I am with Jbarcz1 on this. I dont see what their is to discusse if all the features of the HOI system as you describe it are going to be used, theirs realy not much wiggle room if things are set so ridgidly. It sounds to me the players only real tools are controling the number of RP their empire generates and then selecting new projects when they have sufficient RP avalible and canceling current projects. Its like a closed loop, if we wish to add any kind of adjustability for intesifying reasearch then we must crack the system open and change its inards around substantialy and in essence make a new system.
I dont recall when this HOI system was passed but I am wondering adsactly how much was desided, is it purely a desision that "all tecs will have a cost in reserach points and when the player acumulates the needed points the tec is done" or is it a full blown duplication. Basicaly how much wiggle room is their?
DESIGN: HoI Tech Tree Model
Yeah...so what exactly was passed? I think we need a reminder on the details. I thought only thing passed was organization of techs into theory, applied, and refinement. So far I'm getting, general cat, each tech cost certain amount of rp in which you have to pay for every turn for a set number of turns, can stop project, may slow but not overdrive projects.
Two things were passed:
One is the organization into theory/application. Refinement was suggested and will be discussed in the future once we know exactly what we're refining and how; it's not a part of the HoI model but it probably will be essential for us. It may be that refinement is just another application, but again, that won't be decided until we know more about how we want to refine.
The other thing that was passed is how techs get researched, or, put another way, how RP are used. They are not allocated across projects; projects have an RP cost and a fixed amount of time. For example, if my empire produces 2000 RPs, and I start a 500 RP project, I am spending 500 of my 2000 points every turn until it's done.
In this model, the only way to make projects go faster is to decrease the amount of time and/or RP required up front. You cannot over-allocate; you could research a tech (or have a racial advantage) that lowers the RP and time cost for, say, the 'Culture' category by 5%. There is no mechanism to 'invest double' and we talked briefly about a mechanism to 'invest half' if one was asked for, but as passed, you either start a project or not; you don't just allocate RPs to it.
Lastly, with regard to selecting projects: MOO3 did this easily because you were always working on the next project -- you finished Math 1 and went on to Math 2. This would be a little closer to MOO2 - you finish one and you get a popup that lets you pick the next. This is a UI issue, though, and not part of this question, so let's try to keep focused on what we're trying to get at now.
This is what was passed on the old forums. Not all of the people who were involved in passing it are around now, so it puts me in a difficult position if we set a precedent of going back and revisitng things every time we get new people and lose old ones. All I can say is, the rules were followed, and that's what people were happy with at the time.
-Aquitaine
One is the organization into theory/application. Refinement was suggested and will be discussed in the future once we know exactly what we're refining and how; it's not a part of the HoI model but it probably will be essential for us. It may be that refinement is just another application, but again, that won't be decided until we know more about how we want to refine.
The other thing that was passed is how techs get researched, or, put another way, how RP are used. They are not allocated across projects; projects have an RP cost and a fixed amount of time. For example, if my empire produces 2000 RPs, and I start a 500 RP project, I am spending 500 of my 2000 points every turn until it's done.
In this model, the only way to make projects go faster is to decrease the amount of time and/or RP required up front. You cannot over-allocate; you could research a tech (or have a racial advantage) that lowers the RP and time cost for, say, the 'Culture' category by 5%. There is no mechanism to 'invest double' and we talked briefly about a mechanism to 'invest half' if one was asked for, but as passed, you either start a project or not; you don't just allocate RPs to it.
Lastly, with regard to selecting projects: MOO3 did this easily because you were always working on the next project -- you finished Math 1 and went on to Math 2. This would be a little closer to MOO2 - you finish one and you get a popup that lets you pick the next. This is a UI issue, though, and not part of this question, so let's try to keep focused on what we're trying to get at now.
This is what was passed on the old forums. Not all of the people who were involved in passing it are around now, so it puts me in a difficult position if we set a precedent of going back and revisitng things every time we get new people and lose old ones. All I can say is, the rules were followed, and that's what people were happy with at the time.
-Aquitaine
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!
How about if you research a theory, not only do you get more research projects but you also get a price/time reduction for all projects under that cat which the theory was under.
So basically, cat organizes theory in managable chunks, but have no other gameplay effects. Each theory cost some rp and fixed amount of time and so does a project. When you finish with a theory, you get the price/time bonus that I just mentioned. Then a player can research the theory again to get more reduction and more projects to be researched under the applied section, or the player can go ahead and research projects.
So basically, cat organizes theory in managable chunks, but have no other gameplay effects. Each theory cost some rp and fixed amount of time and so does a project. When you finish with a theory, you get the price/time bonus that I just mentioned. Then a player can research the theory again to get more reduction and more projects to be researched under the applied section, or the player can go ahead and research projects.
This doesn't make sense. It seems inflexible in my view. If your empire can produce 2000 RPs a turn, why are you only able to use 500 RPs per turn on a project. It seems that there is no point in producing extra RPs (ie more research labs) because they would not even be used for projects that only use 500 RPs a turn.Aquitaine wrote: The other thing that was passed is how techs get researched, or, put another way, how RP are used. They are not allocated across projects; projects have an RP cost and a fixed amount of time. For example, if my empire produces 2000 RPs, and I start a 500 RP project, I am spending 500 of my 2000 points every turn until it's done.
In this model, the only way to make projects go faster is to decrease the amount of time and/or RP required up front. You cannot over-allocate; you could research a tech (or have a racial advantage) that lowers the RP and time cost for, say, the 'Culture' category by 5%. There is no mechanism to 'invest double' and we talked briefly about a mechanism to 'invest half' if one was asked for, but as passed, you either start a project or not; you don't just allocate RPs to it.
-Aquitaine
Well, presumably we'd balance it so that more research points will allow you to branch out and do multiple projects, which will give you benefits. You'd have to outdo others by researching broadly into the tech tree instead of deeply.utilae wrote:This doesn't make sense. It seems inflexible in my view. If your empire can produce 2000 RPs a turn, why are you only able to use 500 RPs per turn on a project. It seems that there is no point in producing extra RPs (ie more research labs) because they would not even be used for projects that only use 500 RPs a turn.
I too am worried, however, about the difficulty of balancing it . If we set a fixed cost on the projects, there will eventually be a point at which your economy is producing more RPs than it will ever need and its wasteful to produce anymore. Also, any RPs you produce that are above the sum of the cost of all available projects will be wasted, and if you have a few left over after starting a project, those are wasted as well. How does HOI balance this out?
Empire Team Lead
I was thinking that with the extra rp you accumulate in the bank can be used for cheap/speed up techs by investing into a category. Say you produce 2021 rp and you got three projects going at 1000, 500, and 500rp which adds up to 2000rp. With the extra 21 rp/turn, you save this into the bank and you can use it to invest into a cat for cheaper theories and projects. So maybe the first investment costs 100 rp; you could dump in 100rp after 5 turns from saving up for cheaper and faster projects and all prospective/future projects in that cat.
To avoid cluttering of sitrep, we could allow the player to select on a future project and the AI will automatically research all the prereq to get to the tech you want. You will less likely have excessive rp hanging around too.
Depending on how econ model works, maybe you can trade/convert 1 rp into 1 pp into imperial treasury at no penalties. If you get rp from pp in research labs, you can have the AI automatically control the conversion to match what you are doing in the research so there won't be excess rp.
Basically to get ahead in research, you first get lots of rp; you then invest in cat as much as you want so all the projects are cheap and fast; then you select the project you want. HOI method is less micro intensive compare to allocation method.
Note 1: I think Aq meant 500rp is for one example project. You can have more project going at the same time. You don't micro specifically allote rp into a project. You just click on the project you want and everything is taken care of.
Note 2: Aq meant fixed cost and time for that particular project. It doesn't necessary mean that a level 50 tech cost the same as a level 1 tech. And it's not hard to add a zero at the end of a number to make a tech cost 10x more.
Note 3: There is nothing wrong with researching techs that have reduced cost-benefits over time. Depending how the tree impacts the game, you add more strategy into the game imo, not to mention interest and fun factor. Research is important, but it is just a part of a much bigger game.
To avoid cluttering of sitrep, we could allow the player to select on a future project and the AI will automatically research all the prereq to get to the tech you want. You will less likely have excessive rp hanging around too.
Depending on how econ model works, maybe you can trade/convert 1 rp into 1 pp into imperial treasury at no penalties. If you get rp from pp in research labs, you can have the AI automatically control the conversion to match what you are doing in the research so there won't be excess rp.
Basically to get ahead in research, you first get lots of rp; you then invest in cat as much as you want so all the projects are cheap and fast; then you select the project you want. HOI method is less micro intensive compare to allocation method.
Note 1: I think Aq meant 500rp is for one example project. You can have more project going at the same time. You don't micro specifically allote rp into a project. You just click on the project you want and everything is taken care of.
Note 2: Aq meant fixed cost and time for that particular project. It doesn't necessary mean that a level 50 tech cost the same as a level 1 tech. And it's not hard to add a zero at the end of a number to make a tech cost 10x more.
Note 3: There is nothing wrong with researching techs that have reduced cost-benefits over time. Depending how the tree impacts the game, you add more strategy into the game imo, not to mention interest and fun factor. Research is important, but it is just a part of a much bigger game.
Very well put, skdiw -- that's pretty much it exactly.
I just wrote Tyreth a note on this subject so I'll cut and paste it here:
In HoI, you only produce one resource -- IC, or Industrial capacity. You then allocate your IC among consumer goods (basically food in our game, only it affects dissent rather than population), military supplies, research, and production. So you have four master sliders that control your whole industrial/research/military machine; if you have excess RP (which is rare, because their numbers are small enough that there is almost always a small project you can start) you simply transfer the IC you were using to make RP to something else. But since your IC is gradually increasing (and your tech costs gradually decreasing) this almost never happens unless you're caught off guard and have to suddenly produce a lot of tanks or something, because if you have 5 extra RP, if you wait a little while you will eventually have 15 extra RP and then you can start that 15 RP project you've been wanting.
What would take some work if we want to keep this system is adapting it to larger numbers, because we'll have thousands of RP rather than hundreds (the biggest industrial powers in HoI rarely have more than 150-200 RP total, and the smaller ones max out around 20-30).
The thing I like about this system is that it is much easier to regulate the pace of technology. People say 'oh we will just make things ridiculously expensive late-game' but this never works with MOO; the problem we have is that you have the same techs in a small galaxy, where your empire will max out producing a reasonable amount of stuff, as we do in a large galaxy, where, no matter how expensive we make our technology, somebody will eventually have so many colonies that stuff justs starts rolling in. However, on the flip side, people who are in this position may feel annoyed if we force them to wait X number of turns for a tech; it could be a little silly that it takes the same amount for the Supreme Galactic Empire to research Advanced Farming as it does for the backwards people of Tau Ceti V (or, even with modifiers, not an enormous difference between the two). On the other hand, with the traditional allocation model, if you do make things hideously expensive to keep late game interesting, you're shooting the people who like small galaxies in the foot, because it becomes very difficult for them to get the high-end stuff.
I just wrote Tyreth a note on this subject so I'll cut and paste it here:
In HoI, you only produce one resource -- IC, or Industrial capacity. You then allocate your IC among consumer goods (basically food in our game, only it affects dissent rather than population), military supplies, research, and production. So you have four master sliders that control your whole industrial/research/military machine; if you have excess RP (which is rare, because their numbers are small enough that there is almost always a small project you can start) you simply transfer the IC you were using to make RP to something else. But since your IC is gradually increasing (and your tech costs gradually decreasing) this almost never happens unless you're caught off guard and have to suddenly produce a lot of tanks or something, because if you have 5 extra RP, if you wait a little while you will eventually have 15 extra RP and then you can start that 15 RP project you've been wanting.
What would take some work if we want to keep this system is adapting it to larger numbers, because we'll have thousands of RP rather than hundreds (the biggest industrial powers in HoI rarely have more than 150-200 RP total, and the smaller ones max out around 20-30).
The thing I like about this system is that it is much easier to regulate the pace of technology. People say 'oh we will just make things ridiculously expensive late-game' but this never works with MOO; the problem we have is that you have the same techs in a small galaxy, where your empire will max out producing a reasonable amount of stuff, as we do in a large galaxy, where, no matter how expensive we make our technology, somebody will eventually have so many colonies that stuff justs starts rolling in. However, on the flip side, people who are in this position may feel annoyed if we force them to wait X number of turns for a tech; it could be a little silly that it takes the same amount for the Supreme Galactic Empire to research Advanced Farming as it does for the backwards people of Tau Ceti V (or, even with modifiers, not an enormous difference between the two). On the other hand, with the traditional allocation model, if you do make things hideously expensive to keep late game interesting, you're shooting the people who like small galaxies in the foot, because it becomes very difficult for them to get the high-end stuff.
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!
One other thing--
In HoI, at least, you can never research an application without first having its parent theory; they're always linked. Otherwise it'd be like researching the jet fighters without understanding jet propulsion.
In HoI, at least, you can never research an application without first having its parent theory; they're always linked. Otherwise it'd be like researching the jet fighters without understanding jet propulsion.
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!
-
Impaler
- Creative Contributor
- Posts: 1060
- Joined: Sun Jun 29, 2003 12:40 am
- Location: Tucson, Arizona USA
SMAC solves the small/large map problem by simply multiplying tec costs by the map size (I belive they might be using the square root of the number of tiles on the map or somthing like that). This keeps everything well balanced and time to reserach something is generaly very stable over a braod range of map sizes.
In this HOI model I would imagin the RP per turn a project consumes would be the scaled factor and the turns it takes would be constant.
In this HOI model I would imagin the RP per turn a project consumes would be the scaled factor and the turns it takes would be constant.
Fear is the Mind Killer - Frank Herbert -Dune
Hmm. That's a very interesting solution. Fits well with our second rule: steal what works. 
Just as a heads up - if I seem hesitant to sway from what we passed, it's because I don't want to set a precedent of changing things just because the people involved in the process change. We had a lot of outspoken supporters of this model. However, because the old forums aren't around and the arguments aren't available to satisfy people, I wouldn't be entirely adverse to revisiting it. Must talk more with Tyreth about this though.
At any rate, please don't mistake my defense of the HoI model for personal bias. Just trying to keep the process sane.
-Aquitaine
Just as a heads up - if I seem hesitant to sway from what we passed, it's because I don't want to set a precedent of changing things just because the people involved in the process change. We had a lot of outspoken supporters of this model. However, because the old forums aren't around and the arguments aren't available to satisfy people, I wouldn't be entirely adverse to revisiting it. Must talk more with Tyreth about this though.
At any rate, please don't mistake my defense of the HoI model for personal bias. Just trying to keep the process sane.
-Aquitaine
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!
What happened to the recursive tree? That will solve late game research problem too.
ex. code
int industry_+20% = 0;
int industry_+20%_time = 0;
industry_+20% = 1000 + industry_+20%++;
int industry_+20%_time = 5 + int industry_+20%_time++;
where industry_+20% gives you +20% in industry at the start at the cost of 1000rp for 5 turns. Then it costs 3000, 5000... you can raise it to a power if you wanted. The magic is that tech going on forever, while your empire will grow to a max no matter what size map you are playing (except infinite). The fixed time of 5 turns will also grow in the similar manner if you wanted, or not grow. It's very easily modded so you can stick a scalar factor for map size if you want.
You can use the eras to help orgainize and other ideas and still adhere to everything we passed. You can have the numbers reset after each era so you won't see 378M rp. Say like, "your researchers have dicovered dimensional labs. Now all your rp from dimensional lab is worth 1M of the old rp." So now you generate 378 rp. The game is mostly relative really.
I think recurrsive will make a lot of hardcore research fanatics happy. And it's better than some points for researching future techs like in civ because recurrsive tech does something gameplay-wise.
So far, I don't think we have sway away from passed features. I just didn't know we passed fixed-time thing so ppl assume a rp project allocation system. I like what HOI system has to offer.
ex. code
int industry_+20% = 0;
int industry_+20%_time = 0;
industry_+20% = 1000 + industry_+20%++;
int industry_+20%_time = 5 + int industry_+20%_time++;
where industry_+20% gives you +20% in industry at the start at the cost of 1000rp for 5 turns. Then it costs 3000, 5000... you can raise it to a power if you wanted. The magic is that tech going on forever, while your empire will grow to a max no matter what size map you are playing (except infinite). The fixed time of 5 turns will also grow in the similar manner if you wanted, or not grow. It's very easily modded so you can stick a scalar factor for map size if you want.
You can use the eras to help orgainize and other ideas and still adhere to everything we passed. You can have the numbers reset after each era so you won't see 378M rp. Say like, "your researchers have dicovered dimensional labs. Now all your rp from dimensional lab is worth 1M of the old rp." So now you generate 378 rp. The game is mostly relative really.
I think recurrsive will make a lot of hardcore research fanatics happy. And it's better than some points for researching future techs like in civ because recurrsive tech does something gameplay-wise.
So far, I don't think we have sway away from passed features. I just didn't know we passed fixed-time thing so ppl assume a rp project allocation system. I like what HOI system has to offer.
IMO the important thing is to define the tech costs using a simple yet flexible formula. This is one of the things moo3 got right, IIRC the cost of a given tech was along the lines of:
tech cost = (tech_level * scaling_factor) ^ cost_exponent
Where tech level is a simple representation of the 'difficulty' of the tech (in moo3's case an integer from 1 to 50). Changing scaling_factor allows the overall game tech speed to be changed easily, and cost_exponent can be used to make late game techs even more expensive, if required (by using a value greater than 1.0).
Note that this model doesnt tie you in to a using progressive tree like Moo3's, but it allows you to easily set relative costs for every tech in the game, without worrying about specific numbers until you need to balance things. i.e. you can adjust the cost of techs relative to each other, and the overall tech pace of the game independently of each other. Very very very useful. The point is that youre going to have to balance tech costs empirically at some point, so you may as well make it as easy as possible for yourselves. With the Moo3 approach you can change the entire tech tree costs by changing only two values.
w.r.t. balancing for different sized galaxies, it would be quite easy to tie galaxy size into this model:
tech_cost = (tech_level * scaling_factor * (number_of_stars ^ size_exponent)) ^ cost_exponent
Where size_exponent would typically be 0.5 to give a square root, but could be changed to make galaxy size have more or less of an impact on tech costs (a value of zero would make it irrelevant).
Another thing thats important is to have different tech speeds available to the player - this is something that SMAC got right (there was a slow tech option) and Moo3 got very, very wrong, especially since the tech speed changed randomly with every patch
. Some people like fast tech games, some like slow tech - give them the option.
tech cost = (tech_level * scaling_factor) ^ cost_exponent
Where tech level is a simple representation of the 'difficulty' of the tech (in moo3's case an integer from 1 to 50). Changing scaling_factor allows the overall game tech speed to be changed easily, and cost_exponent can be used to make late game techs even more expensive, if required (by using a value greater than 1.0).
Note that this model doesnt tie you in to a using progressive tree like Moo3's, but it allows you to easily set relative costs for every tech in the game, without worrying about specific numbers until you need to balance things. i.e. you can adjust the cost of techs relative to each other, and the overall tech pace of the game independently of each other. Very very very useful. The point is that youre going to have to balance tech costs empirically at some point, so you may as well make it as easy as possible for yourselves. With the Moo3 approach you can change the entire tech tree costs by changing only two values.
w.r.t. balancing for different sized galaxies, it would be quite easy to tie galaxy size into this model:
tech_cost = (tech_level * scaling_factor * (number_of_stars ^ size_exponent)) ^ cost_exponent
Where size_exponent would typically be 0.5 to give a square root, but could be changed to make galaxy size have more or less of an impact on tech costs (a value of zero would make it irrelevant).
Another thing thats important is to have different tech speeds available to the player - this is something that SMAC got right (there was a slow tech option) and Moo3 got very, very wrong, especially since the tech speed changed randomly with every patch
. Some people like fast tech games, some like slow tech - give them the option.I thought one of the primary differences of playing a small vs. large is a different flavor in techs and not just the same thing but more micro with a large map if you make it all relative, which is why I didn't strongly suggest a scalar factor for the map size in the first place. If you guys want the same playing experience with a small as large, that's fine too. I could go either way. But that's just me and a small note on the side.
Back to my main point (not cost of techs), what about recurrsive techs? Do we want a dead stop after tech 50? Or should we have some basic techs like pp production, food, min go on? I think recurrsion techs would make a better "soft" late research game rather than a dead stop. I think recurrsion adds another dimension to the late-game.
What I am trying to get at is to make the research tree as interesting and as dynamic as possible with other parts of game at all points in the game while making everything simple. Where as in moo3, you get massive amounts of rp anyway and the only decision is to chose what techs to research, compared to HOI or what I imagine, we take one step further and make the whole tech tree a macro consideration. In HOI, as I understand it, since rp = pp because of IC, you have to make the decision of whether to use all your IC for mobilization for a effective war, which in that case, your long-term growth suffers.
Recursive techs doesn't deal with our econ-research model, but it's a simple solution for a better late game, which is my purpose. It doesn't force players to start throwing nova-bombs becuase they already got all the uber techs.
We could also throw in a random cost factor at predetermined prior to the start of a game, a random cost factor that fudges the effectiveness of the tech tree in respect with the rest of game that will enhance the character of each new game.
Small correction: you can do tech slow down/up in moo3 for a branch. Trickier if you want to do individual techs.
Back to my main point (not cost of techs), what about recurrsive techs? Do we want a dead stop after tech 50? Or should we have some basic techs like pp production, food, min go on? I think recurrsion techs would make a better "soft" late research game rather than a dead stop. I think recurrsion adds another dimension to the late-game.
What I am trying to get at is to make the research tree as interesting and as dynamic as possible with other parts of game at all points in the game while making everything simple. Where as in moo3, you get massive amounts of rp anyway and the only decision is to chose what techs to research, compared to HOI or what I imagine, we take one step further and make the whole tech tree a macro consideration. In HOI, as I understand it, since rp = pp because of IC, you have to make the decision of whether to use all your IC for mobilization for a effective war, which in that case, your long-term growth suffers.
Recursive techs doesn't deal with our econ-research model, but it's a simple solution for a better late game, which is my purpose. It doesn't force players to start throwing nova-bombs becuase they already got all the uber techs.
We could also throw in a random cost factor at predetermined prior to the start of a game, a random cost factor that fudges the effectiveness of the tech tree in respect with the rest of game that will enhance the character of each new game.
Small correction: you can do tech slow down/up in moo3 for a branch. Trickier if you want to do individual techs.