Turning SMAX back into strategy game

[bvanevery]

5 PM.  I resume the struggle for human individuality.

MY 2436.  Domai needs 14 turns.  I've got far too many wounded units to continue towards Domai's capitol.  Instead I weaken a city I've bypassed behind me.  I discover that Geosynchronous Survey Pods can be attacked with a probe team after all.  This makes me wonder what was going on at the other base, why I couldn't see any Sensor Array that was nevertheless giving a defense bonus.  A bug?

MY 2437.  Domai needs 12 turns.  I take the city I weakened last turn.  Otherwise I still just don't have enough healed units to move forwards.  I plant fungus around cities I've already taken, to increase their minerals, and hopefully get me more units sometime soon.  A bunch of my mindworms get killed trying to take out enemy mindworms in my path.  They were supported by 1 Sensor Array in enemy territory that I didn't manage to destroy.

I design the Scout Skimmer hovertank unit, a cheap sacrificial unit to take out enemy Sensor Arrays.  A Cloaking Device would have been useful for this, as that also allows one to ignore Zones of Control.  Lacking that, a Hovertank Probe Team will accompany if needed.  I also design a Grav Scout Skimmer, my 1st ever use of antigrav struts to increase range.  I start thinking this is a good idea for probe teams as well and design a Grav Tank Probe Team.  They will take 2 turns to produce instead of 1, but this may be what's needed to stop Domai's project.

7 PM.  I weary of this and save the game.

[bvanevery]

12:30 AM.  The struggle resumes.

MY 2438.  Domai needs 10 turns.  I have to sacrifice some old units to remove a Spore Launcher that's sitting on top of a Sensor Array I need to destroy.  After that it's easy to get through the next base.

MY 2439.  Domai needs 9 turns.  I have a Grav Probe Team with 5 movement that could reach Free Drone Central, if only the roads weren't clogged with so many units.  There's a wall of mindworms and spore launchers in front of me and most of my units are not effective against them, either on offense or defense.  String Artillery is the only thing that works well, inflicting grievous wounds and allowing me to finish them off with a cheap unit.  I've switched to making more of them.  My own mindworm reserves are pretty much depleted from the unequal fights, but I've decided to make a few more of them for tactical reasons.

I've planted a lot of fungus around my forward cities.  In the short run, this leaves me quite vulnerable to enemy mindworm attack.  However this turn, I figure they'll probably just kill Formers, which can be replaced in 1 turn if need be.  Long term, the fungus gives me massive production, 2-4-5 per square.  My empire is about double the size it was, and the rate at which I produce units is increasing.  That's needed because Domai can spam out anything he likes in 1 turn!

Details Assault on Ascent

Assault on Ascent

MY 2440.  With 9 turns to go, I begin an assault on the Ascent to Transcendence.  Taking of Steelfount this turn gave me an alternate route to Domai's capitol, thanks to various rails he halfway laid down for me.  I don't even need to use the expensive Grav probe team to begin the attack.  The odds seem to be the same as any other targeted sabotage I've done elsewhere, but my 1st team dies.  My 2nd team prevails and lives to tell the tale!  This is the single biggest loss of production due to sabotage I've ever seen in any game.

Details Mission Accomplished!

Mission Accomplished!

Domai's doomsday clock is set to 26 turns.  Domai's counterattack is gonna suck next year, but it's definitely worth it to push the clock back.  Now, it is more about making a war with all of his mindworms.

I've realized that I could have the same level of productivity he does, if I want it.  My low factory output was always about fear of global flooding and mindworm stacks, but I think with the Voice of Planet active, it doesn't happen anymore.  I get fungal pops but no stacks, and I've seen no flooding.  Still, I somewhat feel like roleplaying the principle of the thing.  There's also the fact that soon I will take the Bulk Matter Transmitter.  I'm leery of the mineral increase effect it may have, so no need to add fuel to the fire just yet.

3:30 AM.  I do something else for awhile.  5 AM I resume.

MY 2441.  I take an unimportant city.  The rate at which Domai is spamming units is now a problem.  Fortunately the AI seems to be a little too timid to just come overwhelm me.  So I have to build up enough units to crush a base, to end its spamming.  Wet, lather, rinse, repeat.  This game definitely isn't won, but seems winnable.

MY 2443.  Domai came down with a whole pile of mindworms and spore launchers, taking back a city.  But a city full of Trance units does stop the advance.  When his mindworms are in my territory and not supported by Sensor Arrays, they become killable.  Perhaps he had Geosynchronous Survey Pods in his territory bolstering them before?

I've also discovered that I can afford Cost 4 Empath Song units in practice, with the amount of minerals cities have.  After experimenting with Trance and Empath units, I came to favor mostly Empath hovertank designs.  They can destroy Domai's mindworms and spore launchers just fine, and without an "invincible" horde, there's really nothing stopping me from eventually beating him.  He'll spawn, but he'll move his units poorly, and it's only a game of whack-a-mole.

6:30 AM.  Know what?  I'm quitting / declaring "victory".  The game is definitely winnable, it's just an incredibly tedious formal exercise to do so.  There's no point vs. other ways I could be spending my game developer energy.  TEST CONCLUDED.

[Alpha Centauri Bear]

bvanevery,
Remember you asked me to check my technology research timing adding technology values and research priorities into account to make sure they do not invalidate my results. I did. Against my own tree, not the original, though. Ran 16 different research priority combinations against a tree and averaged results. It mostly didn't push discovery times out of levels. I.e. technologies belonging to same level have close average discovery time. Few technologies on edge of level groups can overlap with technologies from neighbor group by 1-2 turns. However, since the standard deviation is about 5 turns, this is nothing and doesn't break general picture. See below.

[Alpha Centauri Bear]

By the way, I also created tree builder program. It does work for me and relieves me from tedious tree reshuffling.
It takes my generic lining structure into account plus linking rules plus whatever dependencies I specify. These are generic dependencies and not immediate prerequisites - they specify that some technology should be a descendant of some others. It also obviously takes desired sequence of discoveries into account and populates the tree. I hardcoded it for my case only as I didn't expect much demand on it in modder community. However, let me know if you are interested in such thing.

[bvanevery]

When working on the tree, I typically need to see all the capabilities and dependencies visually to feel sane.  I think I'm too tired to say more right now, it's 2:15 AM as I write this.

[Alpha Centauri Bear]

When working on the tree, I typically need to see all the capabilities and dependencies visually to feel sane.  I think I'm too tired to say more right now, it's 2:15 AM as I write this.

You mean like a picture or graph with each node containing all the information? That would be a huge work not only to draw it but to rearrange. Unless you are doing it on a magnetic board or something.

[bvanevery]

Actually you can't see the entire tree at once in SMAC, you can only see 1 node.  But I find that if I'm using a faction that has discovered all techs, I can iterate rapidly through the whole tree.  I click on the lower left listing of techs to randomly access the tree.  This is very, very helpful for my dependency shuffling and I do so much fine tuning of dependencies, I really don't think I could work without it.  It takes way, way too long to try to visualize and error check that stuff when only looking at alphax.txt.  Of course .txt editing is the only way to actually make a change, so I do that, task switching back and forth between the game and alphax.txt.  Awful lot of CTRL-ALT-TAB.

The main thing I actually wish for, is to be able to reload the tech tree instantly, instead of having to do the Start New Game... Load Game... F2... scroll down to where I need to be again shuffle.  Reload time is a general problem I've found in games that I've modded.  I am bound and determined to have that capability in my own game sometime "soon".  At the present time though, there's a 1.27 to kick out the door.