Wierd tech selections

[SerapisIV]

I was recently playing a SMAX game and was steadily focusing my research priorities after lifting restriction on getting neural grafting so I could use X-Marines on a heavily coastal Marr. Yet, despite solely having Conquer techs highlighted on research priorities I ended up at one point with Shard Weapons but only with only 3r/p armor and not even Choas weapons discovered yet. I've never had such a lopsided attack/defense run on techs without getting a pretty basic tech like Neural Grafting. Anyone else had such odd occurances?

[SerapisIV]

Oh yeah, in case you couldn't figure it out, blind research was on. Hey I know there have been polls on this but do most of you guys use blind research or turn it off?

[Kassiopeia]

That sounds almost like a bug, I never would of that that kind of an anomaly would be possible.
What comes to your second post, blind research is good to be off on Single Player, but on Multiplayer it's fun because you never know what's gonna pop up. I once cleared myself from a really hectic situation by using very tactically Air Power which I received on pure luck by blind research as a diplomatic trump.

BTW, what is "double-blind" research?

[cbn]

"Double-blind" is playing blind research with all 4 or none of the preferences selected. In theory it gives the most random ordering of the techs.

Multiplayer USUALLY does not involve any kind of blind research since it can be incredibly unbalancing if one player cannot even get say . . . Centauri ecology for a long while. Ditto for Ind Auto or MMI-- a player can get shafted by the tech lotto.


I find that blind reserach is the way to go in Single player.It is just one of many handicaps you can impose on yourself to make the game a little harder

[Horus]

As someone else pointed out some moths ago, just selecting one (1) research path and stick to it, migth provide an even more difficult situation for you (e.g choose "Conquer" if you got a huge map and wants a builder game).

[Bob Morane]

I think blind research sucks.
I like knowing what i will get, rather then playing lotto with the technology.
And if i want to increase the difficulty, i would rather play at higher difficulty level (no, i don't already played at transcend level, i'm quite new with SMAC ) or use super AI's wich are easy to create than blind my research.

[cbn]

Horus

It is true that selecting a single preference can lead to a difficult game BUT it also means that you can plan somewhat based on the techs you are likely to get. You won't know exactly but you can get a better sense than using double blind.

FRom my double blind games I did get the sense that the tech lotto is not completely random anyway. This was confirmed by people that indicated that every tech has a ranking on each of the 4 research categories and that the computer tends toward the higher ranked tech.


A related question from one of my MP games. Is there a way to influence the tech choices you get when playing directed research? In one game, MMI just won't come up even though I have had the prereqs for about 2-3 choices (Neural grafting was slow appearing as well). I am the Hive and already have fusion power, spaceflight and am now working on the TOE tech. I USUALLY get those after MMI. Its frustrating the heck out of me since I am at war with a faction and choppers would make a killing difference. I figure it has to show up next and in the meantime I will just continue harassing him with my needlejets LOL.

[Ned]

I've had games using blind reseach where I believe I've gotten every possible tech one can get without getting Tree Farms before I got that tech. I believe I had advanced space flight and the like. Not getting that tech for so long is a real pain in the a**. It just may be the single most important tech in the game.

Ned

[Avenoct]

To address the original question:

I've never had such a lopsided attack/defense run on techs without getting a pretty basic tech like Neural Grafting. Anyone else had such odd occurances?

Quite often in my experience. A large factor though is tech-trading, which you didn't address. If you trade for a tech that you couldn't have researched, say the preqs for Shard, but not for Chaos, then its easy to see how you get Shard before Chaos. I've never seen blind research skip over a preq, but it does funnel down narrow pathways on occassion rather than getting the full broad base of a particular category.

I suspect there is an answer to this:

A related question from one of my MP games. Is there a way to influence the tech choices you get when playing directed research?

I believe the same 'AI influencers' that categorize what tech-category (Conquer, Build, etc.) a particular tech falls in, and also how popular it is with the AI (by the size of the number), also determine how likely the tech is to come up as an option in directed research.

For example, if you assign the following influencer to say 'Nonlinear Mathematics" [20, 0, 0, 0], it seems to appear in your choices far more often than if you set it to [1, 0, 0, 0]. Obviously, this is when you have the prerequisites for the tech in both cases.

As a side note and further evidence, in the game Aldebaran, I've set the AI influencers quite high in most cases to help funnel the AI down particular research pathways. The secondary effect of this is that, even on Transcend level, the player is faced with from 4-8 tech choices each time, rather than the 1-4 you might see in original SMAC at Transcend Level. (note to self: Better change this as its only the relative difference between techs that funnels the AI, not the actual values, so 20:10 might as well be 2:1 to avoid this multi-tech effect).

Happy Crawlering,

Smack

[SerapisIV]

Originally posted by Smack
Quite often in my experience. A large factor though is tech-trading, which you didn't address.

I traded through the level 1 and 2 techs to parity, however after that, the AI was always behind me and tech and pretty much followed my tech track as well. Not sure how or why. Like I said, it was really just wierd how it happened. But between impact and shard, I got all the techs by straight research and did had all the pre-reqs well before I finally got Neural Grafting.

[Fitz]

FRom my double blind games I did get the sense that the tech lotto is not completely random anyway. This was confirmed by people that indicated that every tech has a ranking on each of the 4 research categories and that the computer tends toward the higher ranked tech.

This shouldn't be the case. Yes, each tech has a ranking based on the catagory, and the most likely tech you will get is the one with the highest total value from the catagories you select. But if you go with no catagories, then all values should be 0.

If you go with all checked, then you will get the tech with the highest total score, provided it is available (1/3 of the techs are not available every time).

If you want to see the tech values, open up alpha.txt or alphax.txt
Each tech has the four catagory values next to it.

I once calcualated the techs you would get if you used all four catagories based on those values. The order was close, but not exact due to some techs being unavaible on some turns.

Ned, the value for Tree Farms is fairly high in Build, and somewhat hight in explore, but not very high overall. Same with some of the techs leading up to it. I played every faction using it's blind research priority once, and I found that only the factions that have Build as a priority get to Tree Farms fast. Morgan was the fastest.

[johndmuller]

It isn't discussed very much for some reason, but one can play Blind Research in an "Active" mode - a sort of chancy Directed Research - it ought to be just right for those who find the Static Blind Research too frustrating, but want to give back a little advantage to the AI. Just figure out what tech you think you want (as if you were playing Directed) and select ONLY that category as your research priority. There's usually only one or two techs from that category for which you have the prereqs, so it is far from random, although the program does not limit itself to just the category you select.

While the research is underway, you also get a clue as to what it is going to give you if you look at the Lab (F2 I think) screen. While in Directed Research, the exact tech will be spelled out, in Blind it is only color coded to the color representing the category of the tech you will get.

I'm not sure when your category specification takes effect - that is, I don't know whether you can change the tech being researched by changing your selection of categories (easy to test though); if you can't, you probably have to make your selection before the previous tech is done which would sometimes put you in the position of deciding how much you wanted to be sure you got the exact tech you wanted (instead of just a category cousin) this time versus goint in the right direction for the next tech. I suppose this system works best if you always have two different techs in mind so you always go to the other techs category each time.

I haven't played this way as much as the Static Blind or the Directed, but when I have, I don't recall getting particularly frustrated, so I guess it worked out OK.

[Mr. President]

Re the original issue: Some of the armors come with Build technologies (eg. Silksteel with Silksteel Alloys (Build 5), Probability Sheath with Probability Mechanics (Build 7)) or even Discover technologies (eg. Psi Defense with Eudaimonia (Discover 11 or 12)).

[Flubber]

BUMP