Author Topic: Question: Pay_for_total_distance  (Read 4488 times)

0 Members and 1 Guest are viewing this topic.

Offline colonyan

Question: Pay_for_total_distance
« on: January 01, 2015, 04:18:24 AM »
0 charges for each station (pays for each stop)
This earns most as customer pays regardless how much route is making unnecessary detour

1 charges for each line customer uses (pays for each line)
I don't quite get this. I assume this is basically option 0(in terms of line changing station) but company loses on lines that are making large detour

2 charges for straight line of customers starting point and destination point (pays straight line distance of starting location to destination location)
This earns the least and most difficult as any detour is net deficit for transporter

Can anyone clarify a little more on option 1?

For example
Assuming entire route to be like letter "G", which have 3 lines, line A being longest one making large curve, line B to going up straight, line C to
be shortest one that going little bit to the left. In short, Line C is net earn. Line B is net earn, Line A is large loss assuming from large detour.

Easier example
This time assume entire route letter "L" starting from top to going down right. Changing once at bottom left, line 1 and line 2.
Line 1 is net gain and line 2 is also net gain. This is identical to option 0.

Last example
letter J. This time only one line. But many stations at 90degree angle of top right corner, and few more along straight section and final curve
section. Based on explanation above, this is situation of setting 2. Making loss as much as line takes distance from straight line between
starting location and destination location.

So in conclusion, option 1 is tad closer to option 0 as customer pays all no matter how far the changing station might be from the start/end straight line.


Offline DrSuperGood

Re: Question: Pay_for_total_distance
« Reply #1 on: January 01, 2015, 05:51:36 AM »
It has nothing to do with lines. It is the payment model based on the distance travelled.

Let me quote what the game says about it...
Quote
# three modes (default = 0)
# 1: the payment is only relative to the distance to next interchange, 2 to the trips destination (default 0 is distance since last stop)

Basically they do the following.
0-> For all carried goods you get paid for the Manhattan distance travelled since last time the convoy stopped. A convoy stops either at a stop or when schedule is being changed.
1-> For all goods deposited at a stop you get paid for the Manhattan distance travelled from the goods boarding point.
2-> Each good at each stop pays you for the Manhattan distance travelled towards its ultimate destination. In theory shipping away will result in you being billed a negative for the destructive distance. This is similar to OpenTTD shipping mechanics.

With 0 active as long as you place a stop before and after destructive shipping points on a line that can service the type of goods carried you will appear to always be shipping positively. This is very useful for commuter lines which often fold on themselves (subways) or have rather chaotic routes (busses) since it turns what would normally be destructive shipping into constructive shipping. This is pretty much required for pak64 unless you want to pollute the map with "point to point" lines all over it. It is possible to use the "cargo bounce exploit" in this mode as you will still be paid for positive shipment by re-visiting the same two stops a near infinite number of times in a long running loop. In servers manual administration is required to punish such abuse.

1 is a slight revision of 0 where you get paid only for the distance between boarding and alighting. This prevents you from using stops before and after destructive segments to generate extra constructive shipping distance. It also make the cargo bounce exploit impossible however people can still use an exchange exploit bug involving multiple stops to get extra cargo distance. In pak64 some types of service become very difficult to run with a profit such as subways and buss services due to the high operating costs involved for minimal area covered.

2 is simply OpenTTD style payment. No abuse can be done since you will always only be paid for physically constructive shipping. Profit from subways and buss services become minimal and highly prone to loss. Playing pak64 is practically impossible as everything has to be point to point to cope with the huge operating cost fraction (a single tile detour can lose you multiple tiles of profit). Only paksets which are maintenance heavy (eg pak128) can take advantage of this since you can afford to detour more (several tiles of detour per positive tile) however it does fall into a hard mode and is only recommended by enthusiasts or people seeking a challenge.

Mode 0 is the most realistic. In real life the cost of shipment comes down ultimately to the actual distance travelled from the point of departure to the arrival at the destination, including all detours made along the way. This is why postage to/from remote areas costs more. These higher costs are often masked by forced price averaging (most local postage services) or with government subsidies (highly remote areas that the government wants to keep populated so have to lower the cost of living in).

The only thing unrealistic about mode 0 is the actual departure logic. Goods should not be sent if the route is total nonsense (re-visiting the exact same 2 stops 100 times before reaching the destination or zigzagging all over the map). A competitor can offer a more direct route (less stops between) and will get priority over you. Experimental solves this by adding some form of maximum life time for goods which means that they will not take a route that is too indirect.

Mode 0 and 1 will earn the same amount if a convoy travels only in one direction through all stops (the classic straight line railway). Mode 2 will earn the same as 0 and 1 for point to point routes (from ultimate source to ultimate destination).

Offline colonyan

Re: Question: Pay_for_total_distance
« Reply #2 on: January 01, 2015, 07:42:51 AM »
Thank you very much for detailed explanation.
So 3 modes have profound relation with the design of the pak.

Hearing that this is based on Manhattan distance, I feel that mode 2 sounds like best option for game to be fair about how much player can earn.
I mean really its Manhattan distance, not the straight line with square root of sum or power 2 of x and y. That's my view though.

Short distance train/metro and buses not making money:
I think this has to do with the locality factor. Not every one can travel to foreign country or far away city regularly. Disparity of income and salary...
I believe  as long as company make profit as whole, feeder lines being break even or red is not an issue.
 
As conclusion, I don't quite see the point of mode 0 and even mode 1. With speed bonus settings, it could be circumvented easily.
Mode 2 seems most cheat proof and balanced as game mechanism. In this game where transportation share is 100% of player's and there's no decline of
development and population, why be so active about protecting player's income base...?

Ok, I see that mode two leaves less freedom on network design. But it is Manhattan Distance... All it need to do is to fit in square and reduce unnecessary detours.

Offline Spenk009

Re: Question: Pay_for_total_distance
« Reply #3 on: January 01, 2015, 11:30:10 AM »
Mode 1 is actually semi-realistic. In areas with different transport providers, people pay for each trip with each provider/company. E.g. in the Netherlands, the distance the bus goes is measured in km and then paid for at the end of the ride (by checking in and out each time you bord and exit). Trains are a bit different, where one checks in at a station and out again at the destination station and the cost of the travel is shown to you by the check-in machine (No increased costs or paying at changeovers). If you switch companies there, you do check out of one and then into the other on the platform.

Mode 2 is sort of applied if you purchase a paper ticket from one destination to another, but the resolution of this model is too low for inner city services (you can't get on a bus and book from the stop you start at to another bus stop in another city).

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #4 on: January 01, 2015, 01:02:50 PM »
1 is probably the most realistic for passengers, especially on long-distance, while 2 is more realistic for mail and goods. Around here, local passenger transport often has a flat rate, or a price that rises in steps based on zones. 0 is perhaps the closest thing to the latter.

Offline DrSuperGood

Re: Question: Pay_for_total_distance
« Reply #5 on: January 01, 2015, 04:16:16 PM »
0 is actually the most realistic for passengers. If your train goes in a huge loop so you end up 1 km away from where you start in a different stop then you will pay both for going to the furthest point away and back to get there. The only thing not realistic about 0 is that you get paid at every stop (and not when they board or depart) but that is trivial and purely due to game mechanics.

1 forces all lines to be very straight in a single direction. This causes all sorts of problems for local pickup lines since passengers then get cheaper than real life tickets (they can travel a long way on your service but only pay for the sort 3-4 km distance between the embark and disembark stops.

2 is completely unrealistic since it would mean that sending something up to the top of Mount Everest from its base is cheaper than sending something from London to Edinburgh. You need to factor in real distance the good is moved and not how much closer it physically gets towards its end destination. If you have to zigzag up a mountain side in real life this is factored into the cost so why should it not be in Simutrans?

If anything there needs to be a new mode which pays for physical distance travelled. This would remove the need for re-direction stations for destructive shipping movements in mode 0.

Mode 0 is the only way to really play pak64 since your operating costs are so huge that you need to constantly ship positive to make profit. Running mode 2 in pak64 means that the only way to make profit with industries is to have them running point to point all the time shipping positive since a single tile of destructive shipping eats the profit from multiple tiles of constructive shipping. Pak128 is more tolerant so can be run in any of the 3 modes but generally people find mode 0 most fun since they often struggle to make profit early on as is.

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #6 on: January 01, 2015, 06:08:47 PM »
2 is completely unrealistic since it would mean that sending something up to the top of Mount Everest from its base is cheaper than sending something from London to Edinburgh. You need to factor in real distance the good is moved and not how much closer it physically gets towards its end destination. If you have to zigzag up a mountain side in real life this is factored into the cost so why should it not be in Simutrans?

This is true for all of them, because they use Manhattan distance. You can just get around it for the others by having intermediate stations, as long as maintenance costs are low for stations. Ultimately, it's down to the simplistic economic model in Simutrans where the willingness to pay is directly proportional to distance and speed. In real life, slow and short travel might have higher ticket prices than fast, long-distance travel, per unit of distance. And for equal distance, slower might some times be more expensive.

Offline colonyan

Re: Question: Pay_for_total_distance
« Reply #7 on: January 01, 2015, 09:31:49 PM »
In real world, company have lot of responsibility of gaining back invested capital as it is real life money. Also there is a situation where public transport is served but due to various restriction(finite infrastructure budget, natural obstacles and other responsibilities), it has to make detour. Thus, charge the users for it.

I also agree that lines all over the map is not ideal. Say there are 9 cities in position of ten number keys on keyborad, having 6 lines to connect just 9 cities is messy.

But in Simutrans, various parameters can be modified in config. Goods price, speed bonus, running and way cost and so on.
And there is no real competition(other than one in network game). No one rides private car.(Although they appear as city car) Thus as game, mode 2 leaves bigger opening for balanced and cheat proof play than 1 or 0. Setting profit margin large seems to allow even in mode 2, the more free form network instead of point to point line. That was my personal impression so far.

So if it was to design pak set for each mode
0 - Vehicles are required to run on highest boarding percentage as player is payed for every meter goods have moved to avoid excessive profit which takes away challenge from the game. Lowest profit margin per distance goods are carried.
1 - Also low profit margin and requiring higher boarding percentage. I don't quite digest this mode yet.
2 - Higher profit margin per distance goods have traveled. Lower running cost with highest way construction cost to prevent ways to be placed all over the place.

Does paks like 64, 128 and 128 britain are designed in mind of be played in specific mode? I assume 64  had no choice to be go with mode 0.

 

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #8 on: January 01, 2015, 10:29:24 PM »
The difference between 0 and 1 is probably only relevant for passengers and mail, except airplanes. At least it doesn't seem natural for me to have more than two stops on freight and airplane lines. Bus lines perhaps suffer the most from not being 0. However, if they don't, then buses will likely be too profitable on straight runs.

Offline colonyan

Re: Question: Pay_for_total_distance
« Reply #9 on: January 01, 2015, 10:55:55 PM »

The difference between 0 and 1 is probably only relevant for passengers and mail, except airplanes. At least it doesn't seem natural for me to have more than two stops on freight and airplane lines.
I think I get the mode 0 and 1. 1 is very close to 0. Except, player suffer high running cost if section(s) of used line(s) are making large detours. That is likely to
not happen often. Thus 1 is  close to 0.

However, if they don't, then buses will likely be too profitable on straight runs.
Two class of buses. Also use speed bonus settings. Especially high speed bonus on target good.
One that are affordable to purchase, but gives low profit margin because of speed bonus setting.
Other that are very expensive to purchase, but gives fair to good profit thanks to higher speed bonus that applies.

Offline DrSuperGood

Re: Question: Pay_for_total_distance
« Reply #10 on: January 02, 2015, 02:11:29 AM »
Quote
That is likely to not happen often.
In reality it happens very often to avoid polluting the map with ways. In pak128 you can mostly ignore it however in pak64 you need to use mode 0 with re-direction stops in order to keep such lines profitable.

Pak64 was designed to be operating cost heavy where most money is spent running convoys rather on maintenance of infrastructure. Making 5 tiles less detour on a route can make a noticeable increase in profit even if it requires bridges or tunnels.
Pak128 was designed to be maintenance cost heavy where most money is spent on infrastructure maintenance rather than running the convoys on it. Saving on 5 tiles of tunnel or bridge on a lowish traffic line will more than compensate for a 30 or 40 tile detour.

Quote
Other that are very expensive to purchase, but gives fair to good profit thanks to higher speed bonus that applies.
Busses mostly run inside cities so are already speed limited by the low city speed limits. If you use busses for long distances you are seriously doing something wrong (no way you can ship >200,000 passengers per month with busses between two stops).

Offline colonyan

Re: Question: Pay_for_total_distance
« Reply #11 on: January 02, 2015, 04:12:13 AM »
Ok, so this is the situation where vehicle are restricted with speed in mode 2.
50km/h city road limit was something I did forgot.

I think mode 2 will require city bus type and high way coach bus. City bus will have lower running cost.
The issue will be highway coach type that have high proportion of city road for its route. .....
Its profit will be to small or go to red.
Real world has highway but simutrans one will over ridden by city road so it is a problem unless player builds by them selves.

All I can think of is to put the speed bonus much higher in general. Like +15% for all. Even for garbage.
Speed bonus could be used to balance the income instead of using it just for bonus income. (Or it is there for this kind of purpose?)

Well, thanks for everyone for feedback especially DrSuperGood.
I have better understanding now.

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #12 on: January 02, 2015, 09:31:46 AM »
Busses mostly run inside cities so are already speed limited by the low city speed limits. If you use busses for long distances you are seriously doing something wrong (no way you can ship >200,000 passengers per month with busses between two stops).

In Simutrans world perhaps. Over where I am now, buses mostly operate outside the city. A decent train service restarted a little over a year ago, but apart from that, it's buses or nothing. And there is still a bus service running along the rail line (mostly, same endpoints).

Offline Octavius

Re: Question: Pay_for_total_distance
« Reply #13 on: January 08, 2015, 07:39:10 PM »
In Simutrans world perhaps. Over where I am now, buses mostly operate outside the city. A decent train service restarted a little over a year ago, but apart from that, it's buses or nothing. And there is still a bus service running along the rail line (mostly, same endpoints).
I assume that in very sparsely populated countries like Norway with a very sparse railway network you have a lot of intercity buses. Over here, in the very dense population, buses are mostly for intracity transport or for rural areas, where they connect the small villages to the somewhat larger villages (somewhat larger meaning about 10,000 people), from where trains bring passengers further. Often the trains and buses are operated by the same company.
0 is actually the most realistic for passengers. If your train goes in a huge loop so you end up 1 km away from where you start in a different stop then you will pay both for going to the furthest point away and back to get there. The only thing not realistic about 0 is that you get paid at every stop (and not when they board or depart) but that is trivial and purely due to game mechanics.
I agree with that. And of course nobody would use that train, but that's a different matter.
If anything there needs to be a new mode which pays for physical distance traveled. This would remove the need for re-direction stations for destructive shipping movements in mode 0.
When I travel by bus, I pay for the physical distance traveled, to 65 m (or €0.01) accuracy.
Mode 1 is actually semi-realistic. In areas with different transport providers, people pay for each trip with each provider/company. E.g. in the Netherlands, the distance the bus goes is measured in km and then paid for at the end of the ride (by checking in and out each time you bord and exit). Trains are a bit different, where one checks in at a station and out again at the destination station and the cost of the travel is shown to you by the check-in machine (No increased costs or paying at changeovers). If you switch companies there, you do check out of one and then into the other on the platform.
To put it a bit more precisely, passengers pay here for the shortest possible journey consistent with the places and operators where they checked in and out using their chip card. Operators can't cheat by using a very slow or inconvenient direct service to scare people into using a detour and let them pay more; if the short service exist people pay for that, even if they take the faster detour.

On buses, one checks in and out when entering or alighting, so one pays for the actual distance traveled. On trains, where one checks in and out when entering or leaving the station and when changing operator, there may be multiple possible routes. For example, when traveling from Roermond to Deventer, one can change trains either in (A) Nijmegen (2:35 hours, 149 km), (B) 's-Hertogenbosch (2:31 hours, 191 km) or (C) Utrecht (2:29 hours, 211 km). On route A, the cheapest route, one uses two different operators. On routes B and C one uses only one operator, the same on both routes, so there is no distinction and in both cases one pays for 191 km. But in every case, it's the distance measured along the track (with the exception that when construction of a line was particularly expensive, they use shorter kilometres).

To complicate matters, ticket price is non-linear. When passengers travel far, they pay less and less per km, and after 250 km each additional kilometre is even free (maximum distance one can travel here is 430km – yes, small country). Then there are complications when changing from one operator to another, making things so complicated that the operators themselves don't put the full mathematical details on their websites, referring instead to the on-line travel planner.

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #14 on: January 08, 2015, 08:27:07 PM »
To complicate matters, ticket price is non-linear. When passengers travel far, they pay less and less per km, and after 250 km each additional kilometre is even free (maximum distance one can travel here is 430km – yes, small country). Then there are complications when changing from one operator to another, making things so complicated that the operators themselves don't put the full mathematical details on their websites, referring instead to the on-line travel planner.

If I order the tickets a well in advance and with a bit of luck, I could travel by train the entire length of the contiuous Norwegian rail network (199 NOK, 26 hours 27 minutes) for less than twice the cost of a regular bus ticket from here to the next city (139 NOK, 1 hour 15 minutes). All trains and the bus is run by the same concern, although with different governmental levels involved.

Offline Spenk009

Re: Question: Pay_for_total_distance
« Reply #15 on: January 08, 2015, 11:04:29 PM »
Dutch national rail ticket prices are time indepdendent it seems. International services are usually facilitated by the DB, who try giving the UK (and accordingly Norway too) a good run for their money in terms of a confusing pricing system.

We cannot simulate advance purchases and the pax just show up when they feel like travelling, so I don't feel inclined to give them any reductions in fares. Like in the real world, last minute tickets don't exist...

Offline Ters

  • Coder/patcher
  • Devotee
  • *
  • Posts: 4895
  • Total likes: 214
  • Helpful: 108
  • Languages: EN, NO
Re: Question: Pay_for_total_distance
« Reply #16 on: January 09, 2015, 06:22:03 AM »
My point is that realistic prices are such a complicated thing that attempting to make prices in Simutrans more like it, is likely to both succeed and fail at the same time. It will become more realistic in one area, and less or equally unrealistic in other areas. At least for passengers. Freight might be more predictable, but I have no knowledge of how pricing works there. Except that the only reasonable way to move goods in Norway is apparently to bring up trucks and drivers from the region from Poland to Bulgaria.