The International Simutrans Forum

 

Author Topic: pak128.Britain Experimental feedback from online game  (Read 6910 times)

0 Members and 1 Guest are viewing this topic.

Offline Sarlock

  • Devotee
  • *
  • Posts: 1340
  • Languages: EN
pak128.Britain Experimental feedback from online game
« on: February 15, 2014, 09:55:20 PM »
Long post warning!


I've been maintaining a list of items during the server game which pertain more to the pakset than to Experimental directly.  Things that are bugs are generally reported as I find them, so these are just balancing issues and tweaks to the pakset that could help things.  This focus is on the years 1750-1770, which perhaps hasn't had much scrutiny before due to being such an early period.  Given that we have well over 2,000 convoys running on the map thusfar, this has been a great opportunity to test out some of the early game mechanics.

Having not really played the Standard pak128.Britain, I am not sure whether most or all of these issues also exist in the Standard version.  That may be the case.

Industry/Freight Revenue

As reported elsewhere, the Cattle Farm-->Dairy run is excessively profitable.  Production rates of Cattle Farms and Dairies are quite high and the profits for milk are high which allows for significant profits to be made from these lines.  This is the #1 source of revenue/profits for all of the top players in the game.  I think half of my profits are from running milk.

Second is the planks delivery line, the Forest / Sawmill --> Builders' Yard run.  Again, production rates are high, allowing for significant profits.

Looking at the Goods list, it's clear why these are the two most profitable freight runs as they are the highest in revenue.  Planks are 0.37 while milk is 0.35.  Compare this to meat at 0.11, a third of the price.

It seems appropriate to drop the revenue for milk and planks down to the 0.10-0.20 range.  Actually, without knowing more details about the pakset, it seems that all freight revenue should be in a tighter revenue range.

Grain is also very profitable but it has a built-in limitation to its potential profits in that bakeries have a very low production/consumption level.  It's nearly impossible to find sufficient bakeries to consume even a fraction of the flour a grain mill can potentially churn out.

This moves us over to factories whose production is too low to be useful in producing profits.  If you look at the online game, almost all of the industrial freight lines concentrate on milk, planks and coal.  There is some use of the textiles industrial chain, ironworks and brewery chain but it is limited.  I have tried other chains to experiment and have found some success with the grain-flour line due to the fact that flour is quite profitable despite low bakery demand.  When infrastructure maintenance costs are included many of these lines are not profitable no matter how you set them up.

The Market line is one of the most difficult.  Many markets only consume 16-40 units per month which is very low.  Harder yet is the fact that the market will only store 5 units of fruits and vegetables at any given time.  You can make a tiny profit from delivering this but when offset against the maintenance cost of stations it is a big money loser.  Meat is even more difficult as the market only stores 3 meat and consumes at a 33% rate.  Fish is marginally better due to the higher revenue value of fish, the 200% consumption rate and the 10 unit storage limit.

I haven't tried the clay pit-->pottery-->china shop line but again I think this poses problems from an end consumption standpoint.  China shops produce(consume) around 32 units per month and have a 32 unit storage.  Potteries have a similar production rate while clay pits produce around 300 units per month, or 10x what a pottery can produce (though the pottery does produce at 200%).

A production rate of around 100/month is a decent number to make a small profit but nothing outrageous.  Get below 50 and it's very difficult to make the line profitable with the small volumes.

Obviously there is a significant balancing challenge with the game in attempting to establish production rate levels and revenues/good levels that work both for 1750 and 1950 (and beyond).  From a gameplay standpoint it's nice to have all industrial chains roughly equal in potential profits, otherwise we run in to a situation like in the server game where everyone makes a dash for the 2-3 most profitable lines and ignores the rest due to questionable profitability.

Ship Operating Costs

Ships have a major advantage in that they can transport vast amounts of materials.  This is, of course, true to reality.  A water-based map, such as the one we are playing, is set up for the majority of transportation networks to be water based, at least until the later parts of the game when all inner-island transport will likely transition to rail (inter-island transport will still be air and water based).

The East Indiaman, with an operating cost of 1.80/km, is very cheap to operate.  In fact, it is cheaper than its smaller cousin, the Brig, at 2.40/km, which is still very low.  Given that you can store up to 1,575 cargo on an East Indiaman and 575 cargo on the Brig, the operating cost should be significantly higher.  You can also put up to 700 passengers on the East Indiaman and 150 on the Brig.  You only need about 30-50 cargo to break-even on a journey (depending on the type of cargo) and less than 20-30 passengers.

I would suggest increasing the operating cost for the East Indiaman to around 10.00-15.00/km (or more) and the Brig to around 6.00-10.00/km.  This will keep them from still being profitable even with a 5% load.  You could also consider, instead, of adding an operating cost to the cargo/passenger holds.  This would create an incremental cost for more holds on the ship and make the player consider how many holds to put on each ship/line.  We also want to give incentive to the player to find backhaul cargo for their routes to help cover operating costs - right now the operating cost is so low that it's not a concern to just add a second line for a cargo that could be a potential backhaul.  Hauling costs right now are only around 1-5% of the revenue generated from cargo (or less if the ship is completely full).

I have been using the Brig for all canal routes.  The increased operating cost is still quite low comparatively and the extra capacity and 3 km/h faster speed easily compensate for that cost.  If, however, the cost was much higher, this would make the player carefully consider when to use the larger ships and most canals would instead be using what they should traditionally be using for that era - horse drawn barges.


Way and Infrastructure Costs

As highlighted elsewhere, the operational cost of the Ship Canal (the largest canal type) is too low at 48.00/km.  The barge canal is 64.00/km.  The Ship Canal should probably be quite expensive to operate, given that it can use large ships with high capacities (such as the Brig).  All of the canals should probably see an increase in operational costs, given how much cargo they can transport compared to a road.  I'd consider doubling the cost of the first three canal types and then increasing the Ship Canal price to somewhere around 200.00-300.00/km.  We want the player to carefully consider when to use the heavy duty infrastructure.

The aqueducts should receive a similar price increase.  They are actually cheaper than the canals they match up to size-wise.  These should be quite expensive to build and maintain - and used only when necessary to cross difficult terrain.

This brings us to docks and storage infrastructure:

Passenger docks, at 300.00/month, are quite expensive to maintain.  They have ample storage, however, at 85 passengers and mail, but when compared to other docks, aren't price competitively.  If the player had no choice but to use these, then it might work, but when you can instead build a goods dock, at 72.00/month and tack on two staging inns (40 mail/passengers/freight each), at 8.96/month/each, you can achieve the same thing for just 90/month, less than a third of the price.  I would lower the maintenance cost of the passenger dock to around 120/month.  Also, given that passenger revenues are fairly difficult to raise in the early years without connecting a lot of limited-revenue routes, you need a lot of passenger docks in order to built up a profitable core transportation service.  Reducing the cost of the docks will help a lot.

I would also raise the cost of the staging inn to be 4x that of the staging post.  The staging post is 8.00/month, I would bump the staging inn to 32.00/month since it has 4x the storage capacity.

I would likewise lower the maintenance cost of the storage buildings, post office, bulk goods bunker, livestock pen and goods warehouse.  The 32 freight storage of the goods dock is too low to be useful and needs to be augmented with at least one building to boost its freight storage up to usable levels.  Road cargo bay should be lowered as well, from 180.00/month to maybe 50-80.  Road freight is difficult to make profitable as it is, due to low speeds and capacities, and offsetting the monthly cost is tough.  Likewise, the coaching stable can be reduced in monthly cost as well.

You could also consider increasing the building cost of some items in compensation: a goods dock is just 187.50 to build and 72.00/month to operate.  If you lower the maintenance cost to 30.00/month you could increase the building cost to 1500 or more.  There should be a higher capital cost for many buildings.  Right now the capital cost is secondary to the operational cost for most infrastructure.



As we move in to later years I am sure I will have much more to add, but this is an overview of the major imbalances that I have run across in the 1750-1770 year range.  I am sure other players have other things to add which would be wonderful.

I am not sure if pak128.Britain Experimental still ties is balancing/cost structure with pak128.Britain or if the offshoot is independent of those figures now.  I'd be happy to offer some of my time and experience with finding a better balance for these early years.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18684
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #1 on: February 16, 2014, 01:46:59 PM »
Thank you - this is a very interesting analysis. As you will be aware, the pakset is a long way from being fully balanced. However, one thing that has been balanced since version 0.8.4 are the relative revenues of different types of cargo: these are balanced based on the rates charged by railway companies, all based on historical sources. See the comments in this .dat file for more information.

This suggests that what needs to be changed for the dairies, etc. is not the price but the distance between industries and the production rates. I strongly suspect that the cattle farms are producing far too much and the dairies consuming far too much to be accurate.

What would be very interesting is an historical breakdown of actual farm production to get a sensible idea of how to calibrate this.

Offline ӔO

  • Devotees (Inactive)
  • *
  • Posts: 2345
  • Hopefully helpful
  • Languages: en, jp
Re: pak128.Britain Experimental feedback from online game
« Reply #2 on: February 16, 2014, 02:19:13 PM »
How about adding cheese as a goods?

In that era, refrigeration technology, although it existed, was still in its infancy and cured or long lasting foods were the staple foods.

The goods to cure food, such as salt, sugar, pepper and other spices, were also popular trade commodities.


Alternatively, since refrigeration is energy consuming, adding some costs to that type of cargo hold would also make sense.

Offline jayem

  • *
  • Posts: 11
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #3 on: February 16, 2014, 06:04:55 PM »
From here (was a basic google plus reading so not too sure of any inaccuracies, will repost/edit as I find more)
http://www.h-economica.uab.es/wps/2006_10.pdf

Milk consumption went from 50(l/person/year) to 100 between 1870 and 1930 (in the UK)

In Barcelona (which was behind Northern Europe milk wise)
1850 milk from cattle (going into the city)
1900 milk from 1400 cows in 180 herds (based in the city) giving 5.5 million liters
  (S: this change would probably occur as the city swallows up nearby farms?)
1920 milk from 7500 in 600 herds (based in the city) giving 25 million liters
        +milk from ? in ? (based outside the city) giving 15 million liters
Then occurred Pasteurisation in spain, city herds stayed the same but creameries increased
1930 milk from 7500 in 600 herds (based in the city) giving 30 million liters
        +milk from ? in 900 herds of 15 companies (based outside the city) giving 50 million liters
 (S: this would correspond to a longer acceptable delivery time, i.e acceptable range of suppliers)


I.E. an average city dairy herd having 10 cows giving around 35,000 liters a year, and an average late rural herd about double
     (the current pack has a dairy giving 35 crates/month, ~ (churns/crate)*1500 liters? so looking about right...although scaling issues exist.

------(further stuff from http://www.igg.org.uk/gansg/12-linind/milk.htm)
before 1960s (~70/40) liter churns used in wagons)
after 1960s (12000 liter tanks used) enough to supply 35K people
(but off course then we run into scaling issues)
« Last Edit: February 16, 2014, 06:12:53 PM by jayem »

Offline AP

  • Devotee
  • *
  • Posts: 1202
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #4 on: February 16, 2014, 06:57:53 PM »
  • Warehousing is expensive to maintain, but staging inns are cheap, so it's cheaper to build a whole row of inns (even with the 10:1 capacity difference and the road tile to maintain).
  • Why is their maintenance for a heap of coal, iron ore, stone or clay, sitting in a pile on the ground. Only perishable goods need storing indoors against the elements.

Offline Sarlock

  • Devotee
  • *
  • Posts: 1340
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #5 on: February 16, 2014, 08:29:11 PM »
@280,535 is a perfect example of a staging inn complex that can store 800 units of cargo for much cheaper than a freight yard [net cost 180/month rather than 900/month using livestock pens].

The problem with balancing revenues against reality is that we are not modeling all aspects of reality here: some goods had a higher transport value because they had spoilage issues (requiring a faster transit time and more expensive cargo holds), higher storage/transport costs, etc.  In the Simutrans world, all these are equal and transporting a bulk pile of coal is the same cost to the player as a hold full of delicate china.  It just effectively biases all players towards the high revenue cargo.

You could mitigate this a bit by making certain holds more expensive than others, but this creates another level of balancing that has to be carefully calibrated.

Offline AP

  • Devotee
  • *
  • Posts: 1202
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #6 on: February 16, 2014, 08:38:24 PM »
Not to mention that warehouses are available from 1750 to 1999 and there is no inflation mechanism in the game.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18684
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #7 on: February 16, 2014, 08:49:24 PM »
  • Warehousing is expensive to maintain, but staging inns are cheap, so it's cheaper to build a whole row of inns (even with the 10:1 capacity difference and the road tile to maintain).
  • Why is their maintenance for a heap of coal, iron ore, stone or clay, sitting in a pile on the ground. Only perishable goods need storing indoors against the elements.

The answer is that there is no way of distinguishing between storage capacities of different types of goods at stations. Anything that can hold bulk goods can also hold a like amount of perishable goods, and vice versa.

Offline AP

  • Devotee
  • *
  • Posts: 1202
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #8 on: February 16, 2014, 09:20:13 PM »
But that doesn't explain the cost differential.

Offline Jando

  • *
  • Posts: 497
Re: pak128.Britain Experimental feedback from online game
« Reply #9 on: February 17, 2014, 12:20:40 AM »
...
Having not really played the Standard pak128.Britain, I am not sure whether most or all of these issues also exist in the Standard version.  That may be the case.

Nice post, Sarlock!

In another thread (A path to balance here: http://forum.simutrans.com/index.php?topic=12763.0) some time ago I compared the balance in Experimental to Britain128.pak on Standard - those two are balanced quite differently.

The most notable difference is (in my opinion) that Experimental has rather high maintenance costs but rather low operating and low capital costs compared to Britain128.pak on Standard. I think that's the main reason why ship routes over sea are so profitable in Experimental. Of course no maintenance has to be paid at all for sea routes, thus the usually quite high maintenance costs in Experimental don't apply here. But those routes still profit from the low operating and capital costs in Experimental.

Personally I would not only increase the operating costs of ships - I would make them much more costly to acquire. An East Indiaman now costs 2.500 cr, I wonder how the game would change if Brig and East Indiaman would cost 20 times the money they cost now. :)


...
 I would also raise the cost of the staging inn to be 4x that of the staging post.  The staging post is 8.00/month, I would bump the staging inn to 32.00/month since it has 4x the storage capacity. ...

I would also completely remove the cargo capacity from the staging inn. Let the player build a staging post (10 cargo) when he just needs a drop-off stop for cargo. Let him build proper docks and warehouses when he needs to handle mass amonts of cargo.


In general: From my experience of playing the game between 1800 and 1900 I would re-balance Experimental along the lines of
A) Much higher capital cost for almost everything
B) Moderately higher operating costs
C) Much lower maintenance cost for almost everything

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18684
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #10 on: February 17, 2014, 12:30:27 AM »
Jando - useful feedback, thank you. I like the suggestion about the staging inn, and will implement that straight away. The others will have to await a comprehensive balancing exercise. However, do you have any idea of the actual construction cost of sailing ships in this era? That would be very useful. Crew numbers can also give a useful guide to the fixed monthly cost when that is implemented. If you have that information, it would be very helpful to post that on the "snippet of relative pricing information" thread stickied in this subforum.

Offline Jando

  • *
  • Posts: 497
Re: pak128.Britain Experimental feedback from online game
« Reply #11 on: February 17, 2014, 01:22:51 AM »
He, Lloyd's Register will know, after all commercial insurance really started with insurance for ships in the age of sail, at least I believe it was so. They even have a historical research service, sadly it's not free. :)

http://lr.org/about_us/shipping_information/Historical_research_service.aspx

I'll have a look around the Internet what I can find in the coming days.

Offline AP

  • Devotee
  • *
  • Posts: 1202
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #12 on: February 17, 2014, 07:47:29 PM »
The answer is that there is no way of distinguishing between storage capacities of different types of goods at stations. Anything that can hold bulk goods can also hold a like amount of perishable goods, and vice versa.

Is that something that could easily be 'fixed'? If you're splitting goods into categories as per the other conversation about industrial chain radiuses, it would just be one more related thing, even if on a per-industry basis. Storage is either outside, inside, or inside (cooled) presumably...

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18684
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #13 on: February 17, 2014, 08:07:54 PM »
Is that something that could easily be 'fixed'? If you're splitting goods into categories as per the other conversation about industrial chain radiuses, it would just be one more related thing, even if on a per-industry basis. Storage is either outside, inside, or inside (cooled) presumably...

It is fairly fundamental and may well require changes to quite a few basic elements of the code in all sorts of places.

Offline zook2

  • *
  • Posts: 321
Re: pak128.Britain Experimental feedback from online game
« Reply #14 on: February 18, 2014, 01:55:55 AM »
I agree that the maintenance costs are too high, compared to construction costs. My initial strategy in my 1860 game was to build as many lines as I could, to connect as many places as possible to the network to generate passenger demand, before my money ran out. While having funds in the five-to-six-figure range, constructions costs for infrastructure are neglible. Only maintenance and running costs will eat you up.

Offline wlindley us

  • Devotee
  • *
  • Posts: 970
    • Hacking for fun and profit since 1977
  • Languages: EN, DE
Re: pak128.Britain Experimental feedback from online game
« Reply #15 on: May 28, 2014, 03:14:20 PM »
Observations in 1853.
  • Narrow gauge locomotives are not yet available, but why does the Narrow gauge menu have only the brick platforms, while Standard gauge is still stuck with only wooden platforms?  That seems backwards, I'd think N.G. would tend to have the cheaper/older materials.
  • Goods annexes: Since livestock productions levels are much lower than bulk, would there be support for about realigning the available additions? Something like the following.  I can derive a new, two-storey Large Goods Warehouse if this all makes more sense than the current.
 
   
Rail
Capy
Mail
Build
Maint
Goods Siding
*
0
0
50
5
Livestock Pen
50
250
100
Goods Warehouse (Small)
100
50
1000
150
Bulk Goods Loading Tower
*
250
0
1000
150
Goods Shed
*
250
0
2000
125
Bulk Goods Bunker
750
0
2000
200
Goods Warehouse (Large)
750
250
5000
250
Post Office
750
2000
250

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18684
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: pak128.Britain Experimental feedback from online game
« Reply #16 on: May 28, 2014, 07:14:38 PM »
The narrow gauge platforms use brick because the wooden platform graphics do not exist; but narrow gauge is not much use for passengers in any event until somebody invents the narrow gauge steam locomotive, for which you will have to wait until 1863, which is when the first actual narrow gauge steam locomotive was produced (and that very locomotive still exists in preservation). Before 1863, narrow gauge was confined to horse working, making it suitable for only specialist uses.

As for livestock pens and the like., the difficulty is that there is no functional separation as to different sorts of goods station extensions: all goods extensions increase capacity for all types of goods. It does not therefore make sense to have different buildings with different ratios of capacity to cost or capacity to size in the same era or else the system will be open to abuse. I did wonder whether to reduce both capacity and cost of the livestock pens, but this would simply encourage players to use bulk goods silos for livestock where space is scarce, as they would then be more space efficient.