|
Something interesting that I picked up from this post is the comparison between Banshees and Vikings. Per minute, constant Banshee production costs you less than constant Viking production.
With constant production, you can get out 1 Viking and ~42.9% of another Viking. Basically, about one and-a-half Vikings per minute equates to 1 Banshee per minute. Thing is though, 1 Banshee has twice the DPS of a Viking; A Viking's 10 DPS vs. a Banshee's 19.2 DPS.
Overall, if you choose constant Banshee production versus constant Viking production, you get about 30% more DPS per minute- which is relevent since you want as much firepower as possible, in as little time as possible- 15 more HP and all of this at a lower cost per minute, making it more sustainable with a limited amout of bases.
Another pro to choosing Banshees over Vikings is that they are great combatants and do not fall into one role, like Vikings do, they can harass effectively and fight head on in an army on army conflict. The only con would be that they have 3 less range than Vikings. The only reason I decided to voice this observation is because there is a thread that I read earlier promoting the use of Banshees over Vikings to combat Colossus based armies.
http://www.teamliquid.net/forum/viewmessage.php?topic_id=202210
Think what you will, I just found it interesting that +3 range makes so much of a difference that Terran's would choose Vikings over Banshees, when Banshees are all around better. And what with Protoss opting for more powerful Gateway compositions, Terrans need as much Anti-Surface firepower as possible.
Another possiblilty would be to produce one Viking and one Banshee at a time off two Starports rather than just one or the other off their respective Starport configurations ( 1 Reactor versus 2 Tech). Just some food for thought.
|
Markwerf, when your insulting someone by calling their post ridiculous, try to spell ridiculous correctly.
Other than that, your comment is right on.
Griffith, you post isn't really looking at the 'cost' of units in a meaningful way. As Markwerf said, it's good for a unit to build quickly because that reduces the infrastructure costs in building that unit. If constant marine production out of 1 barracks drained all of my income off of 3 bases, that would be awesome. I'd make exactly 1 barracks and win every game. Of course, your calculation would make marines look very expensive when, in fact, they're very cost-effective.
What the chart is better for is looking at how good buildings are. Assuming unit strengths are in line with their costs, the more resources per minute I can use with a building, the better. For example, gateways, costing 150 minerals, were too good of a mineral sink when zealots built in 33 seconds, and then 23 after WG research. Using your formula, that's 326 minerals per minute of production from a 150 mineral warpgate. That production was too efficient, so it was nerfed.
Similarly, you'll notice that the factory units are consistently more expensive per minute than the barracks units. A factory has more production capacity than a barracks. This is offset, however, by a factory having roughly double the cost of a barracks.
So if you want a more useful chart, take out the supply costs, and analyze how effective each building is as a resource sink, and then look at the costs of those buildings.
|
This is useful for balancing production off a static economy. Not much else.
In fact, there are a number of online calculators that can help you with this. Here's my favorite:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=155279
In SC2, usually economies are growing. So you need to account for the present cost of the unit compared to the future value. For example, 100 minerals at 5:00 is worth less than 100 minerals at 10:00 in a growing economy since it's a smaller percentage of the economy later than at start, so there is less opportunity cost.
All SC2 costs are investments, meaning they are paid up front and the returns come after a delay of building time. The longer the delay, the more the economy grows between start and finish, and so the less ultimate comparative value the units or upgrade has.
We need an economist to write up a real analysis of SC2 economics, including ideas like present value, "inflation", investment ROI, risk/reward.
|
Yea kcdc has got it right.
Which presents an incredibly confusing fact that you put "Structures" and figured out their "cost" or whatever. You won't be constantly making command centers. It just doesn't make sense for structures.
This only demonstrates how much production you can support on x bases. For instance, reapers are very low so going 5rax reaper is perfectly viable with additional money leftover for expanding. But 5rax marauder just isn't going to work.
|
On March 19 2011 03:54 Thorn Raven wrote:
Think what you will, I just found it interesting that +3 range makes so much of a difference that Terran's would choose Vikings over Banshees, when Banshees are all around better.
Why are you comparing Vikings and Banshees based purely on DPS? I understand when you are talking about just going against colossus armies, but vikings and banshees are inherently different. Vikings shoot only air units, banshees can't shoot air units. I guess make it clear in the beginning of your post that you are talking EXCLUSIVELY about fighting colossus, as they are the only unit that both the viking (not landed) and the banshee can shoot simultaneously.
|
I think another problem with this analysis is the assumption that production always occurs near the supply cap. Over the course of a game, army trades may happen, at which point the supply cost is lifted temporarily.
|
I like this analysis, and I've always wanted to articulate that economy can not be measured only by the amount of resources you're taking in per minute, but also as a function of your production capacity (infrastructure).
To give an example, I was experimenting with a 1rax ghost FE in TvT. On paper, the build was so, so good. I expanded faster than my opponent. I was able to hold off all openings my opponent could throw at me. I was losing games despite these things, though, and I was baffled as to why. Upon watching replays I realized that the reason I was losing is that I had no infrastructure going into the midgame. My production capacity was simply not there, and required a large investment to get running, so my economy could not be considered 'good'.
With a typical analysis of my economy (i.e. how many resources are you pulling in), I would be way ahead - but that's too simplistic, and turned out to be not true in that situation. Equally important to how many resources you're accumulating is the ability to liquidate that money at the same rate.
That said, this thread was worded confusingly, and it should be less about 'cost' and more about 'production capacity'.
|
While production costs per minute may not make much sense for armories/ebay, but they do make sense for factories, and barracks, I often end up with 10+ Factories by late game in my TvPs, and over 10+ barracks in my TvZs, so by that measurement, I definitely need to know the consumption of those two structures.
But you guys are right that its not so much a fixed cost analysis, but more of a cost/minute consumption (hence what I wanted to get across with normalized) analysis.
Terran's mechanics is a lot different from P/Z because we don't have "instant" production. Everything we have is in a continuous flux.
|
Oooh that unit chart is gonna be super helpful for planning midgame strategies/builds.
As iechoic said its more about the "real production capacity of bases" vs "real cost" but semantics are lame.
|
Eh, the thing is supply depot cost is mostly independent of unit cost. I'm not sure how useful that analysis really is. It's headed in the right direction, but you can't do it on a per-unit basis.
|
On March 19 2011 00:58 Griffith` wrote:
I've made a table that I found myself constantly referring to when planning Build Orders. It is fairly straight-forward and should be of use to people who don't want to do manual look-ups and google math. This is more useful for T given their production mechanics, and less so for Z/P, given their respective larvae and warp-in and chronoboost mechanics .
The table is normalized by time and food, example:
On the surface, a siege tank costs 150 minerals, 125 gas, 45 seconds, and 3 food.
3 Food is equivalent of 3/8th of a supply depot, or 37.5 minerals, meaning that a siege tank consumes 187.5 minerals, 125 gas, in 45 seconds => meaning that siege tank production will cost you 250 minerals, 166 gas per minute (though you are getting 1 and 1/3rd of a siege tank).
Note that this does NOT take into account of lost mining time of the SCV when making the supply depots/factories/etc.
Why is this useful? Because income per base is normalized to per minute, ie. 2 base Terran averages an income of of ~1800 minerals and 456 gas per minute. Suppose there was a unit that costs 1500 minerals and 1500 gas but took an hour to build, is it "expensive"? nah. Same analogy goes for upgrades, they're actually really cheap, but just have a high "up-front" cost.
Glad to see I'm not the only crazy numbers person here. I have been working on the same EXACT types of calculations for a terran two base. I have an optimal build against zerg that I am extremely successful with. It involves such:
TvZ
My SOLE mid-game Composition (Marauder Thor Helion):
2 Factory Thor: 600/400 per minute
1 Factory Reactor: 400 per minute
3 Rax teched for Marauder 600/150 per minute
Total ratio: 1600/550.
Analysis: A little gas heavy considering you need the buildings and upgrades, but right around the ball park in terms of that 3:1 ratio of mineral/gas. The build is simply my only working formula for zerg and it works wonders for me.
Now I am looking for an optimal protoss build that does the same. I am thinking:
TvP
My Hypothetical Composition: (Marine, Thor, Tank, Helion with Raven/Banshee/Viking support)
2 Reactored barracks: 480 per minute (note that if you need the marine upgrades, one of your rax will have a lab for extended an time period)
1 Reactored Factory: 400 per minute
1 Factory Tanks: 200/167 per minute
1 Factory Thor 300/200 per minute
1 port (Raven, Banshee, Viking constitutive production) Average min/gas ratio per minute: 155/149
Total ratio: 1535/516
Analysis: An early bio into full blow mech composition with marine support. Also adds ravens and banshees for air support, may include Vikings for any Collosi/Void heavy builds. Haven't implemented the build yet but I think it will work. Ratio is better than the TvZ ratio, as I will need that extra income for a wider range of upgrades.
Now I need to work out a TvT...
|
On March 19 2011 04:51 rapier7 wrote:
Eh, the thing is supply depot cost is mostly independent of unit cost. I'm not sure how useful that analysis really is. It's headed in the right direction, but you can't do it on a per-unit basis.
Well it's valid right until you have your first proper engagement, whereupon it stops being entirely useful because you have spare supply due to losing units.
And obviously when you hit 200 supply you don't need to make depots anymore either.
But it gives a useful top limit to how many minerals you will be spending.
|
On March 19 2011 04:51 rapier7 wrote:
Eh, the thing is supply depot cost is mostly independent of unit cost. I'm not sure how useful that analysis really is. It's headed in the right direction, but you can't do it on a per-unit basis.
Generally, units tend to be 50 minerals/supply. Obviously this isn't followed strictly or anything (immortals and void rays come to mind...), but that seems to be the general rule. It actually reinforces the idea that you need tech heavy units in the lategame because high gas units are more supply efficient.
|
I didn't knew taht, it's really interesting.
|
|
|
This is a fantastic table of data. I have been searching for this.
|
I like the idea of this thread mate, should make it easier to figure out number of production facilities for constant production. Nice one for compiling this
|
I like this a lot because it takes time into account. Someone needs to make another thread where time is treated as a resource because it's the most valuable one IMO. Still this is helpfull for planning BOs, but time is more important for planning strategy.
|
Basically this is a good tool for calculation of Mid Game economy management and figuring out what you can actually afford.
|
Griffith it seems that you have confused your self into believing a poorly constructed set of constructs is insightful.
But for other people here is what he is SAYING.
First he ASSUMES that any production structure that you create is constantly producing a UNIT.
And constant production of SAID unit over a MINUTE is the True cost of that unit.
The Claim that upgrades are cheap stems for the fact that the build time is LONG thus the per upgrade cost is averaged down over a period of time. Thus is the same for the thor and BC comment. No concern is given however to the amount of time it would take to save for this ? as if you were truly calculate in this fashion you would need to know that on 2 BASE your maximum GPM is about 440. And a BC takes 300 gas or 40 seconds of mining time. Similarly for these upgrades.
Perhaps this will someday be a useful line of thinking but upto this point, it is not quite there.
|
|
|
|
|
|
EPT
