Updated Rush Building Chart

Nethog · February 20, 2003, 17:20

I updated my SMAC(X) Rush Building Chart with detailed instructions on how to use it! (BTW, I previously called the chart the "Mineral Cost Chart" which wasn't a very accurate description of what it was used for).

The chart helps you calculate the exact amount of energy to spend for a partial rush build so that the unit is completed on the next turn. If you use the amount that the game suggests you can waste a lot of energy.

BTW, when I was first researched the area of rush building I remember stumbling across a thread that described how you may want to overspend on rush build costs since some minerals are carried over to the next unit in the queue, and how you can save energy when rush building the next unit. I haven't incorporated this into my gameplay, but I probably should. If anyone stumbles across threads discussing this please post the URL here.

Oh, and you can get the SMAC(X) Rush Building Chart at:
http://mywebpages.comcast.net/pmm1/games/games.html

Ned · February 20, 2003, 22:12

I long ago adopted the overspend technique so that I end up with 10 extra minerals for the next rush build. As well, you get the benefit of the stockpile energy for the 10 minerals.

But the overspend seems critical on SPs. If you spend the exact amount necessary, often your SP will not be build the next turn. There is apparently a rounding error somewhere in building SPs.

(There is a similar phenomenon with respect to tech research. Often you will find that you do not get the next tech even though the number remaining plus the research conducted in the present turn should have been enough. Thus you will see 1 turn to go two turns in a row.)

Alinestra Covelia · February 21, 2003, 01:08

I get that phenomenon with the tech quite often, and I think it's usually because of some fluctuation in energy per base. Any number of things could cause this - most usually, a Drone Riot, but sometimes even something as mundane as a Fungal Pop or even an intrusive enemy rover squatting on a nice high-energy square.

Don't know about the SP though. Each time I've calculated the exact mineral costs, I have been fine. Mind you, I was always playing a high Support faction like the Hive, so "sudden support" situations (such as moving Mindworms out of fungal squares, etc) didn't arise.

MariOne · February 21, 2003, 05:56

Nice materail you gathered there, Nethog.
Although, many of the issues you touch have already been covered in depth in the past.
Forgive me if I quote myself, but here are a couple of links, mainly on the overpay issue.
Overpaying problems
Hurrying Production Mineral Loss?
Simply put, overspending, in the sense that you pay even more tha a *complete* hurry, does not work: you can only carry over *current* production excess, up to 10, but *red bricks* are ALL lost when the turn ends.

Then, the table.
Other Rush-Build Tactics
Above it's the link to the thread where I last attached my table about units hurrying costs.
You may also follow back the links to previous threads on the subject.
If you accept a positive, constructive opinion about our two tables, I'd explain why I think mine could be more useful for the average user:
- my table does a further layer of work that you instead leave to the user. Namely, your 2, 4 ,6 steps. You have a spreadsheet, let IT do the calculations! It's still up to the player to devise the number of total missing minerals, and the strictly required ones (you call them "remaining", I dubbed them those you need/want to pay for). But then find the correct row, and follow it till you meet the desired column. Easy as 1-2-3 (wait, that was Lotus, not Excel!

)). This allows for some flexibility too, easy to price at first sight. For instance: your base is producing 9 minerals, you might want to pay for *8* less, so that 1 is carried over, and the turn after you'll reach the 1+9=10 fatidic limit (and take into account the new unit's support too, if that's not free).
- on the other hand, you did too much unnecessary work.Frankly, WHEN do you design a 300 minerals unit??? And WHEN do you want to RUSH it????????? 5000ec for a unit! Sheesh!

Unless in *desperate need*, spending more than 4ec/mineral to rush a unit means you have to revise your approach to production/economy. You might find more convenient to produce/rush cheap shells and then upgrade them. OR, if you really need to rush such an expensive unit with so many still missing minerals, you might EVEN find convenient to switch to a facility, overpay (only case in which it could be worth) for it at 2ec/mineral, then switch back to the unit. Even with the retooling penalty, it could be still convenient if you don't have much more than 10 already accumulated, and your unit minerals would cost 5ec each to rush, or more.

BTW, regarding your note: the easiest way to find out the exact number of missing minerals for an expensive item, is to look in the F4 screen.

____

Regarding the SP hurrying "misfires" Ned, I second Alinestra on that.
And let's put things clear: if on your next turn you find LESS minerals than you projected/expected, that's *always* (in my experience) explainable by some game fact, which causes your production in your base to decrease in the meantime before it got actually applied for the next turn update.
Same for labs, which is even more difficult to track down as it's a factionwide effect.

The real problem with "tech misfires" is when looking in F2 you see that you have already accumulated MORE labs than actually needed to discover it, and you read 0 turns in the top screen bar...
This is harder to explain, but there *are* reasons for it too, I just dont want to make this post any longer...

zsozso · February 21, 2003, 11:39

Somebody has reported on these forums long time ago the exact formula for rush building units - your fractional numbers are probably accurate and come down to the same thing. Anyway the formula goes like this:

1. You pay 2 energy per mineral for the LAST 5 necessary for completion (and any carry-over)
2. You pay 3 energy per mineral for the minerals 6-15 backwards from completion.
3. You pay 4 energy per mineral for the minerals 16-25 backwards from completion.
and so on...

Of course, if you do not have 10 minerals yet, then the cost doubles.
You should over-spend the 10 mins, if you know you will rush again next turn. HOWEVER, if you want to save money (I mean energy) you should also calculate the production of your base, i.e. do not pay for minerals that you are going to collect anyway. OK, now it is getting complicated, so lets take an example:

Your base produces 14 minerals, and it needs 38 to complete the current unit and you plan to rush again the next turn.
10 of the 14 will go for over-production, that leaves 4 for the current unit, unfortunately that goes out of the "cheapest" 5 minerals.
What you need to pay for the other 34mins is: 1*2 + 10*3 + 10*4 + 10*5 + 3*6 = 140

As for the research point accumulation and not getting your tech next turn, the reason is not energy fluctuation. There are 2 problems:
1. The F2 screen reports the tech accumulated incorrectly (this is obvious from the fractional numbers it reprots sometimes, when your labs points actually collected are always integer). You can count the exact number of lab points you earn if you sum up the lab points of each base as reported by the base-screen - I know it is very tedious, but you can see on small number of base cases that it can differ from the F2 claim. You will see the sum is the exact number added to the "accumulated" by next turn IF YOU DO NOT MAKE ANY BREAKTHROUGH.
2. The second problem is responsible for the bigger loss usually. The program will accumulate the lab points base by base, and when it gets to a point that you would make the breakthrough at base X then all the excess lab points of that base are lost. So if you have a science city producing 400 lab points, tech cost is 560 and you accumulated already 500, you would expect that after the breakthrough you will have 340 for the next tech. But in fact, you lose that 340 completely - goes down the drain. This is important if you are in the stage to make more than 1 tech a turn.

Another reason you could end up not completing a tech even though you accumulate enough lab points is drone riots or starvation. Those things do not matter if the base in questions just "contributes" its tech points. However, if it happens, that the particular base is the last one, and its lab points are necessary to complete the tech, then you wil NOT MAKE the breakthrough due to the riot or starvation. That's the meaning of "nonessential production is abandoned" when the game warns you.
In this case, the next turn you see on the F2 screen that you have more points accumulated than the tech cost, breakthrough expected in 0 turns. So you do not lose the points, but you simply don't get the tech, you will get it 1 turn later (unless you are rioting again in all your bases).

MariOne · February 22, 2003, 08:30

Quote:

Originally posted by zsozso
Somebody has reported on these forums long time ago the exact formula for rush building units - your fractional numbers are probably accurate and come down to the same thing. Anyway the formula goes like this:

1. You pay 2 energy per mineral for the LAST 5 necessary for completion (and any carry-over)
2. You pay 3 energy per mineral for the minerals 6-15 backwards from completion.
3. You pay 4 energy per mineral for the minerals 16-25 backwards from completion.
and so on...

Sorry to remark it zsozso, but on this subject what you posted is old, wrong, and already superseded by the info I had just posted.
Just follow back the links inside the 3rd thread I linked to above.
There you will find the *exact* figures and practical methods, no guessing needed, and no rules of thumb letting you one mineral short your vital defender b/c of 1ec underspent due to miscalculations.

In particular, to avoid that some unexperienced leader might be misled by your post...

Quote:

Originally posted by zsozso
- your fractional numbers are probably accurate and come down to the same thing.
...
Your base produces 14 minerals, and it needs 38 to complete the current unit and you plan to rush again the next turn.
10 of the 14 will go for over-production, that leaves 4 for the current unit, unfortunately that goes out of the "cheapest" 5 minerals.
What you need to pay for the other 34mins is: 1*2 + 10*3 + 10*4 + 10*5 + 3*6 = 140

...this is UTTERLY wrong!
what you show it's the progressive increase for a *complete* rush cost (and it' not even correct as the 5th and 6th mineral of each teen do swap their costs...).
But once you determine the # of missing minerals and their total cost, any partial payment is merely proprtional, rounded up to the integer. That is, the 1st mineral out of 38 costs exactly like the 38th out of 38, befre the roundings. In this at least Nethog was correct and very clear too.

This does NOT boil down to the same things.
If you donwnlad my table and look it up, go to the 38 missing minerals row, then follow it to the 34 minerals to pay column.
Cost for a partial hurry of 34 minerals out of 38 missing: 133ec.
No decimals (my table already took care of the *correct* roundings), no calculations to do.
Spending 140ec on 38 missing minerals, you'd have obtained 35minerals (costing 137). You'd have directly wasted 3ec in wrong roundings, and then one extra mineral due to carryover limit. If those 7 wasted ec were necessary and vital somewhere else in the same turn, you'd have harmed yourself because of approximation and superficiality. When no particular effort was required and the correct figures were ready and *easy* to lookup...

Heh, you're the master of fast transcendence, and this speaks for a mastery understanding of game strategies and dynamics and precision in micromanagement to.
Go figure, if you had *also* understood units partial hurrying costs, you might have shaved a further one or two turns from your records

(I shiver thinking how many ec you might have cumulatively *wasted* in incorrect partial hurrying of formers and crawlers in your record-setting games!!

)

zsozso · February 22, 2003, 09:09

OMG!

MariOne, thank you for correcting my mis-information, and I'm sorry for everyone for posting that. I was trying to be helpful... Somehow I misses those threads about the correct hurry form and indeed, I was still using the wrong one wasting some ec.

MariOne · February 22, 2003, 09:44

zso, you are amongst the most contributing posters here, and not just by writing but also by "living example" so to speak (i.e. your many saves attached).

I found out that due to Poly forum merging and archiving policies, the link chain is broken after the first step back.
But I luckily found the link to the original thread.
What's the formula for HURRY cost?

BEWARE!
Someone recently called me "pompous arrogant a s s hole".
Judging from my posts then, he was right!
Also, I'm proverbial for my verbosity and longwindedness.
Well, those old posts of mine were *truly endless*, and said from me....
They at least serve to prove to newcomers that I've come a LOOOONG way to amend myself since then, although I still have a lot to improve.
At least, what I posted there is technically right.
Then I let me carry away by that Adam_Smith...
Only later, resident Aplytoners explained me why it was absolutely useless and to be avoided to argue with him.
Him looking like a math nitwit didn't help either.
I must admit that he never lost his temper and aplomb, his increasingly appalling replies kept almost always extremely polite... I would have reacted much worser to my arrogant posts

Anyway, I hope you enjoy the reading, some posters found it entertaining, I still now had some good laughs at myself in rereading it.

Nethog · February 25, 2003, 15:31

I have updated my SMAC(X) rush building chart AGAIN after reading all of the information you folks have posted here. You can get it at http://mywebpages.comcast.net/pmm1/games/games.html

MariOne, thanks for all of the useful info! In fact I TOTALLY agree with all of your constructive criticism of my original chart, and as a result, have decided to replicate your chart (with of course acknowledging your contribution), as well as add enhanced formatting, notes, and examples. My goal for doing this is to help new SMAC players, as well as help myself remember the details after the next long abscence and return to playing SMAC again

I actually learned quite a bit more about rush building through this excercise, and finally understand the difference between overbuilding and carry-over production which will definitely help my SMAC gaming skills. Thanks again.

If anyone gets a chance, please review the examples I have included for rush building and report any errors, ehancements, or additional suggested examples to include. I have issued the chart as version 0.90 and will release it as version 1.00 in a couple of weeks after gathering all feedback.

As for some of the historical posts regarding the rush building topic, the thread What's the formula for HURRY cost? was indeed FUN reading!

An for my ending comment:

MariOne wrote:

Quote:

...this is UTTERLY wrong!
what you show it's the progressive increase for a *complete* rush cost ...[SNIP]...That is, the 1st mineral out of 38 costs exactly like the 38th out of 38, befre the roundings. In this at least Nethog was correct and very clear too.

Hey, at least I got something right!

johndmuller · February 26, 2003, 03:58

I suppose that the formula for hurry cost is in one of those links, but it seems like with all those teasers, someone ought to put the formula out in plain sight somewhere in this thread; so here it is:

Cost = (M*M)/20 + 2*M
(where M is the number of mins you need to complete the build - doubled if you don't already have the first 10 mins). The game rounds this down to the nearest integer if you buy the full amount; if you buy a partial amount you have to round up the fractional ec's.

If you are looking for things you can conceiveably remember or maybe do in your head, there is a simpler approach which waste a little, but not too much and is easy to remember:

If you need 10 mins (1 standard row), it costs 25 or 2.5 each;
if you need 20 mins (2 standard rows), it costs 60 or 3.0 each;
if you need 30 mins (3 standard rows), it costs 105 or 3.5 each;
if you need 40 mins (4 standard rows), it costs 160 or 4.0 each;
in other words, for each extra row, the unit cost goes up by 0.5 ecs.
If you just pay that amount, you will always have enough (unless you don't have the 1st ten yet) and will only waste a few ecs.

Nethog · February 28, 2003, 13:13

johndmuller, thanks for posting the formula here - it is also included in the spreadsheet.

As for the "simpler" method you post for computing rush costs - I think using the chart is still quicker and easier, and no minerals are wasted.

MariOne · February 28, 2003, 14:00

Sorry for not getting back earlier.
Nice formatting!
Clear examples and explanations! I love #3 particuarly, I would have never been that effective in exposition (in english).

In example #4 tho, the payment ist still "partial", not "total".

Nethog · February 28, 2003, 15:50

MariOne Wrote:

Quote:

In example #4 tho, the payment ist still "partial", not "total".

Nice catch MariOne. Thanks. I have already updated my working copy of the chart and will upload the corrected version of the chart to my website in a couple of days.

Nethog · March 3, 2003, 16:35

I have posted version 1.00 of my SMAC(X) rush buildng chart. There are only a couple of minor formatting changes from the 0.90 version. You can get the 1.00 chart at http://mywebpages.comcast.net/pmm1/games/games.html