Galaxy Size and Tech Costs

For what's not in 'Top Priority Game Design'. Post your ideas, visions, suggestions for the game, rules, modifications, etc.

Moderator: Oberlus

Message
Author
User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#46 Post by Geoff the Medio »

Ablaze wrote:I’m still convinced that no equation can successfully address all the issues. If 15 people join a multiplayer game and then 10 of those people drop out before doing anything this model will break, and no amount of tweaking the equation will fix that.
That's a problem with your equation, not necessarily all equations. And we can't expect to correct for all possible player oddities... Are we also going to worry that the tech rate will be too low if no players do any research? I think we have to assume that if a game starts, most of the players will stay in the game, and those in the game will act "normally" on average...

If we really wanted, I suppose we could attempt to correct for dropped players by adjusting the threshold tech cost at which the increased cost factor levels off, to account for the actual number of players in the game. I'm rather reluctant to do this though, as it starts to get back into changing tech costs, or future tech costs at least, according to the gamestate, which will be extra confusing and problematic...

How to deal with dropped players is really a bigger issue than just trying to keep the tech rate balanced though... I suspect that whatever solution we used to fix that problem as a whole will correct the tech rate balance as well. This might involve replacing dropped players with AIs... (or finding replacement players if possible...)
The level based system, however, would work fine.
It occurs to me that the threshold for tech cost increase of the equation system is somewhat like a "level", except that it doesn't depend on the events system to be implemented... And the threshold value can be calculated ahead of time, as it doesn't depend on the gamestate, so we can display the whole tech tree with costs for techs from the start of the game, rather than having to worry about how to indicate to the player when they've entered this or that phase of tech cost adjustments. If they want to know how much a tech costs, they'll look at the tree... in that case, it doesn't matter to the player whether they're in the pre or post threshold phase. (This isn't like hiding a formula from the player... everything they need to know is visible from the costs of the techs from the start of the game)

Ablaze
Creative Contributor
Posts: 314
Joined: Tue Aug 26, 2003 6:10 pm
Location: Amidst the Inferno.

#47 Post by Ablaze »

Well, I guess this is another case where I'll implement one system, you implement another, and maybe we'll see how it turns out in the end.
Time flies like the wind, fruit flies like bananas.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#48 Post by Geoff the Medio »

Ablaze wrote:Well, I guess this is another case where I'll implement one system, you implement another, and maybe we'll see how it turns out in the end.
Not necessarily... I'm not the one who makes the decisions for what does or doesn't get implemented.

Daveybaby
Small Juggernaut
Posts: 724
Joined: Mon Sep 01, 2003 11:07 am
Location: Hastings, UK

#49 Post by Daveybaby »

I am probably completely missing the point here, but in what way is this problem not solved by having later techs become exponentially more expensive than early game techs?

For each galaxy size / number of players you define a scaling factor and an exponent factor

Each tech then costs : (scalar * tech_level) ^ exponent
Actually maybe (scalar * (tech_level ^ exponent)) would be easier to manage.

Head full of snot this week so excuse me for being slow, but doesnt this solve the problem? It did in Moo3, insofar as it allowed players to mod both the scalar and the exponent and tweak to get any of the following:

:arrow: rapid tech progression at the beginning of the game, completely stagnant at the end.
:arrow: slow tech progression at the beginning, rapid at the end.
:arrow: fairly constant tech progression throughout, at any rate chosen.

All of the above could be achieved by trial and error for any galaxy size. The number of opponents is another factor, since the number at the beginning does not tend to equal the number at the end, but this is easily factored in, and once we have more of a feel for how a typical game plays out, a pretty robust formula could be developed to determine the scalar and exponent factors.

Plus, IMO, there should probably be a 'tech speed' option at game creation (which would only affect the scalar), but that's neither here nor there.
The COW Project : You have a spy in your midst.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#50 Post by Geoff the Medio »

Daveybaby,

1) When the galaxy is full, the rate of RP production for an average empire at a given turn is roughly proportional to galaxy_size / #players. This represents the area of the average empire.

2) When the game starts and until the galaxy fills up, RP production does not depend on galaxy size. When the galaxy fills up, (1) applies.

In both (1) and (2), there are other factors that determine RP production (eg. empire development, rather than expansion), but these are approximated by me as relatively constant between galaxy sizes and numbers of players. This includes tech trading between empires, which is probably a weak assumption.

Your suggestion is could be used to calculate the tech costs for a given level. However, IMO we're better having a "base tech cost" for each tech which can be manually entered and tweaked. This way we're not constrained to the "tech levels" system if we don't want to be... we can still use a power formula such as you suggest to generate the base tech costs that we enter manually if we want to, but we shouldn't limit ourselves to do this (and the game isn't currently set up to do this anyway... it expects manually entered tech costs for each tech).

The main problem with using an power function such as your suggestion is that it doesn't take into account the changing effect of galaxy size as empires grow. Rather than just "scalar * tech_level ^ exponent" or "base_tech_cost ^ exponent", we need a more complicated function of base_tech_cost and galaxy size and (initial) number of players to "correctly" model the transition between growth unrestricted by galaxy size, and full-galaxy empire development only.

If we don't correct model the transition, say by just having adjusted_tech_cost = base_tech_cost * some_scale_factor * galaxy_size or somesuch, we'll end up in situations where at the start of the game, the tech costs are increased significantly before the galaxy size starts to have an effect on RP production. This might mean that at the start of the game in a big galaxy, you couldn't even research one tech until you increase your RP production significantly, but in a small galaxy, you could research 5 or 10 techs at a time without developing at all. This is undesirable, to me.

IMO it's not possible to have just a single exponent value work "correctly" for the whole game for all galaxy sizes, even if the exponent is a function of galaxy size itself (or chosen by trial and error). The exponential curve might sort of fit the RP growth curve for one galaxy size, but if you change the galaxy size, the RP growth curve will change differently than the exponential curve will, making the fit not so good.

Of course this is all theoretical now... but the idea is to get a good set of possible forms of the modifications that might work, so as to have them ready when it becomes possible to use trial and error to pick the formula and constants that work best for all possible situations.

Just about everyone agrees that there should be a player-controlled "tech speed" option at the start of the game. That's not the issue, as you said... sort of.

Daveybaby
Small Juggernaut
Posts: 724
Joined: Mon Sep 01, 2003 11:07 am
Location: Hastings, UK

#51 Post by Daveybaby »

Agree that there will always be some situations where the rate of tech progression is non-ideal, but to be honest i dont think there is any method which would reliably achieve this, because the growth patterns of empires will vary even given the same starting conditions (well, hopefully they will if the game is designed right, it would be awfully boring to have every game play out in the same way). Also, the way empires are predisposed to each other can affect the rate of tech, in one game players may trade a lot of techs, or steal them from each other, or form lots of research treaties, etc, which would drastically accelerate progression through the tech tree. In another game players may do less of this.

Another thing to bear in mind is that there are other things that influence RP generation than just the amount of real estate you control, e.g. techs which increase your research ouput per person, or techs which increase food production, thus freeing up more people for research, terraforming techs, etc. Careful placing of these techs in the tech tree can be used to try to smooth the transition from when the rapid territorial expansion phase ends.
The COW Project : You have a spy in your midst.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#52 Post by Geoff the Medio »

Daveybaby wrote:...i dont think there is any method which would reliably achieve this, because the growth patterns of empires will vary even given the same starting conditions...
...the way empires are predisposed to each other...
...trade a lot of techs, or steal them from each other, or form lots of research treaties...
Yes, I've thought of these issues... but I'm not trying to make the tech progression exactly the same in each game, so much as I'm trying to prevent galaxy size from being an overwhelming determinant of tech progress. Put another way, I'd like to keep the tech rate consistent between galaxy sizes, on average, assume all else is equal, for the whole game (including during initial expansion when galaxy size shouldn't matter, and end-game when it's very significant).
...other things that influence RP generation ... increase your research ouput ... increase food production... Careful placing of these techs in the tech tree can be used to try to smooth the transition from when the rapid territorial expansion phase ends.
If the transition at the end of rapid expansion happens at a significant different time (turn) in different galaxy sizes, even carefully placed techs won't always arrive at the correct turn... which is essentially the main measure of the problem...

Ablaze
Creative Contributor
Posts: 314
Joined: Tue Aug 26, 2003 6:10 pm
Location: Amidst the Inferno.

#53 Post by Ablaze »

I like the level based system because it solves that very thing that you are talking about, dave.

It works like this: You add two events to the game, both of which decrease the cost of every tech in the game for everyone in the game. Event one can be triggered when 80% of the systems in the galaxy have been colonized and event two can be triggered when 30% of the inhabited planets have changed hands in some way.

The beauty of this system is that there is only one variable that needs to be balanced; the strength of the cost reduction. Everything else will automatically adjust to the changing conditions in the game. "Changing conditions" does not just include people who play differently, but also other balance issues.

For instance if it is decided that, in order to make things even, people who explore a system first should get a prize (like MOO2) then an equation based system would have to be readjusted since people would tend to get through early game technologies faster. The level based system, on the other hand, would require no adjustment.
Time flies like the wind, fruit flies like bananas.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#54 Post by Geoff the Medio »

Ablaze, perhaps you should use a different name than "level based system" for what you're proposing... such as "events-based system". Unless I'm missing something, your "levels" have nothing to do with the idea of tech levels, such as daveybaby suggested.

Also, I'm not seeing how your proposed system is going to solve the problem... The game turn of the 80% (and probably the 30%) cost reduction events are going to depend on the size of the galaxy quite significantly... Unless you have an equation such as I suggest in addition to these events, isn't the same problem going to occur before the event happens? And in smaller galaxies, the events will happen sooner, giving them a boost to research, but significantly larger galaxies will keep growing for much longer than smaller galaxies, to a degree that even if the cost reductions are 50% or 90%, some galaxies will be big enough that RP production will still outpace that of smaller galaxies by equivalent factors (2 or 10) after enough continued growth, so the same problem remains...
The beauty of this system is that there is only one variable that needs to be balanced; the strength of the cost reduction. Everything else will automatically adjust to the changing conditions in the game. "Changing conditions" does not just include people who play differently, but also other balance issues.
Well, if you picked a sufficiently simple equation, you'd only have one parameter that needs balancing as well... the resulting equation wouldn't be very good, but see above for probably equivalent issues for a system with only two events... (more than two events might help, which, if you keep adding events, ends up being a discrete event approximation of a continuous function such as I propose) And how will the cost reductions be determined? Will they be the same for all galaxies, or dependent on galaxy size or some other such parameter using an equation? And isn't the exact timing (80%, 30%) itself a parameter that needs balancing?

For instance if it is decided that, in order to make things even, people who explore a system first should get a prize (like MOO2) then an equation based system would have to be readjusted since people would tend to get through early game technologies faster. The level based system, on the other hand, would require no adjustment.
How does cost reductions for techs at 80% colonization and 30% system exchange rebalance the effects of variable tech trading? I'm not seeing the connection here...

User avatar
skdiw
Creative Contributor
Posts: 643
Joined: Mon Sep 01, 2003 2:17 am

#55 Post by skdiw »

I suppose you could say that once the galaxy is 80% filled, the "Galactic Senate Union" (GSU) forms. We can fix the rp growth pre-GSU at one rate than post-GSU at another. As default, we can just assume some formula dependent on the #players and galaxy size chosen from the initial setup. The GSU can vote to change various game parameters such as the rate of research. Every 10 turns, the GSU comes again and you can vote for change in game parameters again. So, for example, if at late-game the techs are going too slow for everyone, the members can vote for a "drop in rp tech cost" bill and the cost of rp will be lower once passed. I think this system is very versitile and adds interest to those players who deliberate chose different galaxy sizes because they wanted a different pace/flavor of the game.
:mrgreen:

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#56 Post by Geoff the Medio »

I seriously doubt that if we have a Galactic Senate other inter-empire governing body in the game, that we'll only start it up after 80% of the galaxy is populated.

As for voting on increasing / decreasing research rates, my initial reaction is a big "WHA?!" on realism grounds...

Upon further reflection, it seems unwise to implement something like this unless it's agreed to with unanimous consent by all players in a game, as otherwise it could be abused severely... ie. researching up really fast for a while, then switching out of research focus to production while simultaneously voting as a block to reduce research rates so that nobody else can catch up, and enjoying an "unfair" and unsumountable lead in tech. I suppose you could view this as a legitimate strategy, but it seems a bit too powerful to me, as there's nothing another player could do to stop it or react and alter their own strategy in response to a block of players doing this.

I'm not sure how this would work for AI players, if at all, but for multiplayer games, it could be a nice addition to keep things from stagnating.

IMO the interface should be outside of the game setting though... So not in the Galactic Senate, but as a server voting system, like ones used to boot griefer players in FPS games. Along those lines, the system could include other things to alter, like enabling / disabling victory conditions, adjusting travel speeds, chosing between systems to determine how long space battles are, how good the AI players are, turn time limits, etc.

Given the above, IMO we still need an equation-type solution for the tech costs issue, though...

Ablaze
Creative Contributor
Posts: 314
Joined: Tue Aug 26, 2003 6:10 pm
Location: Amidst the Inferno.

#57 Post by Ablaze »

Yes, I was thinking of very large tech decreases, something like a factor of ten. Of course there could be more events, with smaller rewards, but they would behave roughly equivalently and just having two events is simpler to conceptualize.

Also, the costs of technology would have to increase much faster then they do in most games in order to account for this effect. I don’t think that people would get very far researching technology at 10 times its intended cost even in a huge galaxy.

Now that I think about it this system has the additional advantage of also making it much easier to catch up if you are significantly behind in technology.

In reality I would probably make the time that the events fire a little more complex, for instance I might fire the first event when colonized systems = (total systems * .85) + 10 (just in case there is something like 15 systems in the entire galaxy, I wouldn’t want the event to be untirggerable if 4 of them happened to be devoid of planets.)
Geoff the Medio wrote:How does cost reductions for techs at 80% colonization and 30% system exchange rebalance the effects of variable tech trading? I'm not seeing the connection here...
A stronger emphasis on technology will be more and more of a waste of effort until these events fire.
Time flies like the wind, fruit flies like bananas.

User avatar
skdiw
Creative Contributor
Posts: 643
Joined: Mon Sep 01, 2003 2:17 am

#58 Post by skdiw »

I think if you aren't satisfy with any approximations of a formula, using events-trigger, using complicated formula, using s-curve, or even tailoring each and every individual tech cost according to galaxy size/# players, then I think you should just appreciate different map sizes plays differently, which can be a positive thing. Remember back in old moo when you fight with lasers all the time in a small map, then in big maps you tech to autorfire or phasors. I know I play a war race in a small map and a tehc/builder race in a large maps because those were good strategies to use. It all balances out anyway since everyone is in the same boat.



GSU isn't overpowered becuase no one would vote yes for a bill of no use to them. You can propose a research boost only for yourself, but I doubt any player would vote yes. However, if you propose a research boost for everyone, then it is balance so ppl may say yes. You can also do under-the-table deals and goto diplomat screen to pay tribute to empire X,Y, and Z in exchange for a GSU favor so you can get a research boost for yourself only. But in that case, everybody is getting something so it's fair.

As for realism, forget rp cost decrease, just say "this bill allows free trade of ideas between scientist from different empire. As result, more ideas are generated. game effect: rp boost for everybody."
:mrgreen:

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#59 Post by Geoff the Medio »

Ablaze wrote:
Geoff the Medio wrote:How does cost reductions for techs at 80% colonization and 30% system exchange rebalance the effects of variable tech trading? I'm not seeing the connection here...
A stronger emphasis on technology will be more and more of a waste of effort until these events fire.
Aagin, I don't see what this has to do with tech trading...
skdiw wrote:I think if you aren't satisfy with any approximations of a formula, using events-trigger, using complicated formula, using s-curve, or even tailoring each and every individual tech cost according to galaxy size/# players, then I think you should just appreciate different map sizes plays differently, which can be a positive thing.
Just accepting the differences is a possibility, but IMO we'd be better off making the tech rates roughly equivalent between galaxies if possible. There would still be variations between galaxies in terms of gameplay, including earlier wars in smaller maps, but if you managed to survive those wars, the tech progress for the rest of the game would go at a rate consistent with other galaxies.

I'm not dissatisfied with all possible formulas. I just dislike several of the suggested formulas (and events-based systems) for the reasons given.
GSU isn't overpowered becuase no one would vote yes for a bill of no use to them.
If a bill passes with with 50% + 1 votes, then a bloc of empires could abuse the system to their benefit, and the detriment of the rest of the empires, even if the rest of the empires don't vote for the bill because it is of no use to them.
You can propose a research boost only for yourself, but I doubt any player would vote yes. However, if you propose a research boost for everyone, then it is balance so ppl may say yes.
My objectionable scenario above did not rely on boosts to only one player's research.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12802
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#60 Post by Geoff the Medio »

It occurs to me that since FO has pooled production, and since most buildings are built one-or-a-few per empire, FO buildings and basically like techs when it comes to galaxy size balance issues. This is different from most other empire games, in which each planet builds its own stuff, so there's no effect on how much PP a single build location has due to galaxy size.

This has an important consequence for FO , which is that if some galaxies have waaay more PP (or RP) than others, then it becomes very difficult to have any sort of strategic limitation based on the amount of PP (or RP) available. Essentially, in a small galaxy, there are very few PP available total, so the player(s) have to pick only one or two of the very expensive buildings (or techs) they might want to build... whereas in a large galaxy, there are loads of extra PP (or RP) from all those extra planets, meaning the player can build (or research) everything... ie. there's no longer any choice involved.

This is rather distinct from most other games where adding more planets doesn't really change what the choices of what to build are on each planet individually... each one can still build (or not) its own copies of each of the buildings that are available. In FO, there's only one or a few buildings of each type for the whole empire, so adding more planets changes WHAT you can build, not just HOW MUCH of it you build...

I suppose the consequence of this is that in FO, we'll have to inflate building costs just like we'll have to inflate tech costs based on galaxy size. Most other games already inflate tech costs based on world or galaxy size, for the reasons described, but not buildings, since they aren't as directly affected by the size-inflation... however FO is different in this regard.

Post Reply