Questions for (nejen) Sikander on "Sikander Spacing"

[UnityScoutChopper]

("Nejen" is this nifty Czech word that can mean, in a single word, "so-and-so/such-and-such but not only him/it." It's pronounced "NEH - yen.")

I finally broke down and peeked at the Dark Side recently: I played a few games with Sikander Spacing and early-as-possible FM. The results were good but not great, and they left me with some questions.

For any newbies here, Sikander Spacing is base spacing two on the diagonal (which I am incapable of ASCII-drawing at the moment).

Questions:

1. This is a q for EVERY fixed-base-spacing player, actually. How in the h*ll do you field enough formers to clear all those base sites? In my experience you have to be REALLY lucky to have a majority of your early potential base spaces fall on unfettered territory, and I'm having trouble understanding the cost-effectiveness of investing in some cases up to 14 former-turns (fung rocky) just to finally be able to use a !@$! base site. OK, 14 is the extreme, but I've seen lots of cases where after the first few base sites, the only alternatives with fixed spacing are investing 6+ former-turns or sending a pod far away (and in the early game, building an RC is often not an option). Where's the turn-advantage in that?

2. Do you build a road network covering "every white space of the chessboard," or leave large areas unroaded forever/a long time? I am a road fanatic, but I see people proposing roadlessness for defense/time savings, which I've never understood much, but...

3. Assuming your beelines are IA followed by EE, about what point do you expect to reach EE using this strategy? I'd like a yardstick so I get a feel for if I'm "doing it right."

4. What kind of variations do you use for the various factions? Are there some for which you simply forget about it and use ad hoc spacing (as I normally do)? E.g. Miriam, Domai, Yang with their late access to CE? Lal with his potential for size-18 bases? The -- shudder -- Pirates?

5. Do you reject tech trades until IA in order to reduce tech costs along your beeline?

6. IIRC, your eventual setup is boreholes on the bases' four diagonals and condenser farms everywhere else. DIRC? (Do I recall correctly?) And again, before clean formers, how in the heck do you whip up all the former-turns needed for a paradigm with no auto-spreading and such a low FOP-produced-per-former-turn ratio? (Yes, I do see its insanely high FOP-per-square ratio and do "get" that.)

7. I presume you use forests in place of the above-mentioned until the WP in the case of CF's and until... mineral restriction lifting for the case of the boreholes? How early do you start ripping up the forests?

Thanks in advance,

USC

Just for the record: my best game of this type was with Roze, where I was up to 20-30 bases and EE by some time in the 2170's, on a huge map. Not too great, but then I had zero useful contact with other factions (I think it was an alien, a bad-hair-day Lal, and no other contact yet). On the other hand, I lucked out somewhat in terms of specified base squares falling on immediately usable squares.

[jtsisyoda]

USC,

There's a great thread on CGN that covers terraforming in general, including some stuff by Sikander himself on this base layout.

http://www.civgaming.net/forums/show...=&threadid=907

In that thread, Curiosity and Archaic posted several saved games (amazing games in my eyes) using "Sikander spacing". If you look at those, you'll see they didn't condensor-farm every square that's not a borehole, but stuck with many 3-food forests. Another thing Sikander pointed out (and the posted games bear out) is using squares on the periphery. Each base in theory has 7 or 8 squares, but in practice, it ends up more like 9 or 10.

[Curiosity]

Follow jtsisyoda's advice and look over at Archaic's game; it's very good, especially in how he sets up his booming.

My game on that thread is meant to be an example of forest-and-borehole terraforming, using arbitrary spacing, rather than sikander's strategy. The Archaic/Sikander strategy definitely gives faster research, and pretty good productivity, too.

I've never been a fan of fixed base placement, but I'm experimenting with a variation on sikander spacing. If you take the tile the base 'ought' to be on, then the base can be moved one tile north-east, south-west, south-east or north-west without interfering with boreholes. Your packing fraction also remains the same, but you have a choice of 5 tiles for the base, at least one or two of which ought to be clear.
It doesn't look as pretty on the mini-map, though.

With tech trading, I trade only for what I need, at any given time. The techs don't go away just because you don't take them now.
That often means no trading until Env. Eco., then trades for Applied Phys, mobility, flexibility, and HEC - leaving only 2 techs to Air Power.

Not too sure when I'd normally get EE... sometime around 2145-2150 with Zak, I'd guess? I've no idea about the Angels, but 2170 doesn't sound too bad.

[Adalbertus]

I played Sikander spacing as soon as I read about it. It was the biggest single jump in achievements I had. One thing to realize (I think it was Sik who pointed it out) is that you shouldn't raise the squares around the borehole but to lower the borehole site to be able to build one.
Initially, I forest many squares, and only later switch to borehole or farm/condensor. To answer your questions for my playing style:

1. I usually send CPs "far away" if they don't find a place, or if I want to secure space. I usually assume that the further in the game, the less long ways are important. Two more turns to your second base place cost one turn for your faction. Two more turns for your eighth base cost 1/8 turn. And after the ninth base I build SPs.

2. I try to connect my bases with roads ASAP because it allows to shift defenders. It is also a defensive advantage for you on border bases to have a forest with road next to it (and no road leading away from it).

3. I'm bad in beelining IA because I want to get HGP and (if possible) CN ... It depends very much on map size and faction ...

4. The pirates are probably the only ones where I would forget about rigid spacing if I ever played them. Except if there would be an airborne custom faction with Colony Needlejets.

5. Yes, unless I get something on the beeline, or otherwise urgently needed.

6.Slowly. Forests first, and then replace them as there is free former time.

7. As soon as bases need to grow, I build a farm/condensor. I don't build many boreholes until mineral restriction lifting.

What you want to terraform, depends very much on your momentary needs. Boreholes are pretty useless before restriction lifting. Before IA, forests with a casual farm (with or without) condensor is best, because you can't afford to many non-specialists per base and you still want to have some mineral input (which you don't get from two worked farms/condensors and three specialists).
With some crawlers, you can get 2-3 formers per base which should be sufficient for your terraforming needs. I tend to delay building base improvements until I get energy, and lots of military until I get choppers (10 of them plus a few drop units conquer the world).

[Lazerus]

Why are you working farm+condensor squares ? they're all nutrients so you could just crawl them,work the forests and still have 3 specialists ??? this would give you the minerals you wanted

[Jamski]

Yes, that's what I do with this strat, crawl the condensors and work the forests/boreholes. Seems a bit silly not to. I might crawl a mined mineral rescorce as well, if I'm lucky enough to get one, but the worked squares should be those that give mixed incomes, and the specialised squares should be the ones you crawl. Or?

-Jam

[jtsisyoda]

A situation where it might be worth it to work a condensor-farm is to make sure you have half your population as workers. Then a little psych makes them talents for a Golden Age.

[Minute Mirage]

1. I too started using Sikander's base placement a while ago, but now I've switched to Blake's version, i.e. bases on "white" squares only, and out of the radius of other bases. I feel that this method is slighty more flexible in regards to base placement and leaves more squares per base, while still maintaining maximum borehole density. I will, however, not clear base sites that are rocky and/or fungus just to be able to stick with this method. Since only half of the "black" squres are reserved for boreholes, it is quite possible to place a base on a "black" square and still get the maximum amount of boreholes.

2. I build roads in every square I terraform, except in the very beginning when I'm getting those 2 nut squares.

3. In my current game (University, large map, average everything) I got Environmental Economics in 2138. I was very lucky, though, because I got Social Psych and Biogenetics from Deirdre, which allowed me to reach SotHB first.

5. Yes.

[UnityScoutChopper]

Written in Discordabetical order....

jtsisyoda wrote:

There's a great thread on CGN that covers terraforming in general, including some stuff by Sikander himself on this base layout.

Been there, read that. :-) I'm eager to see the saves, but I'm partly a bit lazy, and partly struggling with a falling WinNT thread related somehow to IE, called svchost. When it falls, I can only browse; I can't download, save, print, disconnect normally (!), or several other things. It only happens sometimes, but so far my well-intended visits to CGN ended with this thread crash. OK, I'll either fix it or try to manage a visit where it doesn't happen/hasn't yet happened.

Lazarus wrote:

Why are you working farm+condensor squares ?

Are you talking to me? I don't THINK I said I was sending workers there.

Adalbertus wrote:

Before IA, forests with a casual farm (with or without) condensor is best, because you can't afford to many non-specialists per base and you still want to have some mineral input (which you don't get from two worked farms/condensors and three specialists).

You already have 5-pop bases before IA??? I am still concentrating on horizontal expansion in that phase, protected from bureacracy by the HGP if possible. Am I making a mistake in this?

Minute Mirage wrote:

2. I build roads in every square I terraform, except in the very beginning when I'm getting those 2 nut squares.

I wish there'd been more answers on the roads question, but from the two I see, I at least have a nice feeling that indeed I would not do ill even if I made more roads, not less. Incidentally I consider the roads question to be one of the main advantages of pure Sikander spacing: NO OTHER SPACING -- none -- auto-builds a higher number of same-"color" roads for you. In a perfect (from the roading perspective) world, you only need roads on one color (preferably the one holding your boreholes, as the higher the former-turns needed, the more a 'forming project benefits from helpers and thus roads, and boreholes are the "biggest" project of all), and thus we can simplify and say that pure Sikander spacing gives the "densest network of free relevant roads." And that ain't too shabby.

Curiosity wrote:

If you take the tile the base 'ought' to be on, then the base can be moved one tile north-east, south-west, south-east or north-west without interfering with boreholes.

Sounds more attractive to me. (Unless my starting position and/or opening terraforming power (Deirdre) HAPPEN to allow true sikander spacing -- again, it's still a mystery to me how casually people take the opening-game need to have or terraform spots for it.) AFAICT You still have SOME restrictions -- e.g. in a NE-SW base pair, you can't send both the NE base W and the SW base N. But still pretty flexible. The free roads are not as ideally placed, but then if you want you can always switch over to full Sikander spacing as soon as your terraforming capacity catches up with your expansion... an acceptable compromise.

Minute Mirage wrote:

...Blake's version...

...has interested me for quite a while. I always had trouble figuring out from his descriptions if he went black-square with the boreholes to gain higher BH density, or white-square to avoid the roading dilemma. Without even trying, you thus cleared up a key question... thanks. So again... looks like it's perfectly normal around to go even wiiiilder than I do with roads.

Nuff for now.

USC

[Minute Mirage]

I know this is a somewhat old thread, but I felt like answering to it again, since my playing has changed somewhat in the past few months. I did try Blake's base placing, then went back to Sikander's system with more relaxed base placing limits, but now I place bases where approriate. I'd say no fixed base placement system is worth the extra former time required to clear the base sites. Instead, the bases should be placed according to the terrain. However, a base placement system can still work as a general framework, that is, how far apart are the bases, how many condensor farms are you planning to have etc.

I played recently a game on the same map as Sikander and Archaic (saves here), and I crammed my initial bases pretty close together, without no particular placement strategy. However, one problem of this was that I couldn't pop boom those bases up to max size (16) before I got satellites. Since those bases always had all the lab and economy enhancing facilities, I would have liked them to be as full as possible. This has given me an idea of not putting the bases so close in the very beginning, but leaving some space to allow a maximum pop boom. By the time satellites come along, food is plenty and the gaps could be filled with new bases. After all, by this time each base needs only two condensor/farm/soil enrichers to pop boom to size 15, so bases could in theory be placed in every fourth square. I find that the Hab Domes come too late to be worth it in any part of the game.

[Archaic]

A bit busy ATM, so I'll just reply to this small part before making a proper reply to the thread.

Originally posted by Curiosity
I've never been a fan of fixed base placement, but I'm experimenting with a variation on sikander spacing. If you take the tile the base 'ought' to be on, then the base can be moved one tile north-east, south-west, south-east or north-west without interfering with boreholes. Your packing fraction also remains the same, but you have a choice of 5 tiles for the base, at least one or two of which ought to be clear.
It doesn't look as pretty on the mini-map, though.

This is the preferred strat for MP, at least from my perspective. You simply don't have enough time to be slow with placing any bases, and it's worth more to mine the rocky and crawler it anyway. If it's only a fungus though...then usually I'd have enough time to eliminate it with only one former anyway, and place the base smack on.

[Minute Mirage]

Originally posted by Archaic
If it's only a fungus though...then usually I'd have enough time to eliminate it with only one former anyway, and place the base smack on.

That's still six turns that the former could spend otherwise. Sometimes that just might be worth it, but in my experience such situations are rare. In fact, I don't like to use my formers for clearing fungus in the early game unless there's no choice. Instead, I hope for the forest to clear some of it for me and only later do I remove the fungus with my formers.

[UnityScoutChopper]

After many frustrating games of "trying to be Sikander" in the weeks immediately after I started this thread, I'll have to agree that

no fixed base placement system is worth the extra former time required to clear the base sites.

If Sikander can make a fixed base placement -- no matter how sound the theory behind it is -- work, then a big hats off to him. I can't, because I can't field enough former-turns early enough to both a) use terraformers for their usual purpose -- raising tile value -- and b) clearing any square that falls into the pattern but is rocky/fungified. Because... b) is a LOOOOT of squares! I finally did a test run tonight. A really simple one. I took 15 standard-settings large-map starts (3 games' worth -- apparently the scenario editor doesn't "see" alien cities?!?!?), immediately turned on the scenario editor, and counted the number of non-water locations* that fall on Sikander spaces out to radius 4 (i.e. the first 12 "Sikander locations") from each starting location, A) already clear and B) needing clearing. Summary of the data (sorry for the small sample size; I have a low tolerance for boredom):

AVERAGE SIKANDER SPACES LOST TO WATER: 57%
AVERAGE SIKANDER SPACES LOST TO ROCKS/FUNGUS: 23%
AVERAGE % OF *NON-WATER* SIKANDER SPACES LOST TO ROCKS/FUNGUS: 40%

40% is quite a lot -- and that's not counting things like Sikander spaces tucked behind isthmuses, ones just missing a river, ones just barely not being a port, not to mention barely falling upon water.

By the way, it would be a MIRACLE if I didn't make a calculation error during my makeshift foolin' around in Excel. Anyone interested in verifying it can re-use the raw data given below.

Cheers,

USC

---------------------------------------------

START 1:
A) 3
B) 5

START 2:
A) 4
B) 4

START 3
A) 6
B) 1

START 4
A) 5
B) 3

START 5
A) 4
B) 4

START 6
A) 4
B) 5

START 7
A) 2
B) 2

START 8
A) 7
B) 2

START 9
A) 4
B) 4

START 10
A) 3
B) 1

START 11
A) 6
B) 2

START 12
A) 4
B) 3

START 13
A) 1
B) 2

START 14
A) 2
B) 2

START 15
A) 7
B) 2

* Actually, the "water" issue is non-trivial, as although water gets in the way of normal spacing more than fungus/rocks, I would think that giving up whole regions when water awkwardly interacts with strict Sikander spacing is worse than water's normal effect on spacing. But let's ignore that.