posted 11-16-98 12:18 AM ET        

Starcraft- "Once I research enhanced tank weapons level 3, I'll destroy him".
SMAC- "Or will I?"

posted 11-16-98 12:52 AM ET            

Anyway, it does sound like a good idea (the random part anyway). However, I don't know how well it would play, ie how much fun it would be. Science would become guesswork, and you'd never know what advance would lead to what. May be realistic, but I don't know how much fun it would be. And in the end, the game is about fun.

I'm not saying that it wouldn't be fun, just that I don't know that it would. Could get very frustrating. Imagine if it was random, and somehow you had very few military type techs. Make it very hard for someone playing the Spartans.

And I doubt that too many games companies will ever implement such a system. It would simply be too hard to align everything. You'd have to either have interchangeable techs, in a fixed tree structure so that the advances would make sense. Or a purely random system, where one tech advance wouldn't naturally lead to the next.

And imagine trying to structure it so that it would be balanced for every game. Ai carumba, that would be a hard task for a completely random tree.

If someone could ever do it, I'd be willing to give it a go. But I think the problems of balancing would make it unuseable.

posted 11-16-98 09:29 AM ET            

Yeah, but I've been persuaded (a bit) that at the moment it would be impractical. However I feel that the civ style is slowly developing (like introduction of more and more functional UN in MOO2 and now in SMAC, or customisation of units), so eventually we might see some new tech system.

What I didn't like in civ was the lack of realism. But I agree that the playability of the game is much more important then realism.

Some of the features I'd like to see in an 'ideal' tech system:

-account for civilian research, that cannot be strictly controlled (other then by allocating research grants)
-idea of 'concepts' - as discussed on the old forum
-discovery of new concepts might make disovery of other objects more/less likely
-A concept can be simply spotted in an enemy civ and if tech is advanced enough it would become immediately available for research
-You cannot be sure whether a concept can be researched into a real tech (i.e. time machine exists as a concept, but if for example US would start funding a research project, the likelyhood of them succeeding would be small)
-being able to have more than one research project (this has also been debated, but no one agreed on that one either :-) )

I know this is quite difficult to implement, but I've been playing with relatively simple system that uses the concept idea. It is not really a random tech tree (a flexible tech tree?), but it does introduce some ranomness into it that I think exists in the real world without spoiling the game play.

posted 11-16-98 12:07 PM ET            

Basically, what is needed is multiple paths to the same advances. If this had been done in Civ, then it would have been possible to develop a civilization that didn't totally parallel historical civilizations, which would have been more fun.

posted 11-16-98 03:48 PM ET            

Gameplay is obviously of paramount importance, but gameplay does not always equate to fun. As the gaming population ages, we (I) seek more intellectually challenging games. For me, the fun lies in the challenge and the realism of the simulation. Seldom does life have known outcomes, especially in the world of research/invention.

Civilizations never have the best of everything (other than the US? Crud, that destroys that idea...)

"The last 1% costs more than first 90%"
Perhaps that idea could be integrated- research yields a base level of knowledge (though effectiveness should still be varied) which can be refined through further research.

Predictable research outcomes bother me. Dushan had a good point about the time machine. Perhaps there should be blind-end research paths. But they can't be fixed or we would never research the same path again.

posted 11-16-98 04:45 PM ET            

I liked both the way that MOO and MOO2 handled tech advances. In the first you would only get a random subset of the total advances in a category and in the sequel you got to choose which one of three you got from a category. This added to the diplomacy as well as it paid to have friends who you could trade with (and also made having defensive spies worthwhile to protect those techs you had that noone else had). The only problem I had with Civ is that everyone eventually got all the techs. With MOO and MOO2, everyone would have something slightly different and you had to play to your strengths or weaknesses.

From a gameplay balance point of view, I don't think that a totally random tree/list/pool of technologies would work very well. Unless a completely different model is used from the standard "techX yields unitY and buildingZ" model we're going to have to live with what we've got.

As an alternative, you could divide reasearch up into "Theoretical" and "Applied" trees. Make the "theoretical" advances 5x to 10x longer to research than the applied but necessary before you can research the applied stuff. Make the entire theorectical chain available to all players, but make the applied chain have a bunch of different mini advances that the player can only choose a certain number from (3 out of 5, say).

Example:
Theoretical - High Energy Laser Optics (100 beakers)
Applied -
Holodome [building] (+morale) (10 beakers)
Laser Rifle Infantry [unit] (5 beakers)
Laser Tank [unit] (10 beakers)
High Speed Networking [advance] (beakers x 1.5 in all cities) (15 beakers to research)

The player gets to choose 2 out of the 4 applied. The tank would be more fearsome in battle, but you could make 3 infantry for every tank as a trade off.

Finally, the player could set the percentage of research to be spent on theoretical advances and the percentage to be set on applied advances. (e.g. 40% of the beakers goes to the next theory while 60% is spent on using those theories to make applications.) Set one too high and you run out of new applied things to research or having no applied things to research.

posted 11-16-98 04:59 PM ET            

I don't like when you can't research certain techs.

As to the realism - you're right that too much of it can hurt the gameplay. I really don't know if we'd gain anything from messing with the current tech system. But I think that unless someone tries, we'll never know.

To the list of 'features' of an ideal tech system I add this:

-allow for techs that are usable before fully researched. In civ you research maths early on and that's it. In reality, we're still doing into new areas of maths. This could be simulated by having some sort of level of advancement - the more research you do, the higher level you gain. You'd need certain level to be able to research new techs, and also your current level in some techs might influence the quality of units you produce. Note that this would only apply to techs such as maths, physics, construction, metalworking, etc., not techs such as wheel.

posted 11-17-98 07:07 PM ET            

In CIV II for example, you could research Mathematics, say at 100 beakers. Then you could decide to research catapults, say at 40 beakers, or start researching construction at 100 beakers.

If there were enough "applied" advances off each "theoretical" tech, and the research amounts were well balanced, it would make for an awesome balancing act between going for the latest greatest "theoretical" advance, or consolidating your tech with lots of small "applied" advances.

Maybe each "applied" advance after the first would be easier to research (seeing as how you are getting better at utilising the theory. And of course you could trade for both types of tech.

Now that sounds cool. =)

posted 12-17-98 12:34 PM ET            

In Moo2, every advance in a given field of research (Physics, Biology, Construction, Chemistry, Sociology, etc.) led to a small improvement in stats for applications of earlier advances in the same field.

Say you researched Nuclear missiles. Later, when you'd researched Pulson missiles, the nuclear missiles suddenly cost less, didn't take up as much space, and you could build faster nuclear missiles if you wanted.

This sort of improvement continued after you'd outhgrown the tech tree, so it didn't feel so futile continuing with research.

posted 11-18-98 12:49 PM ET            

Upon reflection, I think that it would be better to allow everyone to get all of the techs. Probably would add just as much strategy as choosing which ones to go for.

I think you'd probably want 3 types of buildings too. Things like commercial, private sector research labs whose points/beakers ONLY went towards applied research, things like universities whose points/beakers ONLY went towards theoretical research and stuff like military research centers whose points could be split any which way between applied and theorectical. Any population that is specifically turned into scientists could also have their beakers split any which way.

I also like the idea of the cost for applied research being a diminishing scale regardless of the actual technology. So the first application costs 20 beakers, the next 15, the next 10 and every one after that 5.

I don't think we'd need MathI, MathII and MathIII though. I think just having slightly more descriptive names like Algebra, Trigonometry, Computronics, etc would work. It would also be more suggestive of what the applied research would be. (Computronics for instance would lead to encryption and codebreaking applications.)

posted 11-18-98 01:06 PM ET            

With the theorectical and applied split, you could split wonders into three categories.

1) Wonders that can be constructed immediately after theoeretical research.
e.g. build Leonardo's Workshop after Algebra (or some such theoretical field)

2) Wonders that can be researched as applications after theoretical research.
e.g. research Cure for Cancer after Advanced Genetics

3) Wonders that can be built after researching the application, which you research after a theory.
e.g. build Colossus after researching the application Colossus Planning which you get from the theory of Large Scale Structure Construction

Needlessly complex, maybe. Realistic, yes (but I'm one of these people who says realism isn't always fun so I can't use that as an argument. )

Obviously, for the type 3) Wonders, the application cost would be normal and only the build cost would be high. For the type 2) the application research cost would be very high so these types of wonders would slow down other applications. (Which kind of makes sense, really. If you've got all your scientists working on find the cure for cancer, you don't have anyone left to research the latest in smart bomb technology.)

I think it would add to the strategy though. Little choices like "Do I research the City Walls application so I can start defending my cities or research the Colossus construction so I can start building that?" Knowing that the first application will take 15 beakers and the second only 10. Also adds a bit of a race to some things as spending the time researching the construction could be a waste if somebody beats you to actually building it...

(Would be a great way for marketing to up the technology count though!)

posted 11-18-98 03:01 PM ET            

A subsequent tech (tech A) is cheaper. If you, however, switch to a different direction and research it, and then return to tech A, it will be more expensive than before - you have dissembled the research group and probably the infrastructure as well, so there is a ("real world" (TM)) reason for this.

So the deliberate switching between all techs is possible, but more expensive.

posted 11-18-98 03:50 PM ET            

As for changing research in mid-project. MOO (the original) did something whereby changing the percentages on the research scale had some algorithm to prevent that. I forget exactly what it was, but I think it was something like "when new research is added to a project, it gets added in 10% increments over 10 turns". So if you had 100 beakers in Project A and 100 in Project B, you could take 10 from B and add them to A. B would immediately drop to 90 beakers worth of research, but A would only go to 101 onthe first turn. The next turn A would have 102, then 103, etc.

It's a bit of a computational pain in the butt and is far easier to just shuffle beakers "as is".

posted 12-17-98 12:35 PM ET            

In Moo2, every advance in a given field of research (Physics, Biology, Construction, Chemistry, Sociology, etc.) led to a small improvement in stats for applications of earlier advances in the same field.

Say you researched Nuclear missiles. Later, when you'd researched Pulson missiles, the nuclear missiles suddenly cost less, didn't take up as much space, and you could build faster nuclear missiles if you wanted.

This sort of improvement continued after you'd outhgrown the tech tree, so it didn't feel so futile continuing with research.

Rather neat, is what I still think.