(Firaxis: READ) Railroad Movement in Civilization IV

[Commander Bello]

Basically, in civ 2, someone noticed that a winning strategy was to build cities next to each other everywhere. It meant of course that the later improvements would be less useful, but by that time you would have a certain win anyway. Having the rail capacity depend on the number of connected cities would grant a big boost to this strategy.

I don't see this.
The capacity would primarily be consumed by using the rails, not or at least to a much lesser degree by the distance. If you occupy a cart, you occupy it regardless whether you travel 10 or 1000 miles (except for the fact that it will become available again in shorter time).
Of course, this would require balancing in how the consumption is weighted, but this - at least to my point of view - is less a concern.
As with almost any concept, the balancing is the crucial part.

 

[Birdjaguar]

Basically, in civ 2, someone noticed that a winning strategy was to build cities next to each other everywhere. It meant of course that the later improvements would be less useful, but by that time you would have a certain win anyway. Having the rail capacity depend on the number of connected cities would grant a big boost to this strategy.

thanks for explaining what you meant. In C3C you cannot build cities next to one another. While cramming cities into small places can be useful in C3C, it wouldn't be the sames as in Civ 2.

 

[rhialto]

My point is though, is that it seems in AL's model that the amount of capacity you receive depends on the number of connected cities. If you have a very high density of cities as in the infinite city sleaze strategy, having rail capacity depend on number of cities will only make the problem worse.

 

[Commander Bello]

My point is though, is that it seems in AL's model that the amount of capacity you receive depends on the number of connected cities. If you have a very high density of cities as in the infinite city sleaze strategy, having rail capacity depend on number of cities will only make the problem worse.

Again, I don't see this happen.
Although closer cities might allow for more cities in a given area, in turn each single city will be of less total value since it has less space to spread.
Speaking in terms of Civ3, if you build cities so close that their 21 squares overlap, they will limit each other considerably.
As railroads are a quite late improvement, this will have quite some effect.
But again, this will be a matter of balancing and testing.

 

[MeteorPunch]

Wow, lot's of good options in this poll that would work. My main beef with the system as it is is the sprawling unrealistic looking railroads. I just want railroads in strategic locations (city to city,across continent,etc), not on every freakin' square.

 

[rhialto]

Again, I don't see this happen.
Although closer cities might allow for more cities in a given area, in turn each single city will be of less total value since it has less space to spread.
Speaking in terms of Civ3, if you build cities so close that their 21 squares overlap, they will limit each other considerably.
As railroads are a quite late improvement, this will have quite some effect.
But again, this will be a matter of balancing and testing.

Yes, if their 21 squares overlap, they lose potential. But until the modern age, that won't be an issue anyway. Consider two players...

One builds cities to optimise for modern age, so that city area overlap is minimised. If he lasts into modern age, his cities will domniate, but the cities are far less able to support each other early on, and he will probably spend more time improving tiles that won't be used for a long time.

The other builds cities in a square pattern every 3 tiles. This optimising for a population of 9, or 6 if he even foregoes aqueducts. However, every tile is guaranteed to be used almost as soon as it can be improved. This player will dominate in earlier ages, and is gambling on a win before teh modern age arrives, where the higher productivity bonuses from modern city improvements can be felt. It is well worth noting here that medieval cities are in no way hurt by the close spacing, as although the city areas overlap, there is still plenty of space for all the citizens that each city has available to work a tile. Overlapping city areas only starts to bite in the modern age.

 

[Aussie_Lurker]

Hi guys, have only just come back on, but am VERY glad to see people debating my model ! Also, unlike certain other people, though I do want to win you all over eventually, it doesn't really bother me if you currently disagree with it.
So, lets see if I can allay your fears Rhialto. OK, I think the key issue is that if you connect 1 city to 1 other city you get, for instance, 0.1 CPs-at base tech levels. Building multiple connections between the same two cities will not gain you ANY new capacity points-only the first!!!! You can still build multiple RR connections per city, but all its going to do is cost you money!
That is part of my other balancing factor-within my model, both cities AND their connective infrastructure cost you gold per turn. This actually came about outside of the CP system-more as a means of limiting 'The Snowball Effect', but could also help to restrict abuse of CP generation (also, failure to put aside enough of your national budget for infrastructure and city maintainance will cause said infrastucture to 'break down'-and cities to gain corruption).
The thing is that the BULK of your actual CP's will come, not from building the original connecting tracks, but from improving Rail and Industrial technology, and the construction of appropriate wonders and improvements.
Lastly, I have been thinking about what you said, Rhialto, and from a game-balance perspective, I'm now not entirely sure that CP's should ADD gold to your economy (though I do think that a RR connection should give a city a boost to its income-based on the # of cities that RR is 'connected' to). However, loss of CP's via the movement of units and commodities, or the destruction of infrastructure should LOSE you income-based on your total # of CP's lost. So, for instance, lets say a player has 30CP's, and has a current budget of 1000 gold, then each CP lost might cost you 50 GPT(5% of total income)-meaning that this player could lose up to 1500 gold if he were to use ALL of his CPs for moving units that turn. This means that, at the end of the day, no matter HOW many rails you build, and no matter HOW many CP's you obtain-by fair means or 'foul'-the # you can safely use for military purposes will ultimately be limited by the strength of your economy. This is more for game-play than realism reasons, but will make the # of units in your army of far lesser importance to ultimate military victory than the underlying strength of both your infrastructure and your economy. This, I hope, will all come together in a system that overall promotes QUALITY of a civ, rather than the shear Quantity of land and/or cities a civ has.
Anyway, I hope that helps Rhialto-as someone whose opinions I genuinely respect (like DH_Epics), I do really hope I can bring you over to my way of thinking !

Yours,
Aussie_Lurker

 

[Evie]

I suppose railroad capacity could work. I'm just very wary of adding more numbers of the players to keep track of in general. Every number you add means one more detail to keep track of for the players, and not everyone likes to have fifty bits of detail to keep track of.

As for limiting infinite city sprawling style effects, makes it that you gain 0 point per village (6 and less), 1 point per 2 city (7-12) or 2 points per metropolis (12+). Building a lot of tightly packed cities doesn't sound nearly so effective for rails if cities under size 12 provides you no capacity points.

Alternatively, one could base the cap points on the shield production of the matching cities. Larger cities = more productive = more capacity points.

In any event, the only thing that really matter for me is that the fanatics don't torpedo the game with their inane hatred of instant-movement. They don't seem to realize that while it may "cheapen" the strategies for the more hardcore players, it's about the only thing that make the game bearable in the later stages for casual players (you know, those who do the BIGGEST part of the buying?) who do NOT want to wait two-three hours for their tanks to move from point a to point b (which is not at all impossible with a fair-sized empire on the larger map sizes).

 

[rhialto]

@AL
I think those numbers may need tweaking a little. 30 gold to use a rail is a little steep I think. The point should be to make a meaningful opportunity cost, not to punish the player. In my model, I was thinking perhaps 2 gold for foot units, and rising slightly for cavalry and armour units, up to, say, 5 gold. Other than you having a cap based on "connection points", it's now pretty much the same as my system.

One thing that makes me a little wary is when you say that the bulk of CPs will come from techs (no big issue on this part), and city improvements / wonders. Does that imply a large number of buildings dedicated to rail? While rail is important, I don't think it's important enough to have a more than a single city improvement dedicated to it.

--

Regarding rail based wonders...

For famous stations, we have...

Grand Central, NY - biggest in terms of platform count
Nagoya, Japan - biggest in terms of floor area (but most of this floor area isn't actually train station as such)
Clapham Junction, London - biggest in terms of trains per day
Shinjuku, Tokyo - biggest in terms of passengers per day, 2nd biggest in floor area.
Ikebukuro, Tokyo - 2nd biggest in terms of passengers per day

And for famous lines, we have...

TGV
Shinkansen
Trans-Siberian Rail
Orient Express

I think it is better to not use the famous rail routes as wonders, due to their geographically dispersed nature making them somewhat odd, as wonders are built in a city from an interface point of view. There's still plenty of famous stations.

 

[Commander Bello]

One general comment to the last postings:
As we are talking about cIV, we should make our minds free of the internal algorithms as currently used in Civ3. That means that there might be the chance that in future ALL city improvements (or at least quite a number) will produce "gold" (as a replacement for the economic effects. In that case, a railroad station could add either a (small) percentage or a (again small) fixed amount to their "gold output".
This is a little bit off topic now, but one of the weakest points in the current design of Civ is just the use of the small numbers which in turn leads to big effects if you add to them or subtract.
A town in future could have size 9 for instance (which would be more logical) and a city could have up to 21. A market in turn could not just multiply your income by 1.50, but add 25 gold or whatever other number proves to work in a balanced way.
/* off-topic mode off

[edit] One comment to Oba:

I'm just very wary of adding more numbers of the players to keep track of in general. Every number you add means one more detail to keep track of for the players, and not everyone likes to have fifty bits of detail to keep track of.

This I see as a matter of the interface. The interface should make the consequences of this additional numbers obvious. This means, if the interface would show you where you can go easily, or where you may go but almost exhaust your CPs (let's say by highlighting those areas in different colors when you click on the "railroad button"), it would become very clear for the player what consequences the use of the railroads would have. The only thing he would have to keep in mind would be that there are numbers in the background, but he wouldn't have to count and crunch them by himself.

 

[Aussie_Lurker]

Actually, Rhialto, what I missed out in my last post was also the role of population in determine Capacity Points. Sort of like what Oda said, but based on the AVERAGE city size connected by your rails. So, for instance, a civ with two dozen size 2 cities is going to recieive a HELL of a lot less CP's from their rail than a civ with under 10 size 12 cities-if that makes sense. What I am thinking is that the average city size/10 should be the multiplier to the overall CP's. So, for instance, if the average city size is 15, then your CP multiplier would be 1.5 times.
As for wonders and improvements, I don't think there should be too many. In most respects I was thinking sort of along the same lines as what you were suggesting in YOUR model-just a few key CP boosting Wonders/Improvements.

As for my numbers, well I have already confessed that they are only there for illustrative purposes. They would need to be seriously tweaked for gameplay balance, but should be aimed at making it more expensive for bigger economies to move lots of units than smaller ones.

Yours,
Aussie_Lurker.

 

[dh_epic]

I'm still concerned that capacity points wouldn't limit much movement, since you'd almost always have enough capacity points to do whatever you want. Moreover, if capacity points are determined by cities and their size, then small empires will be the ones most affected -- an imbalance that doesn't need to be pushed any further.

Aussie, just how many CP's do you think forces someone to be strategic? Do you think you're making tough decisions about which units to move if you have 25 CPs this turn? 15?

 

[Aussie_Lurker]

Well, DH_Epic, it isn't so cut and dried as that. As I said, its AVERAGE size thats important in CP generation, so its an attempt at the 'quality over quantity' system. i.e., a sprawling nation covering hundreds of tiles, but with a small overall city size, will have fewer overall CP's than a smaller nation with very large-sized cities. In addition, though, sprawling nations will have to pay a lot more for their connective infrastructure, which makes the COST of moving units much greater. Anyway, in a hurry at the moment, so I will try and answer this in more detail later.

Yours,
Aussie_Lurker.

 

[rhialto]

Just noticed something. If we consider those regions which are big on passenger train traffic, we get western Europe, Japan, and Taiwan. The common point here is that they have high population densities and the nations are generally not sprawling.

Given the above, a CP model that gives rising CP as you get more and more cities seems destined to help sprawling nations. I expect the EU could handle trains being diverted to military use far more easily than the USA because it has so much more track.

Perhaps having the CP based on, say the average population of the top 5 connected cities (perhaps within a token increase for each extra city) would better reflect reality?

 

[Aussie_Lurker]

OK, just for the record, my CP model helps HIGH DENSITY nations over large sprawling nations-largely because I see it in the broader context of my 'quality over quantity' model. To illustrate (and please remember that ALL numbers are for illustrative purposes ONLY!)

Take two civs: one consists of around 3 dozen sprawling cities-created by a combination of 'settler dioerrhea' and outright conquest-spread out over almost half a continent. This has resulted in an average city population/size of 8. The second Civ , OTOH, consists of a dense cluster of only 10 cities, but has an average population/size of 24. The first civ has an average distance between cities of around 7 hexes, wheras the second civ has only 3 hexes between cities-on average.
So, using my model, and assuming a current 'tech' of 0.5CP's per connection, then this is what you get:

Civ A: around 14 'city-to-city' connections, consisting of around 7 tiles of rail per connection. This amounts to a base of 7CP's. However, with an average city size of only 8, this means that his ACTUAL CP's are (7*(8/10))=5.6, rounded up to 6. So Civ A has 6CP's per turn.

Civ B: around 6 'city-to-city' connections, consisting of around 3 tiles of rail per connection. This amounts to a base of 3CP's. However, with an average city size of 24, this means that his actual CP's are (3*(24/10))=7.2, rounded down to 7. So civ B has 7CP's per turn.
This means that the smaller civ can actually move MORE units by rail every turn. Not only that, but if we assume an average per-turn cost of 1gold/3 tiles of rail, then CivA-with (14*7)=98 tiles worth of rail, for a total cost of 33 gold per turn. Civ B, OTOH, with (6*3)=18 tiles worth of rail, for a total of 6 gold per turn.
So, not ONLY does CivB get more CP's than CivA (albiet by only 1 point), it gets it for only 1/5th of cost!!!
Another factor to consider is that, at 7 tiles per connection, CivA is much more susceptible to sabotage, which could lead to a loss of capacity. In addition, CivA has a much larger frontier to protect, making it more difficult to defend with so few CP's.
All of this, IMHO, is an essential part of helping to curb the dreaded 'Snow-ball' effect.

Hope that helped to clarify things for you.

Yours,
Aussie_Lurker.

 

[dh_epic]

I don't know how, but the math worked. Those levels of capacity points are enough to get stuff done, but not enough that you can do it indiscriminantly. Really, I'm still giggling about the aptly named "settler diarrhea".

 

[rhialto]

One question I have is, does the number of CPs depend on the number of nodes connected, or the number of connections between nodes? That is (assume all cities have mutual connections with every other city here):

cities       CP
1     0 or 1?
2     1 or 2?
3     3 or 3?
4     6 or 4?
5     10 or 5?
6     15 or 6?

If CPs depend on the number of connections, a large civ will gain a huge advantage over a smaller one. But if it is based on number of connected cities, that gives a more reasonable limit.

 

[llib_rm]

Both are (3) only substituting RR MP for A-B MP. Same cons apply. (See my reply to ybbor <below>.)

If you're referring to option (3) (all units literally move the same number of railroaded tiles to the RR limit, after which point the RR bonus no longer applies for the remainder of the unit's turn), then you don't get the $3,000.

Why? Let me point out the reasons:

- Program must 'remember' how many RR tiles each unit has moved per turn (i.e. an additional value per unit that must be read each time the unit moves). (If you didn't have this feature, fast units can just move into a none-RR square and then back again to move another 30 squares or whatever the RR MP is.)
- More MM, as you now have to keep track of a unit's 'RR points' (i.e. number of remaining squares that can be moved on RR by that unit) in addition to remaining MP.

Man these arguements are too easy to shoot down, where's Bamspeedy when you need him?

I did vote for #3, at least I am consistent.

For the record, I think RR should cost one movement point MP for any distance greater than three tiles. Less than three tiles should be treated as road movement using one third MP per tile. Units with two MP would be able to board the RR travel great distances, unload the RR, then board & ride again on the same turn. Units with one MP would board the rails, travel great distances, then stop when the rail ends for they are out of movement points. Currently when a unit with one MP goes from rail-road-rail they can continue. In my proposed scenario the one MP unit would stop on the road. This should present some interesting tactics to air missions. It is also quite different than yborr's "rails move at .25"

 

[Greek Stud]

How about something as simple as charging Money for RR. Like a maintenance tax, it cuts from your account from every tile you build a RR on. That way you can strategically place RR or foot the bill and build everywhere.

 

[Aussie_Lurker]

Its tough to know for sure, Rhialto, without testing it in-game. My first instinct is to go the road of greatest game balance, and say NODES is the answer, because it allows the greatest parity between large and small nations. By the same token though, I also consider the fact that, in order to get multiple connections per city, the larger nation will have to build more track which-in turn-means greater maintainance cost. Remember, also, that money paid into maintainance is money NOT being paid into your research budget which, in turn, means that the smaller civ has a better chance of getting ahead on the techs needed to boost the number of CP's per connection. Also, cities themselves would have a flat cost each-with an additional cost per extra 2 population points, yet another way of curbing the snowball effect!!
That said, though, my above example was based on the NODE system, such that you get the CP's for only the FIRST connection that 2 cities recieve (though a connection between a fort and a city, or a colony and a city, or an outpost and a city, will get you an extra, smaller CP boost.)
Hope that clarifies things.

Yours,
Aussie_Lurker.