Maniac
Apolyton Sage
I’d like to present a few guidelines I use for designing a technology tree. I haven’t been able to design a tech tree myself which follows these guidelines a full 100%. Still I think it’s good to keep these in mind and try to approach the ideal as close as possible. Here goes. 
Where to start when designing a tech tree from scratch? Personally I find it useful to identify a few themes around which the mod and tech tree revolves. For a sci fi mod this could be: genetics, computers, physics, terraforming, etc… For an ancient age mod this could be: mounted, metal working, religion, construction…
Your first draft tech tree could then look like a few pillars of: Genetics I, Genetics II, Genetics III & Computers I, Computers II, and so on. You probably don’t want your tech tree to look like that forever, but it’s a useful beginning. It helps you think of possible tech contents, and assign them a place in the tech tree.
How to assign these tech contents is the main focus of this article. The basic principle here is “one choice = no choice”. For a good tech tree the player has to be presented with multiple interesting choices on how to develop his civilization. You want to avoid there being one tech path which is best to follow in all games played. This doesn’t have to mean that all tech b-lines have to be equally attractive in each game and with each civ played. It’s fun to have civs that have synergies with one or a few specific tech paths. But over the course of several games you don’t want a part of your tech tree to remain repeatedly un(der)used, or techs being researched only because they’re a prerequisite of other techs.
To answer the question how to achieve this, I’ll divide all possible tech contents in four categories:
Universal benefits
Specific benefits
Diminishing benefits
First researcher benefits
Universal benefits are tech contents which are always useful for everyone, no matter what civ you play or strategy you follow.
Specific benefits are tech contents only useful for a certain strategy or under certain situations. The difference between universal and specific benefits isn’t black and white. Rather they’re the extremes with between them a continuum full of greys.
Then there are tech contents which have a diminishing value the more you get of the same kind. The first one of its kind is usually very valuable; thereafter the value steeply decreases.
First researcher benefits are one-time boni for the first one to get this tech. Everyone else has no use whatsoever of them.
Now my thesis is that to have a good tech tree, each tech should provide a universal benefit, or a mix of specific benefits, so that it’s likely at least one of them will be of use to you. This counts even if your aim is to have a tech tree encouraging specialization. It’s ok for this benefit to be a very minor one, but there should definitely be one. This assures it’s always better to have a tech than not having the tech.
This has two benefits.
1) It improves the value and fun of diplomacy and tech trading. No longer will there be this bunch of techs which the AI wants to trade to you, but have no value whatsoever for you.
2) It improves AI performance (and also avoids noob traps). No matter how well you try to make the AI, it will never be as good in judging the value of a tech as a human. So by giving each tech a universal benefit, you basically make your tech tree dummy-proof. While it won’t excel, even a random number generator like the AI will be able to research/trade techs without shooting itself in the foot.
With these four tech content categories identified, let’s look at some more concrete examples to make things clear.
Universal benefits
Buildings
These are the biggest and easiest source of universal benefits. Of course they aren’t always useful (eg an Aqueduct isn’t immediately useful if you have lots of health resources), but in the continuum of universal <-> specific, they definitely tend heavily towards the universal side. I’m postulating here of course that the building benefits are designed reasonably well. I’m obviously not going to research a building which costs 200 hammers and gives only one gold and one unhappy face or so.
Unfortunately you can’t use buildings alone to cover your need of universal tech contents. Let’s assume the goal is a tech tree of around 80 techs – more or less the length of all the default civ tech trees. You won’t get much further than 40 buildings (also more or less the average of civ games) without getting repetitive. And adding too many buildings will make it impossible to build them all anyway.
Thus you have to invent some 40 other universal benefits. That’s IMO the hardest part of tech tree design and why I’ve never been able to create a tech tree which fully follows my own guidelines.
Since it’s so hard to think of good universal benefits, I’d suggest as a rule of thumb not to give more than one building to a tech. Better to use that idea for a building for another tech, if you want to reach a good tech tree of decent size.
Terrain improvements
These are another source of universal benefits, at least to a certain extent. Farms, mines and cottages all stay decently valuable because they’re the first source of food, hammers and commerce. But more than three improvements is destined to overlap with the uses of these already available ones. So additional terrain improvements should be categorized as tech contents with diminishing value, having only some specific uses. Windmills, watermills and workshops are less valuable when you already have farms, mines and cottages and aren’t following some specific strategy.
Technologies boosting the yield of a terrain improvement are a very good universal benefit. But you can’t give more than one yield boost per terrain improvement without becoming repetitive.
Bonus resources & their enabling improvements
In the beginning of the game you only have access to a couple resources, but as the game progresses and you expand, you usually gain access to a resource of every kind. So de facto the terrain improvements which connect resources are good to give as a universal benefit to a tech. Likewise with techs that reveal resources.
Here too you can let technologies boost the yields of bonus-connecting improvements. Of course this won’t result in a large economic boost. But remember: the idea behind a universal benefit is just to make having the tech always better than not having the tech. This small benefit might make the difference between wanting or not wanting to trade or research that tech.
Promotions
Since (in most mods) these remain useful throughout the whole game and can be applied to different units, they are a universal benefit, in any case more universal and longer lasting than the units they’re used on themselves.
However note that promotions ensure that it’s always better to have a tech rather than not have it, they’re unlikely by themselves to make a tech worth researching. Thus promotions as a tech content needs to be coupled with some other (probably more specific) tech contents, so that at least some civs will want to research the enabling tech themselves instead of trade it from someone else.
Ability to research other techs
A tech can be valuable if it is an absolute requirement (AND-prerequisite) for other, more interesting, techs, even if the tech itself doesn’t do anything per se. This universal benefit is a big reason why the default Civ4 tech tree doesn’t suck.
However one could raise the question, if the only use of a tech is that it leads to another tech, what's the point of making it two separate techs? I can give the answer myself:
because sometimes you're in a strategy where you want one or more of those specific or first researcher benefits, but you don't want to spend the extra time researching what it leads to in order to get them.
A tech tree where many techs' only universal benefit is that it is required for other techs, isn't a bad tech tree. Still I personally feel the tech tree would be more interesting and the techs in question more fun to research if they also always directly provided a universal benefit. For instance while I rationally know Paper leads to other important techs, I never feel particularly excited when I get it. Rather disappointed or bored in fact. And for a good civ game I guess you want a player to feel happy with every new achievement made, to encourage "one more turn!"
Other universal benefits
I’ll list here some more examples of universal benefits.
Road speed boosts
Worker efficiency boost
Extra trade routes
Extra health & happiness
…
Now let’s look at some examples of tech content more to the “specific” side of the universal <-> specific benefit spectrum:
Specific & Diminishing benefits
Civics
Civics are meant to boost a certain strategy the player wishes to follow. Therefore civics are per definition only of specific/situational use, and cannot be used as the universal or primary benefit of a tech.
Furthermore if the civics are well designed, they should all be somewhat useful, some more than others for certain strategies, but all better than the default civic. This means civics also have a diminishing value. The more civics you already have of a certain category, the less valuable an additional one becomes. So a player is unlikely to want to spend a lot of research points just to get this civic. Unless the civic is very powerful, but then you’ve just created an unbalanced civic system. Not fun either if everyone ends up using the same civic at the end of the game.
So two tips for civics are:
1) definitely include some universal benefit with the enabling techs. Even more so if the civic is enabled by a later-game tech.
2) try to have all civics of the same category be enabled around the same tech level. So that each civic can be the first researched depending on the b-line followed.
Units
Ah, units… These are the prime example of diminishing value tech contents, and as a consequence a big reason why parts of tech trees remain un(der)used.
Let’s look at a Rock/Paper/Scissors inspired combat system. A > B > C > D beats A… The thing is, you don’t need all of these to be successful in combat. While C would be the best unit to beat D, using B or D would do the trick as well. You could be successful in combat by just researching two of these four. Then you could direct your research instead to get some additional economic advantages, or get the next best unit of a certain type earlier: A2 instead of A1, and easily beat the crap out of A, B, C and D.
Other example is a distinction between offensive and defensive units, like archers are in most mods. While archer type units are the best defensive units, offensive units can defend sufficiently as well. So again a player only researching the offensive units will do better than a player researching both. There is a diminishing value in having all options.
As a consequence what military units are researched and used can depend not on the military units themselves, but on what other benefits the enabling tech gives. If you have four possible military research paths, and you only need one, which one will you research? The one which in addition also gives some economic benefit, or the one which only gives a military unit?
Thus in order to ensure that each tech path is used at least some times, and that it can be valuable to research more than one path, each tech which grants a military unit should also grant a somewhat universal benefit. Then you could for example trade that tech from someone to get the universal benefit, but as an added bonus also starting using that military unit, something which you would otherwise never have done.
Diplomacy deals
With these I mean tech trading, gold trading, open borders trading… Only one of the diplomatic parties needs to have the tech enabling these diplomatic deals. As a consequence they have a diminishing value the more civs can do these deals. The first person who can trade techs can make some golden deals. For the last civ enabling tech trading is worthless.
First researcher benefits
Examples of these are wonders of the world, a free great person for the first researcher, founding religions, corporations… When someone beats you to it, these tech contents are worth nothing to you anymore. Yet I still see mods where people reason “Oh, this tech isn’t valuable enough – let’s add some free great person to it.” Biggest mistake possible of course. When designing a tech tree, all techs should still be valuable even if you remove all these first researcher benefits.
Another tip regarding first researcher benefits is: don’t give more than one per tech. One per tech should suffice as a reward for following a certain tech path. If you have two cool ideas for first researcher rewards, why not place them at two separate techs? Then you just increased by one the number of possible b-lines.
The extreme example to illustrate this is the Divine Right tech in default Civ4. It contains nothing but one-time boni. Two wonders and a religion. Completely useless to get from a civ later in the game. Even Nationalism can be obtained through another tech.
This concludes my write-up about tech tree design. Hopefully it is of some use to Civ4 modders all around!
Where to start when designing a tech tree from scratch? Personally I find it useful to identify a few themes around which the mod and tech tree revolves. For a sci fi mod this could be: genetics, computers, physics, terraforming, etc… For an ancient age mod this could be: mounted, metal working, religion, construction…
Your first draft tech tree could then look like a few pillars of: Genetics I, Genetics II, Genetics III & Computers I, Computers II, and so on. You probably don’t want your tech tree to look like that forever, but it’s a useful beginning. It helps you think of possible tech contents, and assign them a place in the tech tree.
How to assign these tech contents is the main focus of this article. The basic principle here is “one choice = no choice”. For a good tech tree the player has to be presented with multiple interesting choices on how to develop his civilization. You want to avoid there being one tech path which is best to follow in all games played. This doesn’t have to mean that all tech b-lines have to be equally attractive in each game and with each civ played. It’s fun to have civs that have synergies with one or a few specific tech paths. But over the course of several games you don’t want a part of your tech tree to remain repeatedly un(der)used, or techs being researched only because they’re a prerequisite of other techs.
To answer the question how to achieve this, I’ll divide all possible tech contents in four categories:
Universal benefits
Specific benefits
Diminishing benefits
First researcher benefits
Universal benefits are tech contents which are always useful for everyone, no matter what civ you play or strategy you follow.
Specific benefits are tech contents only useful for a certain strategy or under certain situations. The difference between universal and specific benefits isn’t black and white. Rather they’re the extremes with between them a continuum full of greys.
Then there are tech contents which have a diminishing value the more you get of the same kind. The first one of its kind is usually very valuable; thereafter the value steeply decreases.
First researcher benefits are one-time boni for the first one to get this tech. Everyone else has no use whatsoever of them.
Now my thesis is that to have a good tech tree, each tech should provide a universal benefit, or a mix of specific benefits, so that it’s likely at least one of them will be of use to you. This counts even if your aim is to have a tech tree encouraging specialization. It’s ok for this benefit to be a very minor one, but there should definitely be one. This assures it’s always better to have a tech than not having the tech.
This has two benefits.
1) It improves the value and fun of diplomacy and tech trading. No longer will there be this bunch of techs which the AI wants to trade to you, but have no value whatsoever for you.
2) It improves AI performance (and also avoids noob traps). No matter how well you try to make the AI, it will never be as good in judging the value of a tech as a human. So by giving each tech a universal benefit, you basically make your tech tree dummy-proof. While it won’t excel, even a random number generator like the AI will be able to research/trade techs without shooting itself in the foot.
With these four tech content categories identified, let’s look at some more concrete examples to make things clear.
Universal benefits
Buildings
These are the biggest and easiest source of universal benefits. Of course they aren’t always useful (eg an Aqueduct isn’t immediately useful if you have lots of health resources), but in the continuum of universal <-> specific, they definitely tend heavily towards the universal side. I’m postulating here of course that the building benefits are designed reasonably well. I’m obviously not going to research a building which costs 200 hammers and gives only one gold and one unhappy face or so.
Unfortunately you can’t use buildings alone to cover your need of universal tech contents. Let’s assume the goal is a tech tree of around 80 techs – more or less the length of all the default civ tech trees. You won’t get much further than 40 buildings (also more or less the average of civ games) without getting repetitive. And adding too many buildings will make it impossible to build them all anyway.
Thus you have to invent some 40 other universal benefits. That’s IMO the hardest part of tech tree design and why I’ve never been able to create a tech tree which fully follows my own guidelines.
Since it’s so hard to think of good universal benefits, I’d suggest as a rule of thumb not to give more than one building to a tech. Better to use that idea for a building for another tech, if you want to reach a good tech tree of decent size.
Terrain improvements
These are another source of universal benefits, at least to a certain extent. Farms, mines and cottages all stay decently valuable because they’re the first source of food, hammers and commerce. But more than three improvements is destined to overlap with the uses of these already available ones. So additional terrain improvements should be categorized as tech contents with diminishing value, having only some specific uses. Windmills, watermills and workshops are less valuable when you already have farms, mines and cottages and aren’t following some specific strategy.
Technologies boosting the yield of a terrain improvement are a very good universal benefit. But you can’t give more than one yield boost per terrain improvement without becoming repetitive.
Bonus resources & their enabling improvements
In the beginning of the game you only have access to a couple resources, but as the game progresses and you expand, you usually gain access to a resource of every kind. So de facto the terrain improvements which connect resources are good to give as a universal benefit to a tech. Likewise with techs that reveal resources.
Here too you can let technologies boost the yields of bonus-connecting improvements. Of course this won’t result in a large economic boost. But remember: the idea behind a universal benefit is just to make having the tech always better than not having the tech. This small benefit might make the difference between wanting or not wanting to trade or research that tech.
Promotions
Since (in most mods) these remain useful throughout the whole game and can be applied to different units, they are a universal benefit, in any case more universal and longer lasting than the units they’re used on themselves.
However note that promotions ensure that it’s always better to have a tech rather than not have it, they’re unlikely by themselves to make a tech worth researching. Thus promotions as a tech content needs to be coupled with some other (probably more specific) tech contents, so that at least some civs will want to research the enabling tech themselves instead of trade it from someone else.
Ability to research other techs
A tech can be valuable if it is an absolute requirement (AND-prerequisite) for other, more interesting, techs, even if the tech itself doesn’t do anything per se. This universal benefit is a big reason why the default Civ4 tech tree doesn’t suck.
However one could raise the question, if the only use of a tech is that it leads to another tech, what's the point of making it two separate techs? I can give the answer myself:
because sometimes you're in a strategy where you want one or more of those specific or first researcher benefits, but you don't want to spend the extra time researching what it leads to in order to get them.
A tech tree where many techs' only universal benefit is that it is required for other techs, isn't a bad tech tree. Still I personally feel the tech tree would be more interesting and the techs in question more fun to research if they also always directly provided a universal benefit. For instance while I rationally know Paper leads to other important techs, I never feel particularly excited when I get it. Rather disappointed or bored in fact. And for a good civ game I guess you want a player to feel happy with every new achievement made, to encourage "one more turn!"
Other universal benefits
I’ll list here some more examples of universal benefits.
Road speed boosts
Worker efficiency boost
Extra trade routes
Extra health & happiness
…
Now let’s look at some examples of tech content more to the “specific” side of the universal <-> specific benefit spectrum:
Specific & Diminishing benefits
Civics
Civics are meant to boost a certain strategy the player wishes to follow. Therefore civics are per definition only of specific/situational use, and cannot be used as the universal or primary benefit of a tech.
Furthermore if the civics are well designed, they should all be somewhat useful, some more than others for certain strategies, but all better than the default civic. This means civics also have a diminishing value. The more civics you already have of a certain category, the less valuable an additional one becomes. So a player is unlikely to want to spend a lot of research points just to get this civic. Unless the civic is very powerful, but then you’ve just created an unbalanced civic system. Not fun either if everyone ends up using the same civic at the end of the game.
So two tips for civics are:
1) definitely include some universal benefit with the enabling techs. Even more so if the civic is enabled by a later-game tech.
2) try to have all civics of the same category be enabled around the same tech level. So that each civic can be the first researched depending on the b-line followed.
Units
Ah, units… These are the prime example of diminishing value tech contents, and as a consequence a big reason why parts of tech trees remain un(der)used.
Let’s look at a Rock/Paper/Scissors inspired combat system. A > B > C > D beats A… The thing is, you don’t need all of these to be successful in combat. While C would be the best unit to beat D, using B or D would do the trick as well. You could be successful in combat by just researching two of these four. Then you could direct your research instead to get some additional economic advantages, or get the next best unit of a certain type earlier: A2 instead of A1, and easily beat the crap out of A, B, C and D.
Other example is a distinction between offensive and defensive units, like archers are in most mods. While archer type units are the best defensive units, offensive units can defend sufficiently as well. So again a player only researching the offensive units will do better than a player researching both. There is a diminishing value in having all options.
As a consequence what military units are researched and used can depend not on the military units themselves, but on what other benefits the enabling tech gives. If you have four possible military research paths, and you only need one, which one will you research? The one which in addition also gives some economic benefit, or the one which only gives a military unit?
Thus in order to ensure that each tech path is used at least some times, and that it can be valuable to research more than one path, each tech which grants a military unit should also grant a somewhat universal benefit. Then you could for example trade that tech from someone to get the universal benefit, but as an added bonus also starting using that military unit, something which you would otherwise never have done.
Diplomacy deals
With these I mean tech trading, gold trading, open borders trading… Only one of the diplomatic parties needs to have the tech enabling these diplomatic deals. As a consequence they have a diminishing value the more civs can do these deals. The first person who can trade techs can make some golden deals. For the last civ enabling tech trading is worthless.
First researcher benefits
Examples of these are wonders of the world, a free great person for the first researcher, founding religions, corporations… When someone beats you to it, these tech contents are worth nothing to you anymore. Yet I still see mods where people reason “Oh, this tech isn’t valuable enough – let’s add some free great person to it.” Biggest mistake possible of course. When designing a tech tree, all techs should still be valuable even if you remove all these first researcher benefits.
Another tip regarding first researcher benefits is: don’t give more than one per tech. One per tech should suffice as a reward for following a certain tech path. If you have two cool ideas for first researcher rewards, why not place them at two separate techs? Then you just increased by one the number of possible b-lines.
The extreme example to illustrate this is the Divine Right tech in default Civ4. It contains nothing but one-time boni. Two wonders and a religion. Completely useless to get from a civ later in the game. Even Nationalism can be obtained through another tech.
This concludes my write-up about tech tree design. Hopefully it is of some use to Civ4 modders all around!
