T-hawk Plays Alpha Centauri

[timmy827]

Fabulous read!  Eager to dust it off now though worried it will all seem horribly unoptimized

[Modo]

There's one loophole associated with this.  Start a turn with a facility as the current build order.  Switch to a project.  Add a supply crawler.  Switch back.  There is no penalty for switching back, so you effectively applied the crawler towards the facility.  I forgot about this when I was listing possibly-bugs early in the the thread.  That's strictly worth doing if you also flip SE along the way to increase the crawler's value.  Upgrading the crawler for this trick can also be more efficient than rushing the facility directly, though most facilities aren't expensive enough for the upgrade to gain much if any.

Was juggling several scenarios in my head to use this, I've come to respect the boxes (food box, prod box) like they were used in these old titles more as opposed to the newer implementations...less prone to exploiting it seems.
As long as the box is independent of the item you avoid a lot of hassle. But just to put it to rest, there's no way to store production in an item that's not currently being produced, is this correct?

EDIT: A couple more minor tidbits from earlier turns

1) in 2113 you mention not wanting further contact with Lal after taking out UN HQ, is contanct in SMAC based on base discovery and not cultural borders?

2) in 2115 you mentioned the original starting Independent (supportless) scout, once it got killed you lost that benefit so only this particular unit has it and it's not like you get one free unit in general?

3) in 2116 you got a free forest growth, do you happen to know if this works based on adjacency with other forests (Civ4 style) or any other factors? In other words, can you reliably influence it?

4) Are the base / citizen automation options of any use to alleviate some of the micro or you're better off keeping manual control?

[T-hawk]

Yes, it's all about the boxes, there is no way to store production in another item or tech.

1) Contact is based on meeting another unit or base, once that happens once you can always contact forever (except sunspots.) What I was avoiding was a specific AI behavior: when you move a unit up next to their unit or base, they will often open communications.

2) Right, no free unit in general. It's possible to get more Independent units: pod-pop or bribe units or capture mindworms far from home, or disband a unit's home base by building a colony pod.

3) Nothing influences it AFAIK. (Although it's really easy to reload and save-scum it.) I do know growth "passes along" between forests: if forest A grows and picks forest B to grow onto, then forest B will attempt to grow instead. This continues until the growth succeeds (finds a valid empty tile) or fails (finds an invalid tile, which is rocky or sea or already occupied by a unit or most terraforming improvements although not condensor/borehole/mirror.)

4) I never used any so don't really know. I think not since the governor always wants to build defensive units you don't need. And it definitely doesn't know how to pop-boom. The one automation option I sometimes use is a former on automatic roads/tubes.

[Modo]

In Civ4 I use the gov to keep citizens assigned correctly and to assign new ones correctly as well in case I miss a city growth here and there so was curious if we can at least get that in SMAC too..altough given how fast and involved SMAC is it wouldn't be needed anyway.

I'm a bit taken atm with this partial rush buy thing, can this game be played efficiently relying only on energy and the partial rush buy and stockpile bug to bypass minerals?

Also, don't remember seeing anything about unit promotions in SMAC so curious about natural unit healing; does it work based on enemy / friendly cultural borders / netural territory as well?

And one other thing I did remember, does a crawler still work during the turn when it's upgraded?

[T-hawk]

A newly-grown citizen is always assigned to something right away, in all of the games, even without any governor active, so I'm not sure what you're asking.  SMAC doesn't have any buttons to focus the new citizen on a particular type of production or specialist job, if that's it.

You'll always have at least some mineral production, it's not an either-or choice, you use both.  Energy can't pay mineral support for units.

Unit healing, I'm not sure exactly, I never had to heal in enemy territory since that territory kept becoming mine instantly.  I think it's 10% outside base including enemy territory to a maximum of 80% health, 20% inside base, or 100% inside base with the right type of facility (command center/naval yard/aerospace complex for land/sea/air, or Nanofactory SP.)  This is the same as Civ 2.  There are no promotions; morale is the equivalent of experience.

A crawler works after upgrading if you select it and give a convoy order again.  I did mention that in the game.

You know, with the amount of time you've spent asking questions you could have played through half a game yourself.

[Modo]

I know, I know...but actually seeing you do it in SMAC has brought me back to other older games that I'm proficient at. Ultimately the goal is to improve one's ability and that's done in a game you're already familiar with, SMAC will wait for the next two free weeks when I'll actually have time to enjoy learning it myself. Adding together the times when I've asked questions unfortunatelly doesn't equal the same amount of time in continous playing sessions.

Yes, I wanted to understand if we could influence how SMAC focuses new citizens and I indeed forgot that in this game the support for units is paid in minerals.

[T-hawk]

Going back to HOMM, I don't know anything about multiplayer of any version of it.

Here's a better description of "why not HOMM" that I remember now.  Back in the day, somewhere around 2000 or 2001 before the Civ 3 Epics, I read a writeup on HOMM 2.  It deconstructed each campaign scenario, establishing the minimum theoretically possible time to completion, in terms of movements of heroes from where you could hire them to the locations you had to reach or conquer.  HOMM is a lot more solvable than Civ in this way, since the movements and goals are so more tightly defined.  Most scenarios could be completed in something like four to eight days/turns.

Luck-abuse (not stated in such terms but basically assumed) would get you the weakest opposing monster stacks, best spells from your own castles, and key skills at level-ups like Logistics (movement).  The author demonstrated an actual execution of his theoretical minimum for something like 18 out of the total 20 campaign scenarios, with the other two needing just a bit of extra time to get enough troops to win the needed fights.  If I were playing and writing about HOMM 2, this is what I would be trying to do, including creep-saving.  Once I knew HOMM 2 was already solved to that degree, I didn't have any need to attempt anything more.  Nor HOMM 3 because I just wished I were playing 2 instead.  There is no other goal besides speed in HOMM because anything not constrained by time can just be "solved" by waiting months to accumulate thousands of troops.

That sort of theoretical minimum is what I find interesting to solve, finding the best possible outcome by some objective measure.  Speed was that measure in HOMM, Civ 5, SMAC, Pirates, and for a while Civ 3 culture games as well.  I kept going on each Civ title because I knew I was smarter and better than the guides and others who had produced some such material.  With Civ 4 I found in-game score more interesting than raw speed, hence my couple writeups on that, but then others on Civfanatics took both the speed and score torches better than I had been doing.  I would do NES console speedruns but others are already doing that better than I could.

The objective measure explains the creep-saving for all of them.  I care about that objective measure, not about what any particular person (such as myself) might encounter on some particular playthrough with some particular random results or suboptimal decisions.  For competitive RB Epics, the goal is defined as your result as determined by all of that stuff along the way, so I found those interesting too (and complied by the rules.)  For Final Fantasy 5, simply finishing at all under a solo or variant restriction was interesting enough to constitute an objective measure, even if it involved thousands of reloads.  For FTL, the entire point of the goal by definition is without save-scumming.  And I even set the one Hardcore goal to do FF5 without ever reloading.

Short version: I need goals, that either stand up to or explicitly disallow creep-saving, and that someone else hasn't already done better.  HOMM is sufficiently constrained on the possible goals that it ran afoul of the latter.

[Modo]

Interesting that all the games you've mentioned and that I know are TBS (from what I know Pirates and FTL aren't and FF is a mix?) so I assume something like Factorio would not strike your fancy?

[abuzeus]

T-Hawk playing Cracktorio

That's an interesting prospect. I was going to reply that he obviously wouldn't play such a game, but then I remembered him playing that bullet hell, Jamestown (So, it's not just planning, but also execution under time pressure). I guess that most Factorio goals are not interesting (building an X spm megabase is just a function of time invested), but time to first rocket might both be short enough and involve enough planning that pitting T-Hawk against Anti might be real fun. Likely that we will never get it though.

T-Hawk: Thanks for the report, an enjoyable and educational read as always.

[T-hawk]

Pirates, FTL, and FF5 are all soft-real-time, you can pause as much as you want so real-time reaction speed isn't a factor.  I like to say they're turn-based with turns progressing as often as you want.

I can play hard-real-time games too, like Jamestown and other shmups.  But they don't work very well for screenshots and written reports.

I hadn't heard of Factorio; I don't really pay attention to whatever's hot on Steam at the moment (FTL and Jamestown were exceptions.)  It looks like an open-sandbox type of game, which don't really appeal to me since yeah most things you can do are just a function of time invested.