Blind and Semi-Blind Research

[Hydro]

I’m a big advocate for Semi-blind research since I think it promotes a bit of realism to SMAC, and makes boring technologic b-lines less powerful. It also makes the game against the AI a bit more challenging since the tried-and-true-hose-the-AI battle plans can go amiss (**gasp**) and you might actually have to work for your dinner.

One of the big criticisms is that Semi-blind research is a tech-lotto. This is not true since you can guide your researchers by setting your research emphasis to Explore/Discover/Build/Conquer priorities or any combination of these. True Blind Research sets all these to null, or selects all. Then you have a true tech lotto.

But, how do you select E/D/B/C to get the best chance to acquire a critical technology? The much-maligned Prima Strategy Guide showed the E/D/B/C strengths of each tech. By looking up the techs you are eligible for you can see what the relative emphases are, and then set your research priority accordingly.

The Prima Guide has been out of print for a long time and is difficult to find. To help those interested in Semi-blind research I summarized their E/D/B/C Tech Ratings, and put all the tech abilities there, too. This is a simple all-in-one tech reference sheet. Note that this table is for SMAC. In SMAX the pre-requisites changed to fit in the new technologies, and the E/B/C/D for these new techs were never published.

I encourage all of you to give up your risk adverse and anal-retentive SMAC/X proclivities (**note the gentle mocking tone here**). Try semi-blind research. You may even like it!

Hydro

[Whoha]

I just set them all if I play blind research(I think thats what double blind is though I'm not sure).

I really find though, that it is a tech lotto, and that not getting IA as fast as someone else is a real problem(for me, not for you people that eat transcend AI's for breakfast).

[AndiD]

Can't you just look up the E/B/D/C values in your alphax.txt?

[GeoModder]

Yes, you can. There next to the tech prerequisites.

[CEO Aaron]

I think the general resistance to blind research is that most players don't want to be jacked for a lynchpin tech for 10-30 turns. IA, CE, EE, and D:AP are highly critical, and missing out on any of them can severely stunt your development.

Does knowing the tech progression make your game easier? Sure, but getting hosed down because you keep not getting a vital tech isn't fun, it's just being beaten up by the pseudo-random number generator. Personally, I'm still finding the AI plenty challenging at Transcend, mainly because I play Morgan on a pretty tight map.

[GeoModder]

Then you need to steal it somehow.

Probing or (when tech steal is on) take a base.
Even trading with the AI is a possibility.

I'm playing a SP game with blind tech and tech steal on (small map of Chiron), and things are going preeetty well by now (MY 2285).

[Hydro]

Hmmm. I’ve never see the AI do anything with the ‘critical’ technologies like IA (I did see the Cyborgs build and use a crawler once, and only once), D:AP (except for taking out unarmored units) or EE, so not getting those ASAP against the AI doesn’t seem that all-important to me. The only place I feel difficulties is if I don’t have some basic armor or weapons when the AI comes calling with dreaded impact rovers (Yang and Santiago come to mind).

What I have found is that not getting that critical tech makes me be more creative in trying to survive the early game, and more appreciative when I do get the tech-of-interest. For instance, I am much more interested in aggressively harvesting Unity pods to get that elusive AA and linking to a NN, and my strategy for almost all the factions relies on this to augment research. I also trade almost any tech for any tech with the AI under the theory that even if the tech isn’t on my target tech I will now not have to waste my time researching it myself. I don’t care one bit that the ‘useless’ tech will increase the time it takes to get my next tech.

In the end I guess I dislike b-lines. They are too staid, artificial and (as I said) boring. Blind and semi-blind research is spicier for me, and that is critical since I’ve been playing SMAC/X since 1999. But, that is a matter of personal taste.

Hydro

[GeoModder]

Well, with blind research on chances are that regardless the AI considers them important or not one of them is bound to stumble over such a key research, trade it on to other AI's and so you might have a chance to trade it as well. As for your crawler issue, in the forementioned game I see a PK crawler on my border, though it's doing nothing in a fungal field. But I had it that the only way an AI is finishing a SP before me was with disbanding crawlers in the producing base.

[CEO Aaron]

It's not that critical techs are dangerous in the hands of the AI, it's that without them, the innate bonuses the AI receives (huge growth and industry advantages) are nearly insurmountable without them.

If I'm playing Morgan up close and personal with an aggressive Yang, I NEED to make sure I get Biogenics, absolutely first thing. Without it, I'm just mediocre meat for a more well-equipped neighbor.

Some factions do very well with random tech, but the truth is that if you're stuck on an island without foils or trying to capture techs but can't seem to get probe teams, you can do everything right and still lose, especially if you're playing in tight quarters.

I also don't think that playing with random tech lends itself to an equitable contest for multiplayer at all, as the player with a more favorable tech awards will find himself in a superior position to a player whose played a more solid opening but has been shafted by the tech lottery. Does that mean that blind research is unfair? No, but it does mean that less of the contest is determined by skill than luck.

[Vev]

However the odds can be rigged, through limiting which techs are available to be researched. But it is still random.

[Hydro]

I think there is some confusion between blind and semi blind research. Blind research requires you to select all or none of the E/D/B/C, and then you are subject to tech-lotto. In semi-blind research the techs are not totally random since you are setting a research priority, which will certainly cut out about half of the possible techs from serious contention. The tech you get is weighted toward your research priority, and then (as has been noted) a random factor applied.

Yes, there is a random factor. But combat has a random factor, too. So is the quality of your initial placement. So is your chance of a good diplomatic relation with that AI. So are the results of pod popping (unless you cheat and reload until you get a good result, and that doesn’t count). All of these do have a critical effect on the game, and can be a tipping point between success and failure. The point is that the basis of the game is the aptitudes of a faction, and the strategy you use to achieve supremacy. Throughout there is luck and chance involved.

I agree that totally blind research takes away this portion of the game from the strategic desires of the players. On the other side, using directed research makes this linear with b-lines making non-b-line techs worthless and a distraction. In the end a person’s preference between directed, semi-blind, and blind research is based on how risk adverse they are.

I also disagree that semi-blind research (or even blind research) is not equitable in multiplayer, and I don’t see how this favors or penalized any given faction. Possible exceptions are the tech challenged factions (Drones, Believers, and possibly the Hive) that can take longer to recover from a turn of bad luck; but that is their penalty they take for their advantage(s). After all, everyone is playing by the same rules, and your faction is as likely to have their research go astray or stay on target as the next faction. As mentioned above, this is simply one more element of chance that the player can have some influence on but not totally control.

Hydro

[CEO Aaron]

I also disagree that semi-blind research (or even blind research) is not equitable in multiplayer, and I don’t see how this favors or penalized any given faction.

Perhaps inequitable was a poor choice of words. A coin toss is an equitable contest, but the relevance of player skill is virtually nonexistent. My argument against blind research is that it makes the game resemble a coin toss more and a chess match less.

Yes, there is a random factor. But combat has a random factor, too. So is the quality of your initial placement. So is your chance of a good diplomatic relation with that AI.

Diplomatic relations are not random. The algorithm is a direct result of a series of arithmetic operations which include your relative power, similar or opposing SE choices, and prior diplomatic history. As for combat and map placement being random, I view that randomness as a necessary evil of the game design. I've often lamented that your opening position is the primary indicator of your likely success (or failure) in all Civ-like games. However, I concede that it is a feature which lends the game considerable replayability, so I forbear. And as for combat, there comes a point in the early midgame where the randomness of combat is swamped by the predictable results of a more powerful economy. It is this feature which attracts me to builder-style play. I _am_ risk averse, and I don't like a momentum style of play where random chance has a large influence on the outcome of my entire game.

I'm not saying that you shouldn't play with Blind research off, I'm just explaining why I generally prefer not to. I want to win my SMAX games based on the strength of MY decisions, not my processor's.

[Hydro]

Just as SE influences diplomatic relations, so does the research emphasis in semi-blind research. Both still have a random factor in them, so in my mind this is a pretty strong argument for semi-blind research.

I also think we are talking past each other. To be clear, I am talking about semi-blind research, not blind research. Your arguments are perfectly valid for Tech Lotto. Semi-blind research (or you could say semi-directed research) is a different animal. Semi-blind/directed research is not a mere dice toss. It might be more accurate to say that you’ve loaded the dice with your E/D/B/C research priority.

Hydro

[johndmuller]

First of all, Directed Research is not completely fair to everyone in a PBEM either. It is entirely possible to be denied the ability to research a desired tech due to the workings of the mechanism which withholds the offering of certain techs each turn. This can even happen on the first tech choice in a Scenario generated game, which can be rather unpleasant if you were planning on something in particular (CentEcol, for example). OTOH, this mechanism is pretty well (if perhaps not perfectly) understood and is reasonably predictable, if not necessarily obliging.

To the best of my knowledge, the algorithm for semi-blind research has not been nailed down, but I would not be surprised if it were possible to manipulate the results of semi-blind research nearly as well as that of directed, if the full mechanism were understood.

It seems quite likely that the scores (in the four categories) of the available techs (the ones for which you have the prereqs) and the combination of categories that are chosen for emphasis by the player would be major factors in the process. For example, the techs could be ranked by adding up the scores of each available tech for all of the selected categories. At the simplest level, the highest ranked tech would be the one you would get; unfortunately, however, I don't think that it is quite so simple as that.

OTOH, it may well be that it is not very much more complicated than that either. For example, if the same mechanism that operates in directed research to withhold offering groups of techs each time were to also serve as a filter on the choices from the ranking method above, it could easily make it mysterious-seeming enough to keep the casual investigator from seeing the pattern - I for one have never tried to keep track of the possibilities well enough to evaluate even this relatively simple possibility.

Another wrinkle I have seen mentioned is the possibility that the game looks ahead 1 additional tech when deciding which techs are to be ranked. In other words, it figures out which techs you could possibly research next and also which additional techs you could research the next time given that you researched any needed additional prereqs the first time, and then after ranking all these techs as above, it gives you the favorite (or its first round prereq if it were a second round choice that was selected). Personally, I think this mechanism is a little too complicated to be chosen (there are also some loose ends, like what if you need to research 2 or more of the immediately researchable techs to get the prereqs for a second round tech - would that be acceptable? and if so, which of the prereqs would you get?).

Well, if it is predictable, it is seemingly more difficult to predict, as I am currently in a PBEM with this semi-blind thing going, and we are at 2259, having gotten Fusion not so long ago in about 2245. While we got Doc:AP around 2210, MMI has not been forthcoming - in fact NeuralG has been holding out on me since then, despite my attempts to tilt at least somewhat in that direction. (I can't say I've been obsessed with NG to the exclusion of all other categories, as there have been other techs that were not unwelcome, but Combat has been selected from at least that point, and I have personally researched at least 8 techs since then and had both of NG's prereqs that whole time without getting NG).

I'll have to say that research progress seems to have been slow in this game, but OTOH, we are none of us real research factions (Humans are Hive, Believers, Cult and Gaians; AI is Sparta and Aliens) and some techs have been researched by more than one human faction.

Even with this quantifiably slow research pace, the tech rate has still been moving along faster than anyone can fully exploit the techs they do have, and aside from my frustration with NG, I have not felt particularly put out by the seeming vagaries of this mode. On the contrary, I have enjoyed it more than the average PBEM and would like to play more games with Directed research turned off.

[Frankychan]

I almost always play Blind....I find that it is much more realistic than directed research. I mean, I can imagine my scientists researching stuff then all of a sudden, they go "Eureka!" and suddenly I've got terraforming.

But how in the world do you do Semi-Blind research.