal to the time it takes to do a round trip to/from the nex/hat/CC (this varies depending on mineral distance, etc.) Hence, you can theoretically max your mineral gain with two workers per mineral. The 'problem' is that the AI will immediately try to find another patch if it is in use, and this results in less than optimal mining (and why more workers will mine faster).
How would gameplay be different if you knew you didn't ever need more than 2 workers per mineral patch. In Starcraft 2, the mining AI seems to be much smarter. It looks like close to two workers/patch will max out mining, as the workers seem to go looking for new patches a lot less than in starcraft..
Particularly, look at the bottom mineral of the center clump of minerals in the below video.
It also feels to me like when workers are added in SC2, they quickly and efficiently find a patch that doesn't already have two workers on it.
If I'm not mistaken, it seems like this could significantly change the game. Less clumped workers would mean less incentive to raid workers. If workers are just as clumped, there is still less incentive to do so because if you don't knock them below 2 workers/patch, you don't really do anything to their income.
I guess you could be right on this, I dont agree that harass will be less useful though. A storm or tankdrop can easily destroy all your mining within seconds and although less time needed to resaturate the income drain will still seriously hamper your production capacity.
I felt exactly that when I played SC2 the last two times. Harass just isn't as effective anymore. Also the units you have are way worse at it than in BW.
harrasement is still worth it. Remember, maynarding. Even if you dont damage their economy directly at the moment, they will still have to replace those workers for a future expansion. An expansion, even in starcraft 2, SHOULD have an over abundance of workers, because of the maynard process.
Harassment is still going to do something, but my concern lies more in how much more precise the game becomes. Like it will be very easy to figure out what the optimal strategy for any of the resource stuff is.
I'm not sure how valid of a concern that is.
ed: for example, you make more workers than the maximum you need, so that you can have your next expansion saturated immediately. So let's say you have all the extra workers automine at the nearest expansion, and when the new expo is about to come up, you just take any above the max you need at that expansion and send them to the new expo. Is that a problem?
I don't feel like it will affect things *that* much, as the general rule for SC for Toss/Terran was 2 workers per patch, and Zerg was 1.5. Not that I'm a good player, but did higher saturations really have *that* much of an effect?
Moreover, if what you say about SC2 is true no good player will have more than 2 workers per patch. More is pointless since they aren't mining faster, unless you're saving workers to transfer to an expo. Doing otherwise would sacrifice minerals that you could be spending on your army. Therefore something akin to a vult drop will always have some effect if you kill a few workers since the player will at least have been planning to use them at another expo.
On January 01 2010 12:45 errol1001 wrote: Harassment is still going to do something, but my concern lies more in how much more precise the game becomes. Like it will be very easy to figure out what the optimal strategy for any of the resource stuff is.
I'm not sure how valid of a concern that is.
ed: for example, you make more workers than the maximum you need, so that you can have your next expansion saturated immediately. So let's say you have all the extra workers automine at the nearest expansion, and when the new expo is about to come up, you just take any above the max you need at that expansion and send them to the new expo. Is that a problem?
Can you elaborate a little? How will this change from SC1 to SC2 to create a concern?
He means there will be no fuzzyness to the economy. In SC1, there is doubt as to how much you get per scv. Obviously after 1 per patch it decreases efficiency slightly due to worker confusion.
But in SC2 your mineral count should be almost exactly # of scvs times amount an scv can harvest per unit of time. The only wild card is the distance to the patch. From the videos we've seen it doesn't look like that plays a very major role. At least I think that's what he means.
However I don't see how this lowers the utility of worker raids. Is it because each worker mines less so each worker you kill is less important?
Rather because each worker mines more, you need to have less workers. And lots of worker raids use splash damage.. so maybe in sc2 the same raid with psi storm would kill 10 workers rather than 20. Maybe, anyway...
I think the idea is that once your economy gets established, its easier to replace losses from harras and get your economy fully fuctional again compared to sc1. In sc1 a fully saturated base, well, I dont really know, say 20 something workers? 20 x 50= 1000 minerals. In sc2, if a fully saturated base is 12 workers, it costs 400 dollars less [and a lot less time, so its even easier to replace than you would immediately think] to replace the damaged economy. I think thats what hes trying to get at. I guess early game, losing workers would have a more substantial effect than in sc1, but later game it would have less of a substantial effect. So running around the map with a shuttle and storm dropping bases can still be useful, but it might not give back the same returns.
It's similar to MBS, I don't think I need to point out how much resistance that met. It makes total sense for them to improve on things like that, but it changes the gameplay. With MBS, that was considered bad enough that effort has been put in to counteract that change (macro mechanics). Anything needed here? Don't know. I can't say, but I can bring up the difference.
I might be missing the subject but Im playing a David Kim card. Tap David Kim to show him killing workers and how it hurts in various replays.
Regarding economy...not a clue, we barely know how the game plays let alone economy. Not to say there cant be pages about possible worker effects on it mind you. Just saying I wouldn't venture a guess down that road.
I read somewhere ages ago that one of the reasons dual gas was implemented was to keep a similar number of workers in each base after the saturation changes took place. The repercussions of this on worker raiding are quite interesting. Killing each worker on gas only causes half as much gas mining damage, but replacing workers on gas (transfer from minerals) impacts more heavily on your minerals than before. The layout of gas geysers will also impact heavily on how defence and raids play out. If the geysers are on opposite sides of the CC like in many BRs we've seen then it is harder to destroy all workers on gas in a single attack, but it is also harder to defend both paths as they are so far apart. Placement of static defence will be very different. If they are together then it is easier to defend them but aoe will be more effective against them. I can't wait to see the sim city (I hate that term) for defending both gas plus the minerals and how the choices of which to guard first in early game impact on the harassment game.
On January 01 2010 17:23 errol1001 wrote: It's similar to MBS, I don't think I need to point out how much resistance that met. It makes total sense for them to improve on things like that, but it changes the gameplay. With MBS, that was considered bad enough that effort has been put in to counteract that change (macro mechanics). Anything needed here? Don't know. I can't say, but I can bring up the difference.
The thing is with MBS is you'll still have to seperate what you're building from because say for example you are going M&M. Marine/Marauder you won't want every single barracks building one or the other you'll still want a split. Kind of like how rax work now with Medic Marine.
If min lines are saturated with 2 workers per patch, people won't make a lot more workers than that unless they are planning to expand soon. More workers means smaller army, after all. So killing workers will be just as effective as in SC1. Or, if your opponent oversaturated to be safe from harrass, you will be able to simply crush him with your larger army.
Yes, workers are way smarter in SC2 than in SC. Yes, it will change the income curve. No, the number of optimal workers per patch won't be changed that much. That was why the reason behind reducing the minerals per trip to 5 but also reducing the "mining time" of this 5 minerals. They try to make travelling time to mining time ratio similar to that of SC1 which should keep income per worker similar to that of SC1. At least at certain number of workers...
Well, even if raiding workers is harder its still totally viable thing to do. The % of your resources that go towards harassing forces might be smaller, but so should the opponents spending on defending against some harass.
Resource harass is not just limited to the effect of income. It also forces your opponent to spend his most valuable resource (time), as well as some of his military on removing the assult. Harass will be valuable regardless of saturation calculations.
On January 01 2010 17:23 errol1001 wrote: It's similar to MBS, I don't think I need to point out how much resistance that met. It makes total sense for them to improve on things like that, but it changes the gameplay. With MBS, that was considered bad enough that effort has been put in to counteract that change (macro mechanics). Anything needed here? Don't know. I can't say, but I can bring up the difference.
The thing is with MBS is you'll still have to seperate what you're building from because say for example you are going M&M. Marine/Marauder you won't want every single barracks building one or the other you'll still want a split. Kind of like how rax work now with Medic Marine.
No you won't. By pushing a hotkey, you only produce one unit at a time. It's not like in Warcraft III.
As far as I know, the mining rate (i.e. how many minerals will a worker mine per second if you exclude travel time) is the same in Starcraft and Starcraft 2, so that means that workers in sc2 will spend 37.5% less time actually mining per trip (because they can only carry 5 instead of 8 minerals). Therefore, the actual number of workers needed on minerals will not be that much lower.
However, due to the improved AI, the income-saturation curve will look a lot differently. In both games, you get very slight diminishing returns as long as the ratio of workers on minerals and number of mineral patches is less than 1 (since the distance from main building and the mineral patches varies slightly, so you can use the closer patches first). As the ratio approaches 2, you get moderate diminishing returns in sc1, every worker you add increases your income by less than the previous worker due to worker AI being very inefficient. Adding even more workers means severely diminishing returns. In contrast, you get only slightly diminishing returns in sc2 when going from 1 worker per patch to 2, since the workers are a lot smarter. However at some point (probably between 2 and 2.5) you hit a brick wall and further workers will just be idle.
Due to this (and the fact that the current macro mechanics increase increase your mineral income (or save minerals in the case of the queen), but not your gas income), I predict that people will tend to expand earlier. Keep in mind though that playing aggressive requires less knowledge of the game, so while there is no relatively stable metagame, we will probably see less economy oriented play than we currently do in sc1.
As for economy raids, they will probably be about as useful as they are now. On the one hand, losing workers in sc2 will affect your current income more strongly. But on the other hand, players will be able to replenish their workers faster. Zerg now have queens, which makes larvae less of a limiting factor, thus making Drones effectively cheaper. Terrans and Protoss will tend to have more CCs/Nexus (see above paragraph). Or to be more precise: I predict that killing workers individually (e.g. with Mutas or DTs) will be less effective against zerg since they can replenish them more easily and since they will tend to have more workers per base. Killing drones with aoe/splash (e.g. storm drop) will be about as effective as before (you can replenish workers faster, but you also have more workers per base). Against Terran and Protoss, killing workers individually will be more effective because the tendency to expand faster cannot offset the greater economic impact that every worker has in sc2. Aoe/splash based harassment will be less effective against Terran and Protoss because they will have less workers on minerals (and gas workers are not very good targets for aoe stuff) and they can replenish their workers slightly faster.
tl;dr version: Killing workers will still be important.
this could be interesting. this is just my theory but...
considering that you only need 9 patches x 2 workers= 18 workers for a saturated line. if more workers doesnt necessarily mean faster mining rate (equilibrium already reached) then we should expect player not pumping more at their main UNTIL they expo. as a result, players will start pumping workers in this timing where right before the expo is finished they will have made enough workers to transfer to the expo thus gain the expo benefit faster then before (instead in sc1 you could make 5 workers before expo finished, roughly from the timing).
so it seems a new window of opportunity will be present where it will have incentive for players to drop at a correct timing. and if enough damage is dealt, the players eco could be literally halted (no workers for expo and few at main) and they could take adv of low army count cause of player expo investment. hell this could help terran players significantly. if they time the drop, their timing for the push is extended longer
ehhh??
the numbers here are a bit of a guess/estimate
god i wish the reaver was here. fucking rape the eco so badly (at least in early and mayb mid)
Drunken.Jedi = genius Dear lord would Blizzard really dumb down the game that much?? like in WC you have what.. like 5 miners and you're maxed? one of the best things about sc1 was the complex.. dare I say organic nature of balancing an economy. One thing to keep in mind through all this that I think Drunken made an allusion to and Heroyi touched on.. was it looks like you're probably going to be able to have more bases in the early to meta game. This means you're constantly going to be making workers to fully saturate those new expansions when they come online. thus you're going to have more workers on the field which takes up a big chunk of your 200 supply. in SC1 60-90 workers for late game (super general) but maybe it'll be around there too for SC2. in conclusion if you go for an econ/macro build you'll still probably be pumping workers in about the same fashion as you were in SC1.
I'm more worried about early game worker phalanxes. Sure it's always a disadvantage to pull workers but if they can fight successfully rather easily (we've seen this a bit before) it may discourage rushing. I'm also concerned about Terran default blocking of ramps that Now are neither situational nor takes skill or knowledge. It really has a chance to mess up early game rushing. But that's a different story alltogether. In short, auto surround and Better pathing will make early game raiding and rushing more difficult I predict. Is this what we want?
On January 02 2010 04:59 Knee_of_Justice wrote: I'm more worried about early game worker phalanxes. Sure it's always a disadvantage to pull workers but if they can fight successfully rather easily (we've seen this a bit before) it may discourage rushing. I'm also concerned about Terran default blocking of ramps that Now are neither situational nor takes skill or knowledge. It really has a chance to mess up early game rushing. But that's a different story alltogether. In short, auto surround and Better pathing will make early game raiding and rushing more difficult I predict. Is this what we want?
I would say yes. I mean why wouldn't you want the workers themselves easier to handle so you just need to worry about the stratergy your macro and micro. Early game raiding should be harder. Atleast that's how I see it anyway.
The thing about Terrans blocking the ramps being situational is when you look at the game currently. Unless your someone with a sick gosu APM of like 400 it's much easier to just wall yourself in. There's no disadvantage to it what so ever so why wouldn't you wall? Unless of course you're doing 1 Rax FE against zerg.
One thing that I have thought about is these situations in starcraft where a player gets pushed out of their natural. Obviously, this is bad, but we see players come back from this. The primary loss is probably making less workers and having one less gas though, I think. The mineral income loss isn't actually -that- bad since you have lots of extra workers on your main's minerals.
What if, on the other hand, your mineral income immediately gets cut about in half, and those extra workers are useless in your base? Will we see people even try to retreat to their main, or will they always just pull the workers to try to push the enemy out - because if they don't, they might as well gg?
On January 02 2010 04:59 Knee_of_Justice wrote: I'm more worried about early game worker phalanxes. Sure it's always a disadvantage to pull workers but if they can fight successfully rather easily (we've seen this a bit before) it may discourage rushing. I'm also concerned about Terran default blocking of ramps that Now are neither situational nor takes skill or knowledge. It really has a chance to mess up early game rushing. But that's a different story alltogether. In short, auto surround and Better pathing will make early game raiding and rushing more difficult I predict. Is this what we want?
I would say yes. I mean why wouldn't you want the workers themselves easier to handle so you just need to worry about the stratergy your macro and micro. Early game raiding should be harder. Atleast that's how I see it anyway.
The thing about Terrans blocking the ramps being situational is when you look at the game currently. Unless your someone with a sick gosu APM of like 400 it's much easier to just wall yourself in. There's no disadvantage to it what so ever so why wouldn't you wall? Unless of course you're doing 1 Rax FE against zerg.
Its not inherently bad, but i guess it will make games last into midgame more easily. Of course no one likes being zergling or zealot rushed, but its still a strategy. This completely nullifies that type of quick harass especially against terrans.
Again, not neccessarily inherently bad, but it is something to consider during balance: people have bitched about almost every aspect of the game, and here is a development that threatens the ability to do the legendary zergling rush and no one is complaining or even a little nostalgic?
Im definitely not suggesting stupid workers, but i think that sending in a zealot and expecting it to take out half the mineral line will be a lot harder than in SC. Good players probably will be able to just micro their probes so they never or rarely die against lone units or small groups of them.
Also, in SC, when you blocked your ramp, you had to consider that zerglings could still get through. Now they cant. Every building blocks perfectly now, which will have a huge effect on zerg with all their short range/melee units, but also zealots, archons and larger units like tanks, thors and ultras.
Hm. The only thing I can think about is if you knew how to block your ramp especially walling you had to know how to do it properly to stop zerglings coming in.
The buildings don't build any faster so 4 pool will still be effective even 9 pool as there is no way the wall will be finished by that point. Even if they block with SCVs then it'll just be how it is now. It's not as if workers have more health.
It will also be easier to defend your workers because they block enemy movement path without hindering their mining path, but your units can path thru your harvesting workers.
Yeah, I guess that strikes me as a little silly. No need to rush a bunker against a 4-pool, just run your marines around in your mineral line... reminds me of those fail banelings from the ZvP battle report.
My point is that small Zerg rushes won't be as effective because workers now have auto surround and don't suck at their pathing which means they are easier to micro. I like to think of them as a single unit doing a damage multiplier attack: with 10 workers you are doing 10*5 damage Which isn't shabby! They still are weak but en masse they will proly be able to take lings. If anything they should just lessen their attacks.
The worker harrass was always SOOOOOO amazingly exciting to watch. Zerglings zipping back and forth, dodging military units and killing workers due to thier superior manuverability. Mabye they can turn off auto sorround for workers, I mean they arnt meant to fight, why should they know how XD.
What you have to keep in mind though, is that SC2's equivalent to 9 pool is probably going to be 10 pool, 10 ovie, 13 queen, which should get the first batch of zerglings out slightly faster, while having better production capabilities and more flexibility than a 10 hatch, 9 pool. This really more than makes up for better worker AI.
However, it's true that 6/7 pool will probably be a lot less viable in sc2 than 4/5 pool is in Broodwar, but I don't think that's a bad thing. 4/5 pool very often leads to a build order win or a build order loss, which are bad for a competitive game.
Which is why most people dont use them. But the option is there.
I dont think blizz should determine which strats are "good" and which are "bad." Im not suggesting they are consciously trying to minimize the possibilities of these strats but they are unconsciously with things like better workers and autoblock ramps.
Im also not saying that "just because it was in BW, it HAS to be in SC2" but I am trying to indicate that the game should be as flexible as possible: if a person wants to go for an ultra fast zergling rush, knowing that their chances of winning are suspect, thats still a risk the player can take.
Making it EZPZ for terran to block their ramp kind of minimizes a certain type of play. Smarter workers also minimizes that type of play. Im sure new strategies with the new cliff-jumping units will emerge for raiding early game, but it just seems... frustrating... that zerg rushes are lining up to be less effective and that ANY rush against (a good) terran will be ineffective.
Again, yes, rushes suck and should by no means be foolproof, but i dont think their counters should be foolproof either.
Overall though, i dont think many people are overly worried about this issue.
Regarding saturation, I remember the devs saying that they were making a point of designing the most efficient worker/patch ratio to be exactly 2:1 Also, as others have mentioned, the smarter AI — particularly "auto-split" — will make diminishing returns less...well, diminishing.
Strangely enough the worker changes (assuming the 2xPatch worker saturation we're talking about here) will result in drastic changes of how Maynarding and early expanding works. An individual worker is now significantly less efficient than in BW (~37% reduction in mining per "travel" interlude) even though workers past the 1-per-patch maintain the same efficiency to compensate (unlike in BW).
This means that it takes a new expansion significantly longer to produce results without transferring workers. Whereas in BW the first 7-9 workers were far more valuable than the following 7-9, in SC2 each of the first 14-18 workers share the same level of efficiency. This means an expansion will scale linearly from workers, as opposed to the non-linear charts that have been posted on this board in the past.
At the same time, Maynarding from your main when your main has approximately ~2 workers/patch will no longer be viable. That's right - you will LOSE money in the maynarding process that will NOT be recouped by sending workers to a fresh expansion. In fact, it's far more efficient (in terms of how quickly you get the same amount of money) to simply rally your main CC to your expansion and pump workers off that way (once you've maxxed out at ~2/patch at the main) than to cut away from your mining force at all.
Essentially, unless you have more than ~2 workers/patch, you should not Maynard at all. If you do have more, you should Maynard exactly down to ~2 workers/patch at your main, and no more.
So what practical impact does this have on the game? Well, first off, it discourages early expansion in response to aggression. Why is this the case? Well, statistically, it takes longer to recoup the cost of an expansion and to gain an economic advantage since an individual worker produces less minerals in the same amount of time. The end result in minerals/minute will be approximately the same from a fully-loaded expo, but the expansion's production will scale slower and anything less than near-full saturation will give less minerals/minute than an equivalent expo in BW. Thus, a situation like you often find in BW with early expansions, where the economic advantage can carry a game if the greedy player holds off early aggression, will take longer to realize.
Now, there's another mitigating factor that still has to be considered: increased worker production can help reach the saturation point quicker. Since we're working on the assumption that the second "set" of workers can now produce about as efficiently as the first, this second wave of workers becomes situationally more valuable than in BW. Thus, the true benefit of the expansion may be more its ability to produce more workers, faster, than its function as a second mining base - which will be especially important when a player pushes to take a 3rd and is able to do a full 2xPatches worker transfer without cutting into active 2x worker force. It may even be valuable to rally extra workers past the 2xPatch mark to sit idle, off to the side of the new expo, so Maynarding will be as fast and efficient as possible without cutting into production. It will take a lot more math and time comparisons to determine exactly how much this "faster worker production" figures into the economic advantages offered by an early expo, and if this can make up for the lesser immediate financial impact it will now have.
Of course, maybe the targeted macro mechanic will totally offset the lesser value of an expo rendering half this analysis moot =]
Exellent post tedster. It will be interesting to see what the increase is in minerals per min for excess of 2 workers per vein too, but I guess that will have to wait for beta.
On January 05 2010 08:33 DeCoup wrote: Exellent post tedster. It will be interesting to see what the increase is in minerals per min for excess of 2 workers per vein too, but I guess that will have to wait for beta.
I'm really hoping they change the timing a little or something because it would upset me to see a neat mechanic like worker transfer be unnecessary in SC2, but based on the 2xPatch saturation that we're seeing and the lack of "overflow" it's looking to be the case at the moment. I hope Blizzard has considered this and I'm pondering posting something about it on their boards because I haven't seen it mentioned anywhere.
I also don't like the idea of having a "max saturation point" for workers such as you see in, say, Warcraft 3. By having what seems to be a linear scale with a distinct stopping point, you will ALWAYS have 2xPatches workers at a base, and thus you aren't making any real decisions with regards to "do I pump more workers for a slight eco advantage and transfer them later or am I fine as-is?" questions.
Hmm I had no idea about this saturation issue, I guess a fix would be to jimmy the workers a little so saturation is at something like 2.2 workers. so extra aren't completely useless, but it's better to have them at an expo...
Easiest way to do this would be to slightly increase the collision on them, so they'd take a split second longer to reach minerals at higher saturations....imo =p Shouldn't affect much, other than walling slightly easier I guess.
Also, just saying, the new workers produce the same money over time that the current ones do, they had to reduce the collected mins from 8 to 5 to stop them collecting too much with the improved pathing. So basically rather than collecting 37.5% less, they have 37.5% better pathing =p
On January 05 2010 04:07 Knee_of_Justice wrote: Which is why most people dont use them. But the option is there.
I dont think blizz should determine which strats are "good" and which are "bad." Im not suggesting they are consciously trying to minimize the possibilities of these strats but they are unconsciously with things like better workers and autoblock ramps.
Well, you're not forced to use those "coinflip strategies", but if your oponent wants to flip the coin you're still along for the ride. You can of course go for a safe opening, but that leaves you vulnerable to risky economic openings like 14 cc or 12/13/14 nexus.
Of course Blizzard has to decide what sort of strategies they want to encourage. After all they are designing the game, so obviously every design decision they make influences what strategies will be viable. And making strategies that turn Starcraft into a game of chance less viable is IMO definitely desireable.
Its not quite the coin flip youre making it out to be. Anticipation and scouting play into how you respond to rushes. The game may not reach, say, carriers, and in that sense you are right: they are dictating the game to an extent, but the game isnt out of your control. Its just another build you have to counter somehow, and that takes skill, scouting and anticipation.
You could also make the point that better workers = more control = less randomness: workers will do what you tell them and wont be retarded, but that will affect the game in other ways.
I just think blizzard should pay lots of attention to these types of early game strategies since they add flavor to the game (even if they arent always desirable). Its just something else for them to consider.
On January 05 2010 08:46 tedster wrote: I'm really hoping they change the timing a little or something because it would upset me to see a neat mechanic like worker transfer be unnecessary in SC2, but based on the 2xPatch saturation that we're seeing and the lack of "overflow" it's looking to be the case at the moment. I hope Blizzard has considered this and I'm pondering posting something about it on their boards because I haven't seen it mentioned anywhere.
I noticed this 2xpatch saturation probably a year ago, whenever the first videos featuring sc progamers came out. And I sat there, figuring someone would post something about it but it never happened and I finally posted something. I kind of figured that important people at Blizzard read these forums anyway, but I don't know how much it matters.
In other words, they must be aware of what they have done with respect to mining. it looks like a very intentional decision. The mining time/worker speed/mineral distances/worker ai have been chosen such that it exactly caps at 2 workers for a far patch, and workers don't go looking for new patches when they go to a near patch (so 2 workers caps short patches too because they aren't dumb). There are some bonuses to it too, I guess - any expansion is going to have the exact same mineral income when it is saturated now, provided the number of patches are the same. Anyway, as I think they are aware and have thought about it, I doubt they will change anything unless there were to be a big uprising like with MBS. And I doubt that has any chance of happening before beta (look at this thread - no real uprising, people are unsurprisingly apathetic since sc2 has been delayed so much)..
I don't like the 'linear' scaling. One thing though. Scaling isn't totally linear. When saturated, every patch provides equal mineral income. But when there is only one worker to a patch, the nearer patches are going to give more minerals (& note there are probably only 3 or so 'near' patches at an expansion..). Still pretty damn linear..
You came to pretty much the same conclusions I did. The new mechanics are very precise and easy to understand and very simple.. I am not sure if I will enjoy watching games of SC2 where the logic behind expanding, behind denying expansions, etc, is very formulaic.
Very well written post, thanks for adding your thoughts.
On January 05 2010 10:43 emikochan wrote: Also, just saying, the new workers produce the same money over time that the current ones do, they had to reduce the collected mins from 8 to 5 to stop them collecting too much with the improved pathing. So basically rather than collecting 37.5% less, they have 37.5% better pathing =p
If this is the case, then a new expansion will produce the same amount of minerals overall as it did in BW (assuming the same number of drones, up to a point), but compared to production at the main it not provide the same "boost" that a BW expansion does. Since Maynarding the second set of workers does not actually produce more efficient mining (and in fact will cost minerals in the short term without a long-term gain) you're not going to see an economic advantage until the new workers you produce to work the new expansion pay themselves back.
It's a far cry from BW, where you would start to see results from transferring even a few workers within the first 30 seconds (!!!) with the increase in productivity being equivalent to the cost of the CC in the first 60-90. So while a BW expansion may not gain more minerals in absolute terms than an SC2 expansion, we're still looking at a bigger overall economic boost as a % of immediate total production.
Not intrinsically a bad thing, but definitely structured to encourage 1-base play more than BW was. Also encourages rushes and early all-in strategies in response to expansions as it will take longer to get ahead for the expandee, giving a larger timing window.
On January 05 2010 03:27 Drunken.Jedi wrote: What you have to keep in mind though, is that SC2's equivalent to 9 pool is probably going to be 10 pool, 10 ovie, 13 queen, which should get the first batch of zerglings out slightly faster, while having better production capabilities and more flexibility than a 10 hatch, 9 pool. This really more than makes up for better worker AI.
However, it's true that 6/7 pool will probably be a lot less viable in sc2 than 4/5 pool is in Broodwar, but I don't think that's a bad thing. 4/5 pool very often leads to a build order win or a build order loss, which are bad for a competitive game.
This is the build I did at PAX to win my beta key <3!!!!
When NonY lost his natural on Outsider, his income dropped to ~1k minerals, while IdrA, on 2 bosses, was earning ~1.3k. NonY lost anyway, but how bad could it be in starcraft 2, if income is cut directly in half in that situation? 650 minerals instead of 1000.. 650 minerals against an opponent with 1300 income?
This actually might add some additional strategy because of the lack of benifit that an overly saturated base provides. - When both my main and expo are fully saturated do I cut worker production because the benifit is so low if I am not intending to expand soon? Or do I build extra workers to counter the losses of worker harass? - Risk vs potential gain (potential lowering of losses) - I just scouted Terrans base and noticed that he is fully saturated but is still pumping probes he is probably planning on expanding and will be slightly lighter on military for the next few mins while he continues to pump probes. - New timing windows
I also would like to point out that there will still be a mineral gain for having more than 2 workers per mineral. The major change is that there is no diminishing returns UNTIL the point where you have 2 workers per base. Yes the returns for additional workers will be very low, but having 3 per mineral will still net more returns.
On January 10 2010 09:00 DeCoup wrote: This actually might add some additional strategy because of the lack of benifit that an overly saturated base provides. - When both my main and expo are fully saturated do I cut worker production because the benifit is so low if I am not intending to expand soon? Or do I build extra workers to counter the losses of worker harass? - Risk vs potential gain (potential lowering of losses) - I just scouted Terrans base and noticed that he is fully saturated but is still pumping probes he is probably planning on expanding and will be slightly lighter on military for the next few mins while he continues to pump probes. - New timing windows
I also would like to point out that there will still be a mineral gain for having more than 2 workers per mineral. The major change is that there is no diminishing returns UNTIL the point where you have 2 workers per base. Yes the returns for additional workers will be very low, but having 3 per mineral will still net more returns.
I think you're misunderstanding a bit about how this will impact the game. It's not that an oversaturated base will matter less - you're still going to see a lot of workers at each base, with extras left over for future expos - but the issue is more that expansions give less of an immediate boost, especially when they are lightly saturated. That's the major balance change - not that an oversaturated base isn't viable, but that a lightly saturated expo is not as valuable compared to the saturated main.
This is a much bigger change to what "light saturation" means to a game state than to what "heavy saturation" means.
On January 10 2010 09:00 DeCoup wrote: This actually might add some additional strategy because of the lack of benifit that an overly saturated base provides. - When both my main and expo are fully saturated do I cut worker production because the benifit is so low if I am not intending to expand soon? Or do I build extra workers to counter the losses of worker harass? - Risk vs potential gain (potential lowering of losses) - I just scouted Terrans base and noticed that he is fully saturated but is still pumping probes he is probably planning on expanding and will be slightly lighter on military for the next few mins while he continues to pump probes. - New timing windows
I also would like to point out that there will still be a mineral gain for having more than 2 workers per mineral. The major change is that there is no diminishing returns UNTIL the point where you have 2 workers per base. Yes the returns for additional workers will be very low, but having 3 per mineral will still net more returns.
Have you actually looked at what the workers do in videos? You say 'having 3 per mineral will still net more returns', and I say 'Yes, you might get 1% more minerals for having 50% more workers'. You make it sound as if having more than 2 per patch will be meaningful. It hardly looks that way.
Having excess workers is one of the best way to fend of a harass. If you only have like 16 workers and get storm dropped and suddently have 6, it's gonna take a long time to replace what has lost. You will also be illprepared for saturating an expansion. Late game, sure it would help increase max army size but there are more reasons than bad AI to make a lot of workers.
On January 10 2010 14:01 MiniRoman wrote: Having excess workers is one of the best way to fend of a harass. If you only have like 16 workers and get storm dropped and suddently have 6, it's gonna take a long time to replace what has lost. You will also be illprepared for saturating an expansion. Late game, sure it would help increase max army size but there are more reasons than bad AI to make a lot of workers.
But if those workers aren't paying for themselves in the meantime it's better to just get static defense/more troops to fend off the harass with the same money.
Also if you have more workers at your mineral line, more of them will die to storms. So if you are producing excess workers you should seriously consider NOT letting them harvest if you don't gain more minerals from it.
On January 06 2010 12:59 errol1001 wrote: I don't like the 'linear' scaling. One thing though. Scaling isn't totally linear. When saturated, every patch provides equal mineral income. But when there is only one worker to a patch, the nearer patches are going to give more minerals (& note there are probably only 3 or so 'near' patches at an expansion..). Still pretty damn linear..
You're right. But it shows too that the problem is very easy to fix (if Blizz decides that it has to be fixed). You could generalize the "nearer patches mechanic" by introducing a very small cooldown (let's say 0.1 or 0.2 sec) on mineral patches, without changing the workers AI. As a result the second worker per patch will not yield the same amount than the first; reintroducing the interesting oversaturating-maynarding process.
On January 06 2010 12:59 errol1001 wrote: You came to pretty much the same conclusions I did. The new mechanics are very precise and easy to understand and very simple.. I am not sure if I will enjoy watching games of SC2 where the logic behind expanding, behind denying expansions, etc, is very formulaic.
From the spectator's point of view (which is largely mine) I think the main problem is not the formulaic shape of expansions making/denying (I'm not sure that it will look like formulaic in the meddle of the battle). It's more than maynarding and managing saturation of bases is one of the most visual part of macro, the most easy to follow and understand as a spectator.
Getting read of the so-called "mindless clicking" part of the macro is quite controversial, but removing that kind of highly refined and strategically oriented macro-managing is a bad idea entertainment-wise.
edit: wording: I'm not a native english speaker at all...
The big issue isn't that it's different - it's that it reduces the strategy surrounding expansions and resource management in several noticeable ways. Since your income will now translate fairly linearly to (# of workers * N) instead of scaling down after the first set of workers/patch, expansions simply serve to increase the total number of workers that can mine at once rather than increasing the efficiency of existing workers. It makes expansions less useful and encourages rushes and early aggression, while cutting down on a few interesting decisions in base-building and resource management.
I am fine with changes to worker AI and saturation - it's the completely linear progression that I'm not a fan of. As opposed to "new thinking" that you described, it's really looking to mean "thinking less".
If total saturation could be reached with a static 2 workers per mineral patch, it would lead to a serious dropoff in required macro in the same vein as WC3; 5 workers per gold mine, period, and anything further is wasteful. Granted the number for an entire base or expansion would be much higher than 5, but a player would still find himself counting the mineral patches, multiplying by 2, and making that number of workers his limit for mining minerals from that expo. It would create yet another skill ceiling in a game that appears to be full of them already. I hope this is not the case. Although if it were, it would make expanding more crucial, as it would be even harder to fight with 1 base vs. 2 or 2 base vs. 3 than it already is. It would probably clean up gameplay a bit if I were honest with myself, but at the risk of making it too linear.
I do disagree, as I said, with your claim that it would make expanding LESS useful. If both our mains had, say, 10 mineral patches (just for the sake of argument), we could both agree that the ceiling for mineral mining is 20 workers on minerals for each player. If the ceiling were not an absolute 20, I could continue on to 25, even 30 workers from my one base after you expand, and perhaps hit you with some sort of timing push before your second base really kicks in and gives you an insurmountable advantage. But with total saturation at 2 workers per patch, I would have to stop at 20 works and be forced to either attack as soon as you expand, or expand with you, because continuing to make workers beyond the initial 20 would be completely pointless.
The upside of this, however, is that we would be seeing MORE expansions (most likely). With a sort of "cap" on the number of workers mining minerals just like there currently is with gas, as soon as you reached that cap, you would be motivated to expand. This could lead to something not often seen in SC, but commonly seen in games like C&C3; fighting over a single expansion, not because of a tactical position it holds, but because both players NEED the resources. In SC, most maps have enough expansions "designated" to either player that the game is decided before any given player has exhausted all of "his" expansions and has to try to kick his opponent off of one of their own just to continue production. Imagine if in SC2, to get comparable production to what you would in SC1, you had to have 5 bases to every 3 you would have in SC1. On a map with only 10 "bases", after your 5th, to keep macroing up, you would have to attack one of your enemy's bases or fight for one he hasn't taken yet, not to kick him off of it, but because you actually NEED it yourself. It would be like playing SC on a map where each player has only a main and a natural, and there is a neutral base in the center. Obviously for SC this wouldn't be balanced, but it is the principle I am trying to get at. Both players would find themselves funneling units into the center in an earnest attempt to secure the base. The fight would not be a fight just for the sake of it, but a fight over the one expansion to tip the scales; whoever is forced to retreat will do so knowing he will be forever at a disadvantage, economically speaking. This would be an interesting mechanic and, I argue, unique to SC because of the large number of bases usually available in the original game. A situation of both players fighting over an expansion, where both players have an equally strong intention of actually taking the expansion and making full use of it after the fight is over, is something somewhat unfamiliar to SC that is seen in many other games, and would add something interesting to SC2. It might add other inherent problems, just as skewing games in the favor of races that have comparably stronger early-game, or making it more difficult for those like Zerg, who simply need more bases than the other, but all of these balance issues could likely be worked out with clever map-making, as was and is the case in SC.
I do disagree, as I said, with your claim that it would make expanding LESS useful. If both our mains had, say, 10 mineral patches (just for the sake of argument), we could both agree that the ceiling for mineral mining is 20 workers on minerals for each player. If the ceiling were not an absolute 20, I could continue on to 25, even 30 workers from my one base after you expand, and perhaps hit you with some sort of timing push before your second base really kicks in and gives you an insurmountable advantage. But with total saturation at 2 workers per patch, I would have to stop at 20 works and be forced to either attack as soon as you expand, or expand with you, because continuing to make workers beyond the initial 20 would be completely pointless.
You're missing the point of why this discourages expansion: expansions would still produce approximately the same "additional" boost to economy over time, but the immediate impact of an expansion would be less. This translates to a larger timing window to punish an expansion and a longer period of time before the expansion gives you a significant econ edge.
It's not that expansions will be less valuable (in the mid-lategame, they might be more valuable since it is easier to fully saturate an expo) - it's that expanding is significantly riskier as an early-game prospect (since it has less immediate impact and a larger window of weakness), which encourages fast, early, repeated engagements in the style of Broodwar ZvZ, which a lot of people hate.
On January 01 2010 10:51 errol1001 wrote: In Starcraft, there is no clear 'optimal' number of miners per mineral patch. Thoughts?
Yes there is, it's 3. 24 workers on 8 mineral patches is considered the best/most efficient saturation :}.
wrong, it depends a lot on the mineral formation, distance from HQ, and also positioning in relation to HQ because the sides are larger on some sides. So it varies from map to map and location to location on each map. Hell even race and upgrades(lair/hive) play a minor role. I've always used 2.5 workers but I still overproduce in order to maynard. (good example is the side minerals on bloodbath , you can put 9 on them if you want but 8 is optimal 2.5x3 = 7.5 (round to 8).
Saying "maynard" instead of transfer is so annoying. Imagine if every time someone made a dropship they called it a "Boxer" and so on and so forth.
Anyway, just because there would be somewhat of a ceiling, if you are planning on a fast expand, you can still make the extra workers and let them mine anyway while you wait for the expansion to finish. I don't think this change would really punish expanding like you say, and it certainly wouldn't turn all the matchups into ZvZ.
Why not? It's an objective fact that BW expos have a large, immediate, nonlinear % boost to economy almost as soon as they go live, whereas in SC2 that gain will be linear and smaller as a total % of economy.
Furthermore, with macro mechanics online at the main base, this % figures to be even smaller than what we're discussing at the moment.
The math is there - if both the first and second set of workers are equally efficient, adding an expo before reaching oversaturation won't have as big an economic impact unless you can build an abundance of idle SCVs to quickly transfer over.
doesn't this mean that if you want to mine optimally you should put your first 6 workers on the 3 closest patches and after that you should add 2 workers to the then nearest patch after that and so on?
also i think the nonliniear mining in BW is overestimated by most In this topic. as this picture shows the mining proces in BW is also liniear up to 2 workers per patch
the pathfinding issues in BW kick In if you keep adding workers beyond 2 workers per patch. it will still change the gameplay somewhat. because if your base gets satured faster it becomes less viable to use one base strategies.
another change is that if your FE is destroyed it will be harder to come back because the workers you have build from your expo are less useful because you can't keep adding workers to your main. If your opponent is able to contain you and get an expension running he will soon be able to outproduce you. but wasn't that the case in BW anyways?
That's not linear at all, though. It's near-linear function for workers already beyond the 1-per-patch base saturation. But with 9 total workers, we are seeing ~3k/5 mins, as opposed to ~4500/5 mins at 18 workers (2/patch).
Also your main being saturated faster actually makes 1-base strategies more viable, since you have more money faster to bust any sort of quick expo attempt, even after they get their expo up and are making back the money they spent. Furthermore, the macro mechanics give you an additional boost to your main base economy that have to be replicated at your expo in some way to gain the same level of returns.
It wont make a huge difference, 2 workers only saturate optimal patches and I think it will be the same in SC2. A single worker on an optimal patch can mine about 72 minerals per min in SC and a sub-optimal patch generates less. The maximum a patch can produce in one minute (Fastest) is about 140 so optimal patches, of which there will be 2 to 3 on a standard 9 minerals base require 2 workers and the rest require 3, in theory you can saturate a Python main with 24 workers, in practice you need 27 due to wandering. Wandering makes a slightly random mess of the mineral gathering numbers between 23 workers and 27 to 28, depending on the mineral layout, with gaps seeming to help reduce wandering. Wander can cost you up to 200 minerals per minute and can take up to 5 minutes to settle down if there are sufficient workers to saturate. If there aren't then it never will. Workers who do not wander will not move much to reach unused patches, as a result the efficiency might be slightly lower as they will just sit in place waiting for the current patch to become available. So it's not clear how the SC2 curve will go, it will top out sooner but you will still want some workers for Maynarding (interestingly it may be worth using distance mining with the excess workers in SC2 while in SC it's too much of a risk of destabilising your workers into wander mode) and may produce less below saturation and above 2 per patch.
Yes there is, it's 3. 24 workers on 8 mineral patches is considered the best/most efficient saturation :}.
No, 3 workers per patch will completely saturate it once wander settles down, which happens quite quickly on 8 and 7 mineral bases as far as I've seen. In theory fewer workers will work but it depends on the exact dynamics of the minerals in a way I don't understand, you can rarely remove more than 1 and removing a builder in this state would destabilise the workers into wander mode, costing you a lot of minerals.
2 workers per patch only generates 70% of the minerals generated by 3 workers per patch and the magic 2.5 per patch people refer to is important on 9 mineral patches at least because this is a strange point where bad wander sets in. 22 workers usually generates MORE than 23 due to wander. Round your 2.5 down if you want to use this rule of thumb. If you're Protoss with a stable 27 worker main saturation it might be worth having a builder probe or to take probes to build from the natural as on bad spawns (9 o'clock Python) it can take up to 5 minutes for the wandering caused by removing and returning probes to settle down, costing you 200 minerals per minute.
from limited PAX experience, harrass is alive and well. Because the game is so un-figured-out, it's very similar to early BW, very fast and brutal games. In my beta-key winning game I essentially won it with a muta harass, which easy dropped him off economically. Because macro is easier, nobody has a bank of money they can fall back on to replenish workers, also, because of what is mentioned, where you don't have more workers then the optimal at each expand, you don't generally have extras to transfer back to a harassed expand. All together, this means that after a harass you have the choice of trying to recover your econ, or making more units, giving the opponent an amazing attack window while you get your economy back in swing.
Maybe the danger of harass will scare people into either keeping a small amount of resources in reserve for remaking workers or making extra workers so they can transfer them back when they have dealt with the harass(assuming of course there is an expo).
I think it will add a little depth to the game because you will have to plan worker creation better as opposed to just following the SC1 rule of thumb, always make workers.
On January 27 2010 12:29 Zack1900 wrote: Maybe the danger of harass will scare people into either keeping a small amount of resources in reserve for remaking workers or making extra workers so they can transfer them back when they have dealt with the harass(assuming of course there is an expo).
Harass never scared SC1 players into doing either of those things, because if you did those things you would fall behind. I'm not convinced that a stronger SC2 harass will affect that either.
For one, extra workers are nigh useless unless you're counting on some of your workers dying. So, in other words, you're spending 150 minerals or more for extra workers that you could have used on, say, a photon cannon. The workers can't defend themselves and so by skimping on defense (unit or building-wise) you're only opening yourself to harass more, while building more defenses prevents harass now and in the future.
So people still won't be building extra workers.
As for holding onto a resource stockpile, people will not be doing that either. Think about it: it's like the worker option only you're most likely delaying your tech or your production while at the same time skimping on defense. If he drops hellions behind your probe line your minerals in the bank certainly aren't going to help you defend your probes, and in the meantime you've weakened your army and made it more vulnerable to a push. Might as well spend them on some defenses so that he can't harass you.
Basically, the principle is: the best strategy is nearly always to reduce worker loss rather than count on just replacing them. And the best stratagy is always to use your resources quickly so that you can capitalize on any resources you have as soon as possible, otherwise that resource difference will set you behind from the other player.
On January 18 2010 18:51 LF9 wrote: Saying "maynard" instead of transfer is so annoying. Imagine if every time someone made a dropship they called it a "Boxer" and so on and so forth.
Cept a lot of people call it maynarding and no one called that "boxering"? LOL?
As long as the worker harassing is efficient and economical, it will still be done. I don't see how this could tip it towards not being worth it, though obviously it will be less devastating. Can't expect everything to be the same SC1 though ^^
I sadly don't have a beta key but from what I've seen the minerals shimer when they are mined (even for Zerg) so it won't be hard to build the optimal number of workers. kind of sad really
EPT
![[image loading]](http://img123.imageshack.us/img123/7841/picture2ok6.png)
