choosing your own tech to research problem

[mrdynamic]

Doea this ever happen to anyone else ? I turn blind research off so that I may bee-line to a particular tech, say, synthetic fossil fuels. I'll currently have reasearched both High Energy Chem AND Gene Splicing, but the of the techs shown to select what can be researched next, Synthetic Fossil Fuels will not be on of them. So Ill just pick somthing arbritary say Polymorphic Software. One my next tech selection, there will be Synthetic Fossil Fuels to choose from. This happens to on several techs as I try to beeline somewhere. What causes that ?? Is it normal or what ?

[bondetamp]

You usually don't get all the choices when playing on advanced difficulties, so there is still an element of chance involved even if it is a good bit smaller.

[mrdynamic]

Sooo... even if blind reasearch is off it is still blind reasearch in a way. You can say I would like better weapons you just dont know if the techs to choose from will allow you to follow your desired oatn. I think that is BS. Reason being is because even though I have the 2 pre-terchs required for mind machine interface, somtimes I have to choose and research up to 3 other techs before I have the MMI option. Which would not be all bad except there are already other factions researching the 2 SPs the MMI tech allows. It seems to happen mostly when I am doing better then the other factions as far is the bar chart goes. Perhaps that is the games way of keeping you from getting to far ahead ? Well, maybe I try blind research and see how things turn out.

[TresXF]

I think that is BS.

do you really want an easy game? i need the challenge and the chance to master the unexpected.
artificial intelligence is still miles beyond human brains - so what to do? give the ai some advantages and the human some obstacles.
no offense - but if you can't accept that you have to go playing with your model railroad

[Kirnwaffen]

I seem to remember a thread a while back about research selections being designed to prevent beelines up certain paths. I'll dig around for a bit and see what I can find.

[mrdynamic]

My question is just a query of, given blind research off, is the game supposed to provide randoms techs to choose from or not. Based on my experience it does provide randomness sometimes, and having blind research on ... definitely is more random.


TresXF , if I was worried about playing an easy game I know I could win I would challenge you to a game of multiplayer ........

[vitamin j]

Now, now, gentlemen, be nice. I remember the thread Kirnwaffen is thinking of. It was very educational. IIRC, I don't think the techs available are actually chosen randomly, it has to do with the highest techs you have in the various categories (build, explore, etc.).

[Flubber]

At any given time you will usually be given a research choice that does not include all available techs. I just accept that sometimes my desired tech will not come up. If you are beelining heavily, the best strat might be to pick some low level non-SP tech that the AI is trading/selling. You will usually get your preferred tech choice on the next tech lotto and carry forward your accumulated tech points and maintain the same tech cost (for that tech). The result is that you get your preferred tech this time just as fast but then have an extra tech that increases the cost of the NEXT tech.

[TresXF]

@ mrdynamic:
i thought the "no offense" and the " " would disable the possibility of any insult.

to make it clear - i was just making fun. i know that those random techs are wanted by the programmers. imho it's pretty boring if you have a 'golden way' to succeeding every game (like in many scripted rts for example).
btw, it's essential to make clear on what level you play - the game preferences are extremely different.

[knowhow2]

to be allowed to research a level 5 tech you need a bounch of level 4s, even more level 3, and almost all the 2s and 1s. So Beeline still require more tech than necessary for that peticular tech.

[MariOne]

Originally posted by TresXF
...the chance to master the unexpected...

...and following posts.

I remember a thread too, long time ago, where several "more or less" sound hypothesys were made about the rule/algorythm determining the techs to be withheld.
The more self-assured was MtG, but I have never verified his assumptions neither.

There is one way tho:
the mechanism with which the techs are withheld is DETERMINISTIC.
After several tests, I could not devise a formula, but I determined that if you put up a basic scenario, in which you just need to:
- use the SAME FACTION you are using in the game
- give it the SAME TECHS you currently know in the game (use Ctrl+F2)
Those are the only variables determining the techs that the game will actually offer you to pick, as far as I could tell.
Then, using Shift+R, you will ALWAYS see exactly the same techs portfolio that you would be offered in the game in the same situation. Of course remember to assign yourself also the tech that you are about to discover in-game (you will have it in the bag when you pick the next).

That is, you can't "calculate" which techs you will be offered or not.
But you can TEST it in advance, thus you can "know it" in advance, with 100% certainty.

If you had time, from a scenario, using F2, Ctrl+F2, Shift+F2 & Shift+R, you could attempt to devise a formula to be expressed in general terms.

[johndmuller]

I must have seen one of those threads because I have been operating on the assumption that given the same set of techs, you will get offered the same collections of new research targets. I have been basing tech-planning discussions with allies in MP on this theory (although without awareness that the 'same faction' was supposed to be a requirement also) and have not noticed any contradictions. While this is hardly a definitive study, it seems to point to a veryl high expectation of consistency at least.

Another theory I heard somewhere is that any withheld techs would come from the same category as the last tech researched. I can't say that I've checked this out - I suppose because it seems to be too much trouble to figure out all the techs you are qualified for - and I can't say that I have ever made any research target decisions on this basis, but it is all too easy to remember this theory, so I always think of it when I wonder what techs I might not get and it may have some influence on my tech selection even though I have no idea whether or not it is true.

[MariOne]

If the set of known techs determines the offer, then the theory that the witheld ones are determined by the category of last one discovered is easily proven false.
Just put up a test where you get to the same "status" from different discoveries, and you see that the offer will be the same.
The offered set is NOT dynamically determined, as I also had initially & erroneously thought.

Mostly, different factions are offered the same techs in the same situations. But sometimes not. I too was at first overlooking factions, and a couple of times (out of 30-40), some player I was trading techs with by leapfrogging reported that he wasn't offered the techs I previewed. I thought that he was lying to me (in-game deception...), but actually the theory proved to be incomplete.
IIRC you can verify it very early, something like give Mobility to Gaia and Ecology to Sparta, and they get a different offer for next tech (just an example, it might be some other factions couple), while most other factions get the same of one of the two.

Personally, I have verified the theory with over 200 observations throughout my pbems, and in some singleplayer test game, and it never failed once.
I agree that scientifically this is not a proof... as the fact that all humans died sooner or later... "SO FAR"... is not a proof that "a human" MUST statistically die...

[RedFred]

While it may be difficult to prove MariOne's hypothesis, you can play two parallel games. Make sure to turn pod-popping off and never trade or probe for tech. Play the same faction in both games. If you play Aki, don't capture bases from factions that have tech that you don't have. Research the same tech in the same order. Compare. I believe that most of us agree (I know I do) that in both games the tech choices you will be offered at every advance will be the same.

So there is no randomness. Question is, how does the program determine what to offer you. Is there hard coding every step of the way for every faction a set of tech choices or is there a simple algorithim to determine what will be offered based on what the given faction has already? If it is the algorithim option, what is that algorithim?

[T-hawk]

See here for Oedo's explanation of how it works in Civ 2. (it's about halfway down the page, or search in the page for 'Heureka')

http://apolyton.net/forums/Forum3/HTML/001533.html

It works mostly the same way in SMAC.

[MariOne]

All my observations deny the two rules reported there, thus also deny your statement that it works mostly like in CivII.

I don't know HOW it works, i.e. I can't answer to RF's million dollars question "Is there hard coding every step of the way for every faction a set of tech choices or is there a simple algorithim to determine what will be offered based on what the given faction has already? If it is the algorithim option, what is that algorithim?".
But I know that it DOES NOT work *that* way (CivII/oedo's).

If you prepare a scenario from scratch, without setting anything except the techs known to a certain faction, it has no history.
The tech offer always matches the tech offers observed in several different actually played SP & MP games, absolutely regardless of the history (i.e. how you got there, in which order, following which path) and the context (i.e. the conditions of the other factions and any other I could imagine).

I'll go with my method.
When I need to know it, I'll put up the simulation, and will hold it as bible.
This is useful if you are contracting to get a (human) player's next discovered tech in a trade (leapfrogging, or boosting an arretrated player to get from him the single one tech you're still missing).
You can preview whether he'll be offered something interesting to you too, and if after getting something from you in advance, he tells you he wasn't offered what you were expecting from him, you'll know that he's lying to you.

[johndmuller]

Marione, I don't think that oedo's/CivII mechanism is necessarily inconsistent with your observations:

1) Given the same set of techs, the "joker" techs would be the same;

2) the on-on-off alternations which partially determines whether a tech is available or not changes with each new tech and thus seems to be essentially a counter of your techs and would therefore always have the same value given the same set of techs;

3) if the two statements above are true then the tech offerings would behave just as you observed.


(to account for the slight variations you noted w/r different factions, perhaps the "joker" techs vary slightly by factional and/or category preferences - a feature either new to SMAC, not relevant in CivII, or swept under the carpet by oedo)

Probably because I didn't really look at the algebra much (since he prefaced it with a disclaimer), I didn't really understand what the implications of the quote below would be, so there may be a fly in my ointment.

starting techs, however, behave like never researched techs. that´s why they play into the formula above.

[T-hawk]

The practical upshot of the system remains true in Alpha Centauri as it was in Civ 2. If a tech you want is unavailable for research, be careful about what you do while researching an alternate tech -- if you acquire a total of 3 technologies via research/trade/pods/artifacts, your desired tech will again be unavailable at your next research choice. I've confirmed this effect several times while playing.

[Fitz]

Actually Mario, I carefully read the Civ2 thread a few months ago and concluded that the basic premise could be applied to the AC system. I unfortunately don't have the time right this second to reread it and find my case in point, but I remember your prior comments on it being determined solely by faction/given techs, and took his basic premise that it also was determined by the tech order inside a file, and figured it was based on the alpha.txt tech order (or another txt file order) in addition to the given faction/techs, regardless of the order to arive at those techs. Try rereading it if you find the time and you'll probably see what I mean. That would make it a simple three part determination, fixed on those three variables (tech order in file, known techs, faction [or faction's free tech])

You can probably test that idea with your test simply be rearranging the tech order in the text file and see if you get discrepencies with two tests, one right after the other. If you do it a good ten-twenty times and find no changes despite significant reordering, I'll withdraw my hypothesis.

[Zaphod]

RedFred, this "bump" is for you!

[Chowlett]

unfortunately, the thread referred to above does not appear to exist anymore, either live or in the archives. If anyone has a copy, reposting could prove quite useful.

[RedFred]

Thanks Zaphod.

The reason that I wanted this bumped is that it has been determined that some PBEMs do not follow the correct tech rules for a given faction. I now know of two instances when tech that should have been available was not. In the first instance I was playing the Cult and I was unable to take the ultrafast beeline to SoHB that the Cult is known for. In the second instance Zaphod did not have the Datatech opening tech options normal for that faction.

We believe that this may be a CMN issue. Unfortunately I still haven't gotten up to speed on CMNing. But is it possible for this problem to occur if factions are somehow reslotted when setting up the game. Or is it something more simple like not distributing the normal techs given to a faction at the start of the game?

[Gregurabi]

I'd be interested in hearing about any strange PBEM tech choice results as well. In both of the PBEMs I've played so far -- one as Lal, one as Morgan -- I was not able to choose Centauri Ecology as my first tech. I've never had that experience in SP play.

I'm also using the Linux version of SMAX. I don't know if that's relevant here.

[Mongoose]

RedFred, I wdon't think it would be a reslotting issue because you can out any faction in any slot in SMAX. I've not ever seen what you describe while playing SP; factions have been in every which slot during those.

I am sure that missing the starting tech would have an effect. I suspect that is the reason for the alternate choices. I wouldn't care to argue that the missing starting techs are the SOLE reason, though.

I have determined that what other factions are in the game, or were in the game, affects the choices offered later in the tree. That might be part of the problem here.

[Capt Dizle]

Wow, a whole nother subject I'd not thought to think about. This game makes my head hurt!

[Fitz]

RedFred, if I may further hypothesise:

If the tech order is determined by the three points I made before, ie - tech order in file, known techs, faction [or faction's free tech], maybe there is a fourth factor I missed. That factor would be # of tech choices made so far.

Thus, if you have a seven player game, the 7th player would be getting options based on the alogithm that would apply to the first player in the 7th turn of a single player game.

This could easily explain how strange results pop up in MP that you would never see in SP. If I recall the "missing" thread properly, it went through a 3 turn iteration of options every time you discovered a new tech. If this is the case, perhaps that iteration is not seperate for every player in MP, but instead occurs as you move from player to player.

The real test (or first test, I should say) of this would be to see if the first player always recieves the first set of tech choices that he would recieve in SP.

[big_canuk]

Furthermore then, a 3 player PBEM would give the same results as in SP. 3 player PBEMs are along with 4 player, probably the most common. MariO, any idea if your testing was mostly with 3 player PBEMs? Were any of the suspicious results with games with other #'s of players?

bc

[Straybow]

Careful about re-ordering the techs in alpha.txt. Some of them have functions that are not assigned via lines in alpha.txt, and those assigned functions may go to the [i]slot]/i] rather than the tech name or preqs. Now, wouldn't that be an odd surprise.

[johndmuller]

I performed the search that T-Hawk suggested and found the thread with no difficulty (today at least) - it was apparently repathed or whatever in one of those upgrades it is now located at: http://apolyton.net/forums/showthrea...hlight=Heureka
the "&highlight=Heureka" part is probably optional and doesn't seem to work in any event, but if you scroll down the thread a good bit there is a post by oedo that is the jumping off point to the theories in question.

As to CMN'ing (I'm taking that to mean using a Scenario) having an effect, that would be quite possible as the CMN needs to add in each faciton's starting tech manually as the game does not do so automatically. If there is a special "starting tech" routine or flag that a game started in the normal way would experience, it is entirely possible that a Scenario-started game would not execute that routine or set the flag. Further, it would most likely not increment the counter that is implied by the 1-out-of-3 withholding mechanism postulated in the theory, which could also lead to a different bee-line pattern. (In the same fashion, Actual in-game construction vs Scenario Editor placement of "goodfacs" led to a lot of confusion viz-a-viz clean minerals and the ED calculations - apparently due to the SE's failure to increment the >pop< counter when goodfacs were inserted by would-be theory testers.)

As to the 1-out-of-3 withholding theory, I have been assuming that to be true in my PBEMs for a while now (in fact because of this thread), and have not experienced any disappointments. In practice, it comes down to assuming that the tech will not be withheld twice in a row, for the most part, which is easy enough to deal with. The added consideration of making sure that you don't get exactly 2,5,8,... new techs in between by other means is also useful to keep in mind.

I've never gotten any deeper than that into this theory, but insofar as I have tested it in play, it hasn't done wrong by me that I noticed. For sure this is not a proof of anything, but it has worked well enough that I have on several risked pactmate relations (and ridicule / nerd accusations) by engaging in complicated manouvers to get to Industrial Auto relying on this theory as an underpinning - no egg on my face yet .

[RedFred]

b_c, in both cases I know of where early game tech choices weren't as they should be, were four player games. I have not played enough PBEM to know if this is merely coincidence or if there is some special magic to games involving more than three players.