[SC2] Macromanagement in Starcraft II - Page 8

On September 12 2009 20:56 Freezard wrote:
Nice article but I don't know why you guys take the beta so seriously lol... the game is nowhere near completion but I love you anyway.

On September 12 2009 21:33 dcttr66 wrote:
i got about halfway through...i don't really agree about the part where you don't have a choice...just because using the resource gathering skills is kindof pointless if you've mined out the map already.

On September 12 2009 21:54 tedster wrote:

Mining out a map happens incredibly rarely and by then the game has already been decided 95% of the time.

On September 12 2009 23:46 0neder wrote:
I have to say, if I imagine SC2 in current form w/out macro mechanics, I imagine a better game with more balance between micro and macro.

On September 12 2009 20:56 Freezard wrote:
Nice article but I don't know why you guys take the beta so seriously lol... the game is nowhere near completion but I love you anyway.

On September 12 2009 13:52 citi.zen wrote:

I am also not convinced you can scout the T that easily after the first rack is complete. They probably want to rush to mule anyway, and then make the rest of the racks, or whatever else they choose. By then no OL will be in their base, correct? [1]

By the way, the charts show mining efficiency without mule/obelisk, correct? Accounting for those would make the minerals disparity wider in favor of P/T, and much more imperative for Z to expand to keep up. [2]

On September 12 2009 13:57 integral wrote:


You make some very good points about things I ironically overlooked (and why I overlooked it) in a post about not misinterpreting the graphs, notably that zerg is spending faster than terran and protoss and this skews the graph. It's hard to extrapolate from the tail end of that graph whether dM/dt for zerg would be comparable to the terran and protoss curves, but after looking at it again I think that it would, plus it'd be pretty hard to argue that it wouldn't. [1]

Protoss' mineral mining rate w/mechanic is going to be 1.2x (not 1.25x, unless 6/5 is now 1.25... but I got the other math wrong so ) that of normal workers w/o mechanic, which as you point out falls significantly short of the advantage the zerg mechanic gives. [2]

I'd want to factor in build times and building costs to truly graph out a function dM/dt that included expansions, [3]

but as it stands I concede to the "zerg queen mechanic also enables faster economy" argument, which makes a few of my prior points invalid. (e.g. that there will be some theoretical intersect at which point the mining rate of protoss and terran would surpass that of the zerg -- and thus a point in the midgame where protoss and terran would reach a greater number of minerals mined overall. This still might happen in a real game due to other limits such as population cap and the typically limited number of mineral patches on starcraft maps, but there is no denying the current advantage zerg has both in theoretical expansion and production capabilities.)


edit: if the unit cost/effectiveness ratios are balanced, dM/dt is what needs to be emphasized and focused on in terms of economy balancing. Any significant imbalance in the dM/dt of unchecked expansion will have to be made up for with other gameplay aspects. [4]

edit two: clarity or lack thereof

On September 13 2009 00:08 JWD wrote:

I've seen this argument a lot around SC2 threads lately, and it's completely broken. The logic seems to be that, because SC2 isn't finished, there's no point in heavily critiquing it. But the only reason you can assume that the finished product will be better than what's out today is because of the critique (and resulting improvement) the game will receive! Especially in cases like this, where Blizzard employees seem to be missing a crucial balance concept, the voice of third parties like TL motivates improvements in balance/gameplay.

So, please, don't post here just to say that TL is being too serious, or too good, or too thorough at its SC2 coverage.

On September 13 2009 00:17 DefMatrixUltra wrote:
[1] The theoretical and realistic models of the mechanic might collide and disagree at strange times. The thing about the model is that the constant 'C' becomes slightly smaller as you build more drones (because there is such a thing as oversaturation). At some point, having too many drones is not advantageous enough without having an expo to use them at.

Another point that makes a Zerg fall short of its theoretical maximum is that a Zerg's build is more optimized if he builds one type of unit at a time (this is generally true). So building a drone, then a ling, then a drone is usually not the most optimal way to do it. It'd be better to go drone drone ling instead (of course this is assuming the regular course of your build and not that you're being attacked at the moment and need emergency defense). So at some point when the Zerg decides to make an army, he will pretty much stop building drones altogether. Protoss and Terran will still be getting a fairly theoretically maximal output from their mechanics because they will rarely have to stop worker production.

However, I think in a realistic sense, the Zerg is able to reach the saturation required for his build earlier because of the larva mechanic. This allows him to switch production to something else (possibly a contain or a set of defensive units to cover an expansion [like a Terran would do in BW]). So it could be that because of the playstyles of the different races, Protoss or Terran might benefit more in minerals over time (simply because trying to follow closely to the theoretical model of the larva mechanic means not building enough units to cover expansions needed). Even if this is the case, the fact that Zerg has such a powerful and flexible early game really makes me think that a possible mineral advantage later on is pretty moot. Advantages early on tend to snowball in an RTS like SC (this is often called the slippery slope effect).

[2] You're right here. For someone reason I used 4-->5 instead of 5-->6. Maybe it was just too late or something.

[3] Doing this is actually a complete nightmare. It's much easier to talk about the ideal economy and then talk about the kinds of expenses that take away from ideal economy. To actually do this, you would need a specific build order and you would need to set up a differential equation for each milestone. So from 0 seconds to first supply depot, you'd solve the DE for one function. Then from that point to Barracks, you would solve another DE for another function. And you would do this many many times and add the results all up together by evaluating the functions along the time intervals to which they apply.

This becomes pretty much impossible after the first engagement with the enemy (unless you are doing it after the fact... though that's pointless because you could just look at the replay and see how many total minerals they had mined). You can't even loosely predict what an enemy will do or even the actual outcome of a single engagement so you definitely cannot predict it accurately enough to write out a model.

Also there are some problems you run into with using continuous math with these discrete things that will affect even a 100% accurate model.

[4] This is one of the core aspects of BW to me. It's also the reason I find the Zerg so interesting because they have so much potential for economic advantage.

On September 13 2009 00:08 JWD wrote:

I've seen this argument a lot around SC2 threads lately, and it's completely broken. The logic seems to be that, because SC2 isn't finished, there's no point in heavily critiquing it. But the only reason you can assume that the finished product will be better than what's out today is because of the critique (and resulting improvement) the game will receive! Especially in cases like this, where Blizzard employees seem to be missing a crucial balance concept, the voice of third parties like TL motivates improvements in balance/gameplay.

So, please, don't post here just to say that TL is being too serious, or too good, or too thorough at its SC2 coverage.