Rush Buying - Maximise your EC today!

[Kody]

It's kindof complex you see. I wasn't advocating that rushing crawlers is the best way, rather that it is a method that is possible and useful in certain circumstances.

Whether to rush depends on a number of factors.

- Your Industry rating (crawler upgrading costs/mineral changes depending on industry)
- Your skill at managing resources (how well you can compound the effect of having extra crawlers, and how well you're able know when to rush buy)
- Whether stockpile is allowed (rush buying tends to increase the amount of stockpile energy you get)
- Amount of free terraforming you can place crawlers on (effects how much the resources compond)
- ETA for the special project you're building. (Longer you have crawlers the more the effect of the crawlers compound up)
- The number of special abilities you can add (changes the cost of crawler upgrading)
- The armor level you can add (changes the cost of crawler upgrading)

I don't actually have proper figures for anything other than +3, but I spent a few days thinking about it for other industry ratings after talking to Minute Mirage. The following are my estimations.

Assuming you're really good at managing resources and rush buying, plus stockpile energy is allowed and you have spare forest to harvest and you can only add +2 armor and trance. If you intend to build up for about 5 years I would rush buy at industry rating of 0.

With -1 industry and same conditions as above 8 years before you start saving up. With -2 industry I think crawler upgrading is where you should spend most of your money for SP building.

Ofcourse when you're borderline and getting close to the time you can think of crawler upgrading you don't suddenly stop rush building, you still rush when the cost is obviously cheap. Rather you reduce the amount you rush and tend to save up more and more money as you get closer to the time you will need money to upgrade the crawlers.

The above stated some of the more optimal variables for a rush building strategy.

Above average player, with no stockpile energy, +3 armor and trance and no spare forests/mines to harvest. The industry rating of maybe +2 +3 or above would be required so that 5 years to the build you start thinking of saving up some money for upgrading.

___________________________

When I was last in the ACDG playing as the Hive I was paying for minerals at a rate of 2.3 ECs/mineral. However due to smart resource management and the stockpile energy bug I was actually getting an estimated minerals for 1.7 ECs/mineral (compounded for a 4 turn ETA to build). Something like 1.8 ECs/mineral when actually used numbers directly from the game and only checked the effects for 2 turns.

You calculate the stockpile by taking the percentage of the production that you rushed and then multiplying it against the amount of stockpile that you got, then you add in the number of turns you sped up the crawler multipled by the number of minerals that you crawled.

To take into account the time factor you take the overall percentage increase for stockpile and minerals for the past 2 years and do compond interest on the industry growth generated by your rush in proportion to your total factional production. You compound the average mineral/stockpile growth of your faction against the portion of achieved growth due to your rushes. Once you gotten the number of minerals that you have effectively achieved by the time of your SP build you can then find the mineral to EC rate.

It takes a while to calculate and I was originally working off a real turn to generate the inital figures. I wouldn't know how to figure out the cost figures for -1 or -2 industry without actually planning out a couple of turns for a real Cult or Spartan game. I have a feeling crawler upgrading would be far cheaper in that situation.

I think for the hive infantry crawlers only ever go down to 2 EC/mineral with clean trance plasma armor, until fusion reactors. Rover crawlers or needlejet crawlers are cheaper for the actual upgrade, but the problem is they're offset by requiring more minerals and hence have a slower build time, so they gather less resources, and reduce the amount of stockpile. Consequently rover crawlers for upgrading is about the same as infantry crawlers assuming stockpile allowed and available forests/mines to put crawlers on.

Once you get can make an trance/plasma armor/clean crawler (about 2 EC/mineral) and an average player is running the turn. With +3 industry the cost of rushing would probably be 2 EC/mineral (assuming about 4 turns to estimated build time). So basically comparable with crawler upgrading because when you have researched up to clean I think SPs start only taking 3-4 turns at most to build.

Hrmm so long winded I am...

[johndmuller]

Kody, if I'm understanding you correctly, it seems like you are saying that the stockpile energy bug tax rebate program pays off on the basis of (normal production of base) plus (extra mins rushed/purchased) rather than merely on the basis of the (normal production of the base) alone, which had been my presumption (not based on any personal testing or observations, but instead, I think, on postings in other threads to that effect).

To sketch out an example, say we are producing 20 mins per turn at the base and we are building a unit that costs 40 mins, and that stockpile energy is in the queue.
A): If we start with nothing accumulated and do nothing special, we will finish the build on the second turn with nothing to spare and no carryforward to the next build.
B): If we start with nothing again, but this time rush an extra 10 mins (for 30ec) on the second turn, we again finish the build, but this time have a 10 min carryforward for the next build.

My current understanding of the rebate plan is that we would get bonus ec's to the tune of 10 extra ec's following the second turn in both scenarios, but I am getting the impression that you are saying that we would get 15 ec's under scenario B, where we anted up the rush buying credits - Is that the way it is? or are you talking about something else or some other situations?

[Kody]

When you rush your productions your productions they finish more often in some cases. For example you're producing 20 mins per turn and your at 0 industry so crawlers cost 30 minerals. If you adjust your production so that you have 10 minerals queued and every turn rush it an extra 10 minerals then you'll finish a crawler every turn opposed to every 1.5 turns.

The increased stockpile energy is then one extra stockpile every 3 turns. So that's 10 ECs every turn instead of 6.67 ECs every turn. You need 30 ECs to do the rush every turn, but you get 3.33 ECs back every turn due to the increased stockpile energy. So that's actually 26.66 ECs for each rush.

Over 6 turns the rush causes crawlers (either 1st or 2nd or 3rd) and 4th,5th,6th to build 1 turn earlier you get an extra 2 minerals each crawler (4 crawlers build 1 turn earlier). The 2nd rush causes either 4th or 5th or 6th crawler to build 1 turn earlier so 1 crawler builds 1 turn earlier. Assuming you have forests available to put the crawlers on in the mean time, you get 5 extra crawler turns at 2 minerals each so 10 minerals extra.

That's an extra of 1.67 minerals per turn (10 minerals/6 turns) in additional to the amount your rushing. Add that to the 10 minerals your rushs to get 11.67 minerals.

So the rate is 26.66 ECs divided by 11.67 minerals so the mineral rate is 2.2857 ECs/mineral and we haven't taken into account the compound interest yet.

So at first glance the EC/mineral rate is 3 ECs/mineral, but due to the way the game works you actually get a rate of 2.2857 ECs/mineral over 6 turns. This is at 0 industry the effects work better when you have a better industry rating as you tend to rush at 2.3 ECs/minerals flat rate before all the extra stockpile and minerals are taken into account.

[girlfight_club]

Kody, you should to be doing something else on V-day!

But excellent stuff!!! You should pdf it and post it somewhere, or someone should get into some datalinks strategy mod!

Chiron must be infomercialled!!

[MariOne]

Re: Rush Buying - Maximise your ECs today!

Originally posted by Kody
Well Contruction Expert Nethog has compiled some useful data on the ins and outs of the construction industry.
Excel Rush Build Costs - Nethog's Pricing Analysis

Well, it's nice to see that the interest in SMAC thrives on.

It's not easy for everybody to search for old topics and find the right threads covering the subject they looked for, so it's good that new players present to new forum readers the old same issues again.
Hey, after all, there's plenty of new actors every year testing themselves with ole good Shakespeare, despite it has already been so many times so finely represented in the past 400 years!

So, I've got no problem with fresh blood piggybacking and rehashing the work that has been done by old vets in the past 4-5 years in this forum.

More than once I've already linked to old threads about units hurrying costs, and rushbuilding in general, when yet a new discussion popped up on the same subject (sorry if I can't be bothered right now to do it once more...)

It would just be NICER if at least credits to previous work were acknowledged.
It's not matter of copyright, but "NetHog's" Excel table is actually MY table.
He had built one table, but then I showed him the one I had done years before, and he adopted mine completely.
The layout, conceptual organization and actual formulas are all those of my original table.
He added the examples tab, where I only offered the chart and made any observation and explanation in the threads. And also there indeed, his #3 and #4 examples were added after I explained him in a thread the related rushing tactics (which were also common knowledge amongst the veteran players community).


...this all, just for the sake of precision...

[Kody]

Since we're going to start giving credits and all.

* I learnt from one of Mongoose's post about the changing to facility trick for 4 ECs per minerals when under 10 minerals on the civgaming strategy boards.

* The MariOne build table I guess I should call it, was where I figured out on how things are rushed. I studied it for a while and got and handle on how things should be rushed after I adapted it a little (no I'm not going to release my adjustments as they're shoddy).

* Reading Ogie Oglethorpe posts taught me the basic concept of turn advantage and how it accumulates.

* Observing Buster's every move through infiltration in the ACDG made me question the idea of partial rushing everything and lead me to realise the usefulness of the roll over minerals.

* Vev taught me the idea of crawler -> prototype rushing.

* Mongoose's posts taught me about unit upgrading. I think I followed it up by reading posts by a couple of other people unfortunately I forget who.... might be Curiousity.

* Minute Mirage and I had a long discussion via PM a while back on crawler rushing. We both had come up with the idea seperately and we hammered out some of the variables with industry and such. I originally felt only the factions with industry bonuses would benefit, however Minute Mirage felt otherwise. So I did the calculations and it turned out he was correct.


The anything else mentioned are things I've worked out by myself from fiddling and putting ideas to together.

[Snowflake]

In fact why couldn't we start a library like that in the CGN? There are so many good works here buried somewhere. If somebody could lead the project, then people can submit completed work, or nominate old threads to be included as links. There are quite some good skills and advices that are really worth of learning about for anybody who is interested in SMAC.

I wonder which mod should we contact for this request?

[Mr. Harley]

You also need to understand the ins and outs of designing upgrade crawlers. A crawler with special abilities and less armor always costs fewer energy to upgrade to compared to a crawler that has the same mineral prodcution cost but better armor (fission please).

To upgrade a standard infranty crawler to a hypnotic trance synthametal armor crawler 0-2t-1 cost 90EC, but to upgrade it to a pulse armor crawler 0-3p-1 costs 120EC. They both get you 80 minerals when you turn them in, so the cheaper one costs you 90EC for 50 extra minerals. If you don't permit the stockpiling energy after unit production - most PBEM games don't - it's a pretty good purchase for any faction, unless you are inefficiency crippled. The break points vary with the technology - reactor, but your always better off with special abilities versus extra armor. Just a heads up.

[Minute Mirage]

Originally posted by HongHu
In fact why couldn't we start a library like that in the CGN? There are so many good works here buried somewhere. If somebody could lead the project, then people can submit completed work, or nominate old threads to be included as links. There are quite some good skills and advices that are really worth of learning about for anybody who is interested in SMAC.

I wonder which mod should we contact for this request?

I actually went through the past threads and collected those that I felt were the most interesting. I'd still need to sort them out sometime.

[MariOne]

Kody, in case my words and attitude were ambiguous, I actually praised the effort of collecting and representing old and probably otherwise forgotten knowledge, and didn't deny your positive contribution in reorganising and dramatising it, adding something new from your flour bag too.

I was just a bit because I *thought* to remember that NetHog had added references to my original work in his file, and I realised that wasn't the case....
Other from that, I didn't want to take any credit,and I concur that names such those of Mongoose, Ogie, Buster should be revered here.
___

That said, you might want to edit your initial post, where you wrote : "So if you want to rush buy something you should always try and spend the money on productions that have more than 10 minerals queued. Facilities cost 2 ECs each mineral above 10 minerals, or unit productions that are nearing completion and are over 10 minerals each. "
you should substitute "at least 10".

Also, as someone else noticed in the original Unit Hurry thread, exponential is ok for common sense language, but mathematically speaking the cost increment is not exponential (constant^x), you could say it's "of the 2nd order" (x^2)....
But here I'm just nitpicking.

I would not let instead pass the expression: "Looking at the economic characteristics of rush buying it can be seen that the last couple of minerals for units are always the cheapest".
This is THE misleading statement that started my whole interest on the issue!
AT A GIVEN MOMENT, all minerals cost the same.

When you miss 20 minerals to completion, all of those 30 cost the same 3 ec. If you pay for 15, you pay 15*3=45. If you pay for 18, those 3 more cost the same 3 ec, and you pay 18*3 =54. If you wanna complete and also pay for the last 2, they still cost 3ec and you pay 60.
You need to wait some turn and be left with LESS MISSING minerals to completion, to see the price for all of them going down....
Such misundertanding actually had led some expert teammate of mine to make the wrong rushing decisions....

What you said is true ONLY in the sense that when you miss just 4 or less minerals to complete the unit, each of those costs 2 ec, just like a facility. That can be kept in Mind when rushing for instance synth garrisons with a Planned-Wealth Hive (should you ever want to do such a thing).
I SEE that's what you really meant, reading the following sentences. But the intial phrasing could stick in some new player's mind and mislead him, I guess it requires rephrasing, to avoid any possible misunderstanding.
___
Regarding the Stockpiling bug, my extensive past tests lead me to support jdm's interpretation. But I should review them to make sure, now.
___
Re: crawlers cashing into several simultaneous protoypes of the same design (combined with a skunk in 1st base), I just add one word to help you understand the *extent* of its effectiveness: PBs
Combine it with crawler upgrading, and you have a whole nuclear arsenal *overnight*. So, whatch out when your opponent becomes able to build missiles.... He might do it faster than you expect.

___

I once even built a table to decide exactly WHEN it was convenient to rush a facility and then switch to a unit.
But it was too complex for the average player, and very seldom applicable, so I figure I dropped it....

Remember, that excess accumulated are wasted (like NetHog reported in the #4 example I told him to add to MY table ).... WHEN production is completed. That is, next turn....
If you SWITCH right after having paid to add those excess minerals, they DO count for the halving...
This is the only case when it might be worth to overpay for a facility (because it NOT that facility that you actaully want to produce)....
Imagine, you HAVE to produce a 6 rows unit under normal industry and you have right 10 minerals acumulated (or few more so that you just lose 1 or 2, or even better you WERE already producing a facility before you suddenly realise you NEED a unit THERE and NOW). The ~50 missing minerals cost you 4.5 each. Imagin you produce 10 so you have to pay for 40*4.5=180ec.
Now, switch to a facility, even a cheap one. You have 10, you need to end with 50, you'll produce the last 10.
To have 50 when you switch back, you need to get to 90 (you'll thus lose 40 in the switch). You must then pay for 90needed-10there=80, *2ec=160. You just spared 20ec.
That's what jdm already explained you.
The catch is in such case it's irrelevant whether you pay to accumulate 90 minerals on a facility costing 160, or just 40. Even if you pick a facility costing 40, you don't have to worry for those 5 rows of red bricks you just bought, as they'll turn white again after the switch.

Not that revolutionary you'd say.
Well, think to this.
I *stressed* that given the missing amount, all the units minerals cost the same. Well, that's true also for the *extra minerals.....
Imagine you were producing a recon rover. Two rows, only few minerals still missing.
You spot an incoming army. IF you have that rover ready TO USE next turn, you might preemptively attack the invaders. But a laser is not enough, you need, say, at least Chaos weapons. BUT if you upgrade the rover after production, you can't use it for that turn, and the turn after it will be too late. BUT, if you switch to a Chaos Rover before rushing it, the missing Minerals become too much and the cost unaffordable.
Imagine too that you agree to not DW upgrade at the beginning of a turn, or that anyway you also have other ReconRovers in service and you will not have the money to DW upgrade them all... And finally of course that you can't bring in military help from any other base.
Solution?
If you have 4 or less mineral missing, you can directly rush the Recon Rover, overpaying for it as described above. No need to pre-switch to a facility, as the minerals you pay for the rover cost the same. Yes, non just the missing 4, but even the 80 excess ones you're goint to rush before the final switch.
This is becoming a bit complex. It should not be forgotten when you have already some minerals accumulated on an initial unit (say, an Impact this time) that would get lost on a pre-switch to the intermeditate facility.... In such case, rushing the initial unit might still be worth even up to 10 minerals missing.....
This is mostly just theory anyway, I almost never found myself in emergency conditions requiring such tricks. Almost.

____
If you want a breakdown of "cost per added row" in upgrading crawlers under the most usual conditions, that is, "crawler upgrade convenience" tables, I had built several.
Once you reach Fusion reactors, cost can get as low as 12ec per added ROW. Yes, that independent from Industry, you'll have then to figure out what that means per mineral under your actual industry setting. Even at +4 (! Drones Only!) that is still 12/6=2ec per added mineral, and crawler rushing can't beat that one.
Before Fusion, but with TWO abilities (picked between Drop\Clean\Trance\ECM), you can still go below 15ec per row (160/11) which still rivals a +3Industry crawler late rushing.
Only while you are limited to Fission and one crawler ability (provided you have Synth and Plasma Armor), best effectiveness tops at 18ec per added row (90/5), and even that way, a +2Industry would allow you to only rush a crawler after you have 18/24 minerals accumulated in it, to retain a marginal convenience over upgrading....

Not to mention that Upgrading can be done on the fly, when needed (indeed they call it "instabuild"), while investing the same money in crawler rushing must be planned and carried out well in advance and over a longer period.
This doesn't mean that I don't rush crawlers too, but only because of turn advantage in collecting resources, not because its "alleged" cost-effectiveness over crawler upgrading.


=======

Anyway, big fun your examples, I specially LOL @ the skunks one!

[MariOne]

Originally posted by Kody
I think for the hive infantry crawlers only ever go down to 2 EC/mineral with clean trance plasma armor, until fusion reactors.

Again OFF THE TOP, as everything I wrote above.

basic "supply crawler" (SC): 3 rows
Trance SC: 5 rows
Synth SC: 5 rows
SynthTrance SC: 8 rows (?)
PlasmaTrance SC: 9 rows

Upgrade costs: (NEWrows+armorINCREASE)*10
Trance: 50
Synth: 60 =(5+(2-1))*10
SynthTrance: 90
PlasmaTrance: 110

Cost / ADDED Row (independent from Industry)
Trance: 50/2 = 25
Synth: 60/2 = 60
SynthTrance: 90/5 = 18
PlasmaTrance: 110/6 = 18.33

So best effectiveness you can get before Fusion and before one of BioEng(Clean) \ AST (ECM) \ MMI (Drop), is 18ec per added row.

Hive starts at +1 Industry, which is 9 mins/row.
This is indeed 2ec/min you can buy by upgrading.

But to have crawlers you HAVE to discover IndAuto and the prereq PlaNets, and Planned is even efficiency harmless for the Hive, so it's a MUST.

At +2Industry, 18/8=2.25 ec/min
Let's assume you wont' disdain to also adopt Wealth:
+3Industry, 18/7=2.57 ec/min

____

ONCE you discover tho one of the above Special Abilities (SA), things change.
They also change a bit less with less expensive SAs, which now don't pop to my mind tho.

Synth, 2SA: 11 rows (8 added), upgrade 120ec, 15ec/row (added)
Plasma, 2SA: 14 rows, upg. 160ec, 14.55 ec/row

+1Industry: Synth 1.62 ec/min, Plasma 1.67
+2Industry: Synth 1.82 ec/min, Plasma 1.88
+3Industry: Synth 2.08 ec/min, Plasma 2.14

I provided the figures...
I leave you to complete the work drawing the conclusions within the framework you set up, thanks

We could say tho at first glance that crawler rushing competes with upgrading only if you take well into account all the factors that you PERFECTLY (kudos) described in your crawlers post, and their compound effect over time.

The time deadlines you gave have little meaning tho, because how long you need to spare depends on how much you make every year....

[Kody]

Kody, in case my words and attitude were ambiguous, I actually praised the effort of collecting and representing old and probably otherwise forgotten knowledge, and didn't deny your positive contribution in reorganising and dramatising it, adding something new from your flour bag too.

I was just a bit because I *thought* to remember that NetHog had added references to my original work in his file, and I realised that wasn't the case....
Other from that, I didn't want to take any credit,and I concur that names such those of Mongoose, Ogie, Buster should be revered here.

There's lots of people that we've gotten information from over the years. After your little comment I thought it might be good to mention them. Some of them tend to avoid flaunting themselves so most people will never realise what they have contributed.

_________________________
Updated the first suggested changes you made.

What you said is true ONLY in the sense that when you miss just 4 or less minerals to complete the unit, each of those costs 2 ec, just like a facility. That can be kept in Mind when rushing for instance synth garrisons with a Planned-Wealth Hive (should you ever want to do such a thing).
I SEE that's what you really meant, reading the following sentences. But the intial phrasing could stick in some new player's mind and mislead him, I guess it requires rephrasing, to avoid any possible misunderstanding.

Look at the 2nd table that shows the per mineral costs. You'll realise that I meant exactly what you first thought I meant.

http://apolyton.net/upload/view.php?file=33915_CrawlerRushbuying.zip

The EC per mineral cost lowers the closer you rush it to completion.

Regarding the Stockpiling bug, my extensive past tests lead me to support jdm's interpretation. But I should review them to make sure, now.

It's like 0.5 cents in your bank account, it's there but you can't actually withdraw it until you get another 0.5 cents to make it 1 cent. Basically what I'm talking about is a 0.5 stockpile, you can't actually get the extra ec's from the stockpile until you add another 0.5 stockpile. The example I gave after JDM's post did multiple rushes and stockpiles and showed the additional stockpile energy generated when you continue to rush. I hope that explains it better.

the *extent* of its effectiveness: PBs

I don't particularly like the concept of PBs that why I negelected to mention them.

Remember, that excess accumulated are wasted (like NetHog reported in the #4 example I told him to add to MY table ).... WHEN production is completed. That is, next turn....
If you SWITCH right after having paid to add those excess minerals, they DO count for the halving...
This is the only case when it might be worth to overpay for a facility (because it NOT that facility that you actaully want to produce)....

There are gaps in what I wrote, problem is gets rather time consuming chasing down all the tricks and possibilities and outline the specific circumstances. *adds it to the list of things he hasn't mentioned*

Even at +4 (! Drones Only!) that is still 12/6=2ec per added mineral, and crawler rushing can't beat that one.
.....
This doesn't mean that I don't rush crawlers too, but only because of turn advantage in collecting resources, not because its "alleged" cost-effectiveness over crawler upgrading.

Crawler rushing gives you hidden cost rebates, partly because of the turn advantage and partly because of stockpile. So with +4 industry you'll probably be able to get something a fair bit cheaper than 2 EC/mineral.

What I've been trying to do is add those rebates back into the inital cost, that's something which people have neglected to do conciously. From what I can see the people that do rush, just feel that rushing is better in certain circumstances without being able to exactly say why or when.

So I tried my best to outline the reasons why and tried to give a rough estimate of when.

Anyway, big fun your examples, I specially LOL @ the skunks one!

Thanks.

I provided the figures...
I leave you to complete the work drawing the conclusions within the framework you set up, thanks

*Shudder* no way! It took me a long time to figure out all the variables for a specific circumstance.

There's too many factors to say definititively that under these set of circustances you're better off rushing or upgrading. If you really need want a 100% certain answer than you need to look at an actual turn and point out each production queue to say rush here, or upgrade this crawler.

+3Industry: Synth 2.08 ec/min

So about 2ec/min, I wasn't sure of the exact number as I had the game uninstalled to stop me from playing it.

The time deadlines you gave have little meaning tho, because how long you need to spare depends on how much you make every year....

They have little meaning, but at least give people a feel of about when from what I've experienced. So a near enough example at least gives a starting point to figure things out from.

It's sort of like explaining complicated maths to a group of investers. You can't give them the formulas and expect them to work everything out when it took you over a week to look at one circumstance properly. Instead you give rough estimates based on what you done.

[Adalbertus]

Sorry if it wasn't mentioned before but the posts got quite long in this thread:
When I tend to have sufficient money to regularly rush builds, at first I focus on rushing those who have two turns to go. Why?
Assume your base produces 22 min/turn, and you're about to finish the build next turn. Say seven mins are missing. Then you'll waste 22-10-7 = 5 mins (you'll need seven, you can keep ten more and five are lost).

[jtsisyoda]

MariOne wrote:
I would not let instead pass the expression: "Looking at the economic characteristics of rush buying it can be seen that the last couple of minerals for units are always the cheapest".
This is THE misleading statement that started my whole interest on the issue!
AT A GIVEN MOMENT, all minerals cost the same.

I thought "the last couple of minerals for units are always the cheapest" referred to not losing out on the rounded up fractions. If it's 2.5 ec/min to rush 2 mins, it's 3 ec's to rush 1 min. I remember situations where it would be 2.5 ec/min to barely finish, but 2.25 or so to completely finish. Or am I missing the point?

[Mr. Harley]

I used to try to figure the rush buying as some sort of sequence, i.e. 3 min 2 min 2 min 2 min 2 min for the last five minerals. However, all you need to do while in a game is use your windows key, bring up the calculator, and use the formula ( minerals being purchased/total minerals left * energy cost to completion ), always rounding up. I used to do calculations like that in my head, but age and now that I'm playing more PBEM means that I use the calculator.