|
Introduction
This introduction will cover the absolute basics of mining within SC2, as well as the idea of maximum and optimal mining. If you have a grasp of these simple ideas, feel free to skip the introduction:
+ Show Spoiler +Within SC2, all races have the ability to increase their income rate by building workers. All workers have 2 different phases of mining; Travel Time and Harvesting Time.
Travel Time refers to the amount of time it takes for the worker to return the collected resource to the base. Harvesting Time refers to the amount of time it takes for the worker to collect the resource from the specific source. Travel Time and Harvesting Time combined can be referred to as a single cycle of mining.
As you increase the amount of miners on a base, the amount of mining cycles increases per set period of time, resulting in a higher income. Adding more workers to a base is referred to as saturation or saturating the base. There are two saturation level that can be obtained in the current form of SC2 per base.
At 16 of 24 workers, all workers mine as efficiently as possible. When one finishes its Harvesting Time within the mining cycle, the 2nd worker on that patch finishes or nearly finishes it's Travel time. If you add additional workers beyond this point, the workers will have to wait between Travel and Harvest time, resulting in a diminishing return on workers for the base. The point at which no workers suffer from diminishing returns on income is referred to as the optimal saturation point.
At 24 of 24 workers, all mineral patches are fully saturated. Any workers after 24 will have no substantial difference on the income rate earned from this base. The point of this plateau is referred to as the maximum saturation point.
Asymmetrical Mining refers to the ability to gain an ever increasing amount of resources from having more bases. Currently, the game supports symmetrical mining at a 3 base economy for all races, which means that cost-efficiency of units is the ultimate determining factor of success. Asymmetrical mining is a more appropriate method of economy, as it allows a player to use their economic advantage to throw away less cost effective units to whittle down a more cost effective army. Losing these units does not matter as much, as the person with the more powerful economy can continue to whittle, while the lesser economy individual loses their cost-effective army.
Hypothesis
I hypothesize that increasing the Harvesting time of the workers will result in a large enough decrease in income rate, in which more bases beyond 3 will be required to optimally mine with the limited supply cap of workers.
Procedure
Belshir Vestige LE 1.0 was downloaded using the SC2 Editor for purposes of testing out this hypothesis.
In order to determine the effect of the experiment, two different controls were used. Using the 'Test Document' feature, 24 workers were built out of the town center to establish maximum saturation income rate. The amount of minerals gathered to a set time were recorded; the set time was established as t=0. The amount of minerals gathered after 1 minute was recorded, and the 1 minute trial was repeated in triplicate. In addition, a 2nd control of 48 probes spread across 3 bases (according to optimal saturation) was used to determine the current optimal income rate on 3 bases. As above, a set time was established as t=0, the minerals gathered to that point were recorded, and 1 minute later, recorded again. This was also completed in triplicate.
After determining the income rates of the standard game procedure, the map was altered in a way that probes, drones and SCV's would spend additional time on the mineral patches, gathering minerals. This was accomplished by modifying the Gather ability in the data section of the SC2 editor. The field 'Stats: Resource Time Multiplier +' was altered from 1, to a value between 1 and 2.5. This multiplier acts on the harvesting time, by multiplying the standard time by the coefficient provided by this field. The standard game has a coefficient of 1. As above, after the number of workers required were established, a set time was set to t=0, the minerals gathered to that point were recorded, and 1 minute later, recorded again. These were all done in triplicate.
Data and results
The following data has been obtained thus far:
Controls
+ Show Spoiler +
Control - 24 workers, 1 base, maximum saturation, Resource Time Multiplier = 1
Time(mins) Minerals Gathered Income
05:30 990 -
06:30 1800 810
07:30 2620 820
08:30 3445 815
Control - 24 workers, 3 bases, unsaturated; Resource Time Multiplier = 1
10:00 3940 -
11:00 4960 1020
12:00 5980 1020
13:00 7010 1030
Control - 48 workers, 3 bases, optimal saturation; Resource Time Multiplier = 1
08:00 3220 -
09:00 5160 1940
10:00 7110 1950
11:00 9040 1930
Experimentals - 1 base
+ Show Spoiler +
Experimental - 24 workers, 1 base, maximum saturation; Resource Time Multiplier = 1.5
Time (mins) Minerals Gathered Income
6:00 955 -
7:00 1530 575
8:00 2090 560
9:00 2650 560
Experimental - 24 workers, 3 base, unsaturated; Resource Time Multiplier = 1.5
10:00 3205 -
11:00 4065 860
12:00 4915 850
13:00 5775 860
+ Show Spoiler +
Experimental - 24 workers, 1 base, maximum saturation; Resource Time Multiplier = 2
6:00 155 -
7:00 575 420
8:00 995 420
9:00 1415 420
Experimental - 24 workers, 3 base, unsaturated; Resource Time Multiplier = 2
10:00 1865 -
11:00 2590 725
12:00 3315 725
13:00 4035 720
Experimentals - 3 base
+ Show Spoiler +
Experimental - 48 workers, 3 base, optimal saturation; Resource Time Multiplier = 1.75
10:00 2090 -
11:00 3520 1430
12:00 4960 1440
13:00 6400 1440
Experimental - 48 workers, 6 bases, unsaturated; Resource Time Multiplier = 1.75
15:00 8880 -
16:00 10440 1560
17:00 12010 1570
18:00 13580 1570
+ Show Spoiler +
Experimental - 48 workers, 3 bases, optimal saturation; Resource Time Multiplier = 2.5
14:00 2955 -
15:00 3980 1025
16:00 4980 1000
17:00 6015 1035
Experimental - 48 workers, 6 bases, optimal saturation; Resource Time Multiplier = 2.5
19:00 7970 -
20:00 9200 1230 (outlier)
21:00 10500 1300
22:00 11815 1315
23:00 13125 1310
Conclusions and Interpretations
Based on the data, we can conclude that increasing the harvesting time of the workers by a set coefficient (The 'Resource Time Multipler' field) does affect the overall income. In addition, spreading workers across more bases has a greater effect when the Resource Time Multiplier is increased. However, this multiplier cannot increase indefinitely, as increasing beyond 3 would essentially make maximum saturation at 8, regardless of other workers. A value somewhere between 1 and 2 is likely ideal, as it does not punish 1 base play as much a higher value, but still has a significant effect on asymmetrical mining effects.
A second noticeable effect of increasing the 'Resource Time Multiplier' coefficient is a decrease in the overall income earned per minute. However, this is easy enough to fix, as there is also a 'Resource Amount Multiplier' option within the gathering ability field, indicating that the income lost through increasing this multiplier can be regained by increasing the amount workers gather per trip.
Further Considerations
This was a very rudimentary experiment, just to test one variable within the SC2 editor. However, I would claim that the results obtained indicate a successful hypothesis test. From here, it would be reasonable to continue adjusting the Resource Time Multiplier to find an appropriate difference between playing on smaller bases versus playing on larger numbers of bases.
|
You figured out that increasing the time it takes for a worker to mine also decreases your income? I could've told you that without having to test anything... Am I missing something, or what? I'm not really sure what the point of this is...
|
No. I experimentally figured out how to recreate the Asymmetric mining effect from BW without dumbing down worker AI. If you look at the experimental results, you can see that spreading the workers across more bases gave an increase in mining, as opposed to current symmetrical effects. Therefore, increases the time it takes for them to mine reintroduces the asymmetrical mining effect.
|
You use 24 workers per base for 1 base, yet 16 per base for 3 base. I don't understand why.
|
On October 25 2013 12:39 Rasera wrote:
No. I experimentally figured out how to recreate the Asymmetric mining effect from BW without dumbing down worker AI. If you look at the experimental results, you can see that spreading the workers across more bases gave an increase in mining, as opposed to current symmetrical effects. Therefore, increases the time it takes for them to mine reintroduces the asymmetrical mining effect.
well yeah its quite logical actually^^
Right now workers dont travel as much inbetween the minerals cause they mine at the "perfect speed".
If you slow them down they will travel more and you will get more income when you spread them out, so yeah ^^
|
On October 25 2013 12:34 Penguinator wrote:
You figured out that increasing the time it takes for a worker to mine also decreases your income? I could've told you that without having to test anything... Am I missing something, or what? I'm not really sure what the point of this is...
What hes trying to attain is a broodwar style of mining. The longer it takes a unit to mine, the longer the 2nd harvester has to wait before he can begin mining. Thus making it slightly more optimal to have less saturation and more bases where you can focus 1 worker on 1 patch instead of wasting time waiting for the previous harvester to finish
|
For what it is worth, I really feel like a change in mineral patches or mining distance would greatly improve the decision making variety in way that helps the game in every way for every person.
|
On October 25 2013 12:42 iTzSnypah wrote:
You use 24 workers per base for 1 base, yet 16 per base for 3 base. I don't understand why.
I use 24 workers on 1 base to establish maximum saturation value. I use 16 per 3 base, as the consensus on optimal saturation is 16 workers on each base for 3 bases. Too much above that and you tend to neuter your army supply.
On October 25 2013 12:42 The_Red_Viper wrote:Show nested quote +On October 25 2013 12:39 Rasera wrote:
No. I experimentally figured out how to recreate the Asymmetric mining effect from BW without dumbing down worker AI. If you look at the experimental results, you can see that spreading the workers across more bases gave an increase in mining, as opposed to current symmetrical effects. Therefore, increases the time it takes for them to mine reintroduces the asymmetrical mining effect. well yeah its quite logical actually^^ Right now workers don't travel as much in between the minerals cause they mine at the "perfect speed". If you slow them down they will travel more and you will get more income when you spread them out, so yeah ^^
Well, I understand it's logical, but a lot of people spit on logic without hard evidence. And at least this way, it's a start towards asymmetrical mining and games being more spread out to accommodate for more bases.
|
On October 25 2013 12:47 vaderseven wrote:
For what it is worth, I really feel like a change in mineral patches or mining distance would greatly improve the decision making variety in way that helps the game in every way for every person.
I originally considered this, but you get some very weird looking mineral patches if you try to make a larger distance, or similarly if you move it too close, then aoe on workers becomes much too effective.
I decided I would try to target the harvest time instead.
|
Essentially the number of workers required for one base saturation is lowered, and the max income rate per base decreases too.
|
On October 25 2013 12:47 Rasera wrote:Show nested quote +On October 25 2013 12:42 iTzSnypah wrote:
You use 24 workers per base for 1 base, yet 16 per base for 3 base. I don't understand why. I use 24 workers on 1 base to establish maximum saturation value. I use 16 per 3 base, as the consensus on optimal saturation is 16 workers on each base for 3 bases. Too much above that and you tend to neuter your army supply.
16 workers per base is the optimal saturation, that is not dependent on how many bases you have. All you did by using 24 was make the 1 base results non comparable to the rest of the experiment.
|
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal?
|
On October 25 2013 12:53 mewo wrote:
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal?
As I understand it, 3 bases of 16 mining workers provides a sufficient income to build up your army, while still allowing for a sizeable army. 16 more workers onto an extra base provides extra bank basically, at the cost of 16 army supply.
I suppose everyone has their own preferences, but I used it as a baseline for this experiment.
|
On October 25 2013 12:52 iTzSnypah wrote:Show nested quote +On October 25 2013 12:47 Rasera wrote:On October 25 2013 12:42 iTzSnypah wrote:
You use 24 workers per base for 1 base, yet 16 per base for 3 base. I don't understand why. I use 24 workers on 1 base to establish maximum saturation value. I use 16 per 3 base, as the consensus on optimal saturation is 16 workers on each base for 3 bases. Too much above that and you tend to neuter your army supply. 16 workers per base is the optimal saturation, that is not dependent on how many bases you have. All you did by using 24 was make the 1 base results non comparable to the rest of the experiment.
I set the 24 workers on 1 base as a baseline against the control of 24 workers on 1 base. It was meant to illustrate that increasing the mining time increased the effectiveness of spreading workers out across other bases. I suppose I illustrated it poorly though.
|
The advantage of the BW econ isn't that it forces you to take bases faster. Its almost completely the opposite actually. In BW the immobile race can stay on few bases, and still have a decent income with "oversaturation". Meanwhile the mobile race can get an economic advantage by being 1-2 bases ahead. Sc2 econ differs from BW econ as it makes it possible for the immobile race to even out the economic disparity by playing a defensive economic-based game.
Thus it doesn't matter whether you as as mech'ing player is behind 4base to 5 base, you will still have a similar econ. In BW your econ will be slightly inferior, which means it incentivies you to stay on 3 bases and attack rather than attempt to take a 4th base. This means;
- BW econ --> Incentivies the immobile race to attack
- Sc2 econ --> Incentivies immobile race to turtle.
This is extremely important to understand, because Sc2'ish problem is - completly contrary to what you believe - that you can't actually attack on 2-3. Just go and watch some BW vs Sc2. You notice that terran in TvP BW often times will stay forever on 3 bases. In TvZ he will often stay a long time on 2 bases. This makes it possible for the immobile race to attack while not worring about counterattacks --> Opens up for offensive games.
So really the reason why Sc2 is often times such a boring turtleparty is that the threat of counterattack is way too high. This is partly due to the fact that the economy rewards you for taking a 4th base instead of staying on 3 bases and also due to the efficient clumping of units.
Asymmetrical mining is a more appropriate method of economy, as it allows a player to use their economic advantage to throw away less cost effective units to whittle down a more cost effective army. Losing these units does not matter as much, as the person with the more powerful economy can continue to whittle, while the lesser economy individual loses their cost-effective army.
We don't have assymetrical mining in Sc2, but we have really really assymetrical economy. Terran in sc2 --> much better econ than protoss. Zerg alot better than protoss etc.
Both those matchups are - contrary to what your theory would predict - awfull'ish turtle. BW econ isn't about FRB or differences in economic mining efficiency early game. Instead it is about making a difference in 4base to 3base income while maintaining A) An even larger difference between 5base to 3base income and B) Roughly matching the strenght of each race in the early game --> So all of the races can play offensive early game.
|
On October 25 2013 12:59 Rasera wrote:Show nested quote +On October 25 2013 12:53 mewo wrote:
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal? As I understand it, 3 bases of 16 mining workers provides a sufficient income to build up your army, while still allowing for a sizeable army. 16 more workers onto an extra base provides extra bank basically, at the cost of 16 army supply. I suppose everyone has their own preferences, but I used it as a baseline for this experiment.
Dont forget the 6 workers mining gas, thats total 66 workers on 3 bases.
|
On October 25 2013 12:53 mewo wrote:
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal?
And 22 if you include gases. That should be the case but if you look at say, a typical protoss or terran game, getting on 4 base with optimal saturation means you need 88 workers and that limits your army supply to be 22 less than your opponents, though that's all very elementary.
Unless you're zerg and plan to make 10-100 spines later on or something, not many people (no evidence, but I think it's quite apparent) would want to cross the 70 worker mark. That makes staying on 3 base, until stuff gets mined out, very viable even against players on more bases (though eventually increased gas income might give the person with more bases an advantage). Some people also feel this also has a lot to do with the symmetrical supply armies that all races can create (200/200), though that isn't relevant to this discussion.
From what I've read, this seems to be the "problem"..
No significant income advantage except in gas + Danger of dying while spending money / positioning army to establish 4th = little to no incentive for players (particularly P & T) to get fourth / fifth base unless the main / natural is getting mined out.
I don't really know how different BW resource collection is from SC2 but it will defo take a lot of testing before such a big change can be made. At least the problem is being identified!
|
On October 25 2013 12:53 mewo wrote:
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal?
Doesn't Riot do this every new season in League of Legends? What am I missing here? Has Blizzard dropped support for StarCraft 2? I was under the impression they wanted to make StarCraft 2 the best eSports game possible...
Sure, it might shake up the balance. It would also make the game way more interesting to watch, which greatly outweighs the negatives. Balance patches can come after the changes.
|
Hi ras, nice writeup. I like SAIYENTS. :D Below I pick apart a lot of it at length, but my intention is to help you out. ^^
This has been discussed quite a bit in the past, have you looked at any old threads about mining in SC2? You might save yourself some experimenting if you find old data, or change your mind on how you want to address the problem. Also, the SC2BW mod has an AI workaround that achieves BW mining pretty much anyway. While I agree with the sentiment that'd it'd be great to achieve a similar effect with a simpler tweak, I don't think it's possible. Did you investigate the overall behavior of the workers after increasing harvest time? I'm pretty sure they will just pair up as normal, not "bounce" as in BW, which is the cause of the economic benefit of extra bases in that game.
To address what you did present:
I hypothesize that increasing the Harvesting time of the workers will result in a large enough decrease in income rate, in which more bases beyond 3 will be required to optimally mine with the limited supply cap of workers.
I would recommend you look at this statement again and try to tease out all the assumptions and pieces of reasoning that are hidden in it, because it's really saying a lot more than it's meant to I think. Working from the back, the "worker supply cap" isn't a clear idea, because surely you're not referring to having 200/200 workers. If you're referring to "around 60-70 workers", that is a strategic choice that depends on a multitude of factors and couldn't really be pinned down for the purposes of such a thesis. If you mean the number of workers for saturation, that's a simple statement (good), but it doesn't necessarily have anything to do with "optimal mining" in the context of having more than 3 bases. That, again, would be a strategic factor that you couldn't determine offhand.
Vaguely, I understand the intention but the idea is not precisely stated. To back up further, "more than 3 bases will be required" makes sense if you're saying that given a certain number of workers that saturate 3 bases in normal SC2, lengthening the harvest time will have the result of requiring additional bases to achieve "optimal" mining. Since this is what I think you meant, I'll leave it at that. But I'll point out that this is not really the stated goal of the larger project.
The first clause "increased harvest time will lower the income rate enough", isn't really clear in how this will create the outcome you hypothesize. The way I think you mean it, you're correct, but if you state it more precisely you'll see why this happens in a way that probably won't lead to a good final solution to adjusting the SC2 mining "problem". Specifically, the reason you need more bases is because you're oversaturated when the harvest time is increased. The oversaturation workers need to be moved to another base. At that point, you'd have the same income as you would normally on 3 bases, but you'd have 4.
To be even more explicit, increasing the harvest time will allow workers in transit to return to the patch by the time the other is finished harvesting and starts their transit. Overall, the income will be decreased because minerals/time has been decreased. However, this will also cause the patch to be saturated with only 2 workers (if the harvest time is increased sufficiently). This really just means bases will become saturated faster. It's really just another version of the Fewer Resources per Base project. If you increase the minerals harvested per trip to make up for the slow income rate, you've really just recreated gold bases.
There is a subtler point though for the case where the saturation isn't actually reached at 2 workers, but only moved in that direction. This would mean the returns of a 3rd worker on a patch is really minimal and you'd be much better off putting that worker at another base. This is what I assume you're going for, and what you are trying to say by "optimal mining". However, you must realize that this is already the case. We don't ever see players intentionally go above 18-20 workers per standard 8m2g base, certainly not the technically full saturation 24. That's because they'd send them to another base where they are actually giving you reasonable income. So the game is already operating the way we would want.
Why then do we have the supposed 3 base cap? It's not strategically worth it to create another vulnerability that won't give full returns. And it's not worth getting full returns from 4 bases at once because the supply you tie up in workers for 4 bases will leave you unable to defend them. Of course there are gas considerations in certain matchups but you see the point.
So, realizing I'm sure you've already been through this, in theory depressing the saturation and requiring worker supply to go to a 4th base sounds good, but it's going to mess up game balance no matter how you accommodate it with counter-tweaks, which is probably the biggest strike against it. I think in order to make a point, you'd essentially need to test game balance in an environment with adjusted mining, and of course that is a huge challenge.
Anyway, I hope this helps and I'd be glad to discuss any of it.
edit: really interesting post, hider. thanks.
|
On October 25 2013 13:34 Arco wrote:Show nested quote +On October 25 2013 12:53 mewo wrote:
You'd have to rebalance the whole game. Maybe for sc3, but for now isn't 16 on each base enough to make 4 base ideal? Doesn't Riot do this every new season in League of Legends? What am I missing here? Has Blizzard dropped support for StarCraft 2? I was under the impression they wanted to make StarCraft 2 the best eSports game possible... Sure, it might shake up the balance. It would also make the game way more interesting to watch, which greatly outweighs the negatives. Balance patches can come after the changes.
So naive. Give Blizzard a few million dollars, and maybe they will, 2 years after LotV is released.
I don't even know how you can ask a question as dumb as "has Blizzard dropped support for Starcraft 2" when they are developing an expansion at this very moment.
And never ever again compare SC2 to a f2p game.
|
|
|
|
|
|
EPT
