News:

Use the "Forum Search"
It may help you to find anything in the forum ;).

[0.8.4] Mail is extremely profitable

Started by ӔO, November 25, 2012, 03:44:03 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

ӔO

Especially in early era, as apparent by the online game.


I would think the problems are thus far
1) revenue is too high
2) too much mail is generated
3) mail does not seem to be affected by refunds
4) mail carriages have too high a capacity for their maintenance
5) mail does not have journey time tolerances


Possible ways to fix
a) Lower revenue from $0.23 to around 75~50% of that.
b) reduce generation factor
c) increase maintenance for carriages


Other ideas that are not entirely pakset dependant
a) reuse passenger generation code, but setup in inverse, where mail prefers long distance over short distance (may not be entirely historically accurate, unsure)
b) enable refunds for mail and other goods
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Junna

Do you have to attack anything that allows even the slightest profit. :(

I don't think it is a problem whatsoever.

I don't like the idea of making it so difficult to play.

o_O

Quite a bit of mail is generated, especially with the low capacity of some of those mail vehicles.  It seems like mail would be more of an add on rather then something you are hauling vast quantities of. 

ӔO

Quote from: Junna on November 25, 2012, 05:07:33 AM
Do you have to attack anything that allows even the slightest profit. :(

I don't think it is a problem whatsoever.

I don't like the idea of making it so difficult to play.

the problem is, it is extremely profitable and by extremely, I mean horse carriages receiving 70k income for mail.

And that wasn't the whole line, it was a single horse carriage that did that.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Junna

Quote from: ӔO on November 25, 2012, 07:10:40 AM
the problem is, it is extremely profitable and by extremely, I mean horse carriages receiving 70k income for mail.

And that wasn't the whole line, it was a single horse carriage that did that.

Was that for a single trip? That does seem quite excessive.
Passengers should perhaps pay a bit more early on though, considering how slow they are until that 48km/h train is available... Income from pax is very limited unless you have quite an extensive network, which is very difficult immediately upon start (born out further by the many newly set up players that go bankrupt - realistic, in a way, but it's kind of bothersome when all infrastructure vanishes. Except canals I guess? Player 8 made good usage of old abandoned canals from a defunct company, I noticed.)

jamespetts

The 70,000c for a trip was probably a bug of some sort, but I cannot find a way of reproducing it so that I can track it down and fix it. It only occurred once so far as I know.

The main problems, I think, are that the mail carrying road vehicles of the era have too much capacity and that too much mail overall is available to transport. I think that it is correct that mail has a slightly higher revenue per unit than passengers. I have corrected the capacity of the early mail road vehicles for the next version of Pak128.Britain-Ex, but the volume issue is a more difficult one.

At present, on the Bridgewater-Brunel server, which is running a game in about 1793, enough mail is generated to fill enough mail carriages to clog the roads of small towns. This is clearly vastly excessive. Passenger generation is constrained by journey time tolerances: even if passengers can get to their destination, if they cannot do so within an acceptable time (and what counts as an acceptable time for passengers is randomised between a range fixed in simuconf.tab per unit of passengers generated), they do not travel. Mail generation, however, is not so constrained, partly at least because it was thought that people would send mail even if it would take a long time. However, I wonder whether this is correct: firstly, if mail takes a long time to get to its recipient, it is longer before the recipient replies to it; in chains of correspondence, which is a primary usage of mail, a reduced transmission speed would greatly reduce the volume of mail per unit of time. Secondly, there are many occasions on which an item of correspondence is only useful if it can reach its intended recipient within a certain time, and, if there is no way of the item of correspondence doing so, it will not be sent.

Putting this together, is there something to be said for applying the journey time tolerance feature to mail? If so, ought it be the same as for passengers (which would be very easy), or ought it be adapted somehow? If so, how might it be adapted?

Another possible adjustment is simply to generate fewer units of mail per unit of passenger. Currently, for every four units generated by cities, three are passengers and one mail. For those who are interested, the line of code is:


const ware_besch_t *const wtyp = (simrand(400, "void stadt_t::step_passagiere() (mail or passengers?"))<300 ? warenbauer_t::passagiere : warenbauer_t::post;


I could change this 1:3 ratio to 1:4 or 1:5. However, if I do that, I wonder whether there will not be enough mail produced later in the game. It would be very useful to know what the real life ratio of mail to passengers is (and then adjust that to take into account our bags of mail measure and work out the number of items of mail in a bag, etc.).

I wonder also whether we need to reconsider the "bag of mail" measure: units of mail, currently measured in bags, are generated like individual passengers from city buildings and industries. However, in reality, although individual buildings and industries generate multiple trips of whole passengers, they do not, except for very large buildings, generate entire bags of mail. Even if they did generate entire bags of mail, the contents of that bag would not all be destined for the same place. Is there something to be said for replacing "bags" of mail with "items" of mail, of some small fraction of the weight of a bag, but also the same fraction of the revenue potential of a bag, and concomitantly increasing the capacity of mail carrying vehicles by 1/ that same fraction?

Another possible way of dealing with it is to have a time based mail demand .tab file. The reason for this is that the demand for mail has changed a great deal owing to social, economic and techincal factors outside the realm of transport (and therefore outside the realm of what can be simulated directly) in a way that passenger demand has not. The penny post in 1840 saw a vast surge in mail volumes, whereas the advent of e-mail from the late 1990s onwards has seen a precipitous decline in mail traffic. This would take a more detailed coding effort than changing the ratio of passengers to mail, applying the same journey time tolerance settings as for passengers to mail or changing the measure of mail from "bags" to "items".

I should be very interested in people's views as to which of these measures (and it might be that more than one at once is appropriate) is/are best for dealing with the mail balance issue currently seen in the Bridgewater-Brunel game.

Edit: Incidentally, I don't think that making a higher proportion of mail travel longer distances will help the situation: longer distance travel is more profitable, and this will not abate the excessive volumes.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Quote from: jamespetts on November 25, 2012, 10:48:24 AMI wonder also whether we need to reconsider the "bag of mail" measure... Is there something to be said for replacing "bags" of mail with "items" of mail, of some small fraction of the weight of a bag, but also the same fraction of the revenue potential of a bag, and concomitantly increasing the capacity of mail carrying vehicles by 1/ that same fraction?

Sounds good, but can the game handle such small fractional value? I mean, if we're simulating letters, then I expect a carriage to carry a few thousand of them at a few grams each... if it all rounds-down to 0 (and thus you get divide by zero errors) or rounds up too quickly it may cause issues.

jamespetts

#7
Yes, this is an issue: and the mail:passenger ratio might need increasing, rather than decreasing, if this is done. How about "bundles" of mail? An intermediate size between individual items and whole bags? Each "bundle" could be 1kg.

Edit: After discussion in the online game, it seems that the above idea won't work terribly well through issues with storage capacity of station buildings. We seemed to be getting about one full bag of mail per hour in the game in the 1790s in a moderately urbanised area, whereas in the mid 20th century, a similar area would probably generate only about 2 40% full bags of mail per day. Taking a working day to be 13 hours, that works out at 0.06 full bags of mail per hour. That suggests that, even in the high era of mail, the amount of mail generated should be about 6% of what it is now.

Mail volumes have also varied considerably over time: the invention of the penny post greatly increased volumes, whereas the invention of the telephone and then the internet greatly reduced volumes. This rather points towards the varying over time solution. To do this properly, though, I should need access to historical statistics of mail volumes generated. Does anyone have any idea where I can access this information?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

o_O

The USPS has some statistical data here, although it only goes back to 1886 and it is from the US
http://about.usps.com/who-we-are/postal-history/pieces-of-mail-since-1789.htm

To get actually useful data you would need to control for population and smooth out the economic fluctuations and wars. Just glancing at it looks like USPS mail per person actually increased until around 2000, and maybe it still is if you could control for the great recession.  This doesn't even account for UPS and fedex who handle mostly mid sized packages.  Including private companies and measuring the tonnage rather then units of mail I suspect a huge increase since 1990 or so.  For every letter someone doesn't send due to the internet they buy a package online and get a dozen junkmail letters. 

...the thought of news stand industry chains converting into junkmail factories in the modern era amuses me. 

jamespetts

Interesting indeed! I wonder whether there are any comparable British data?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

The other thing to bear in mind is business mail, if we end up merely estimating volumes. Any large office has/had a Post Room because of the volumes involved.

el_slapper

Quote from: AP on November 26, 2012, 08:10:33 AM
The other thing to bear in mind is business mail, if we end up merely estimating volumes. Any large office has/had a Post Room because of the volumes involved.

indeed.

The office building I'm right now in, faking work(not always faking, though), is 9-level plus ground level(not counting restaurant & car park underground). 1/4 of the ground level is used for the mail service. I think we are 30-40- people per level, & 3 people work full-time for internal & external mail.

Though I dion't know how much of mail is internal mail. I guess a lot, but I shall inquire.

jamespetts

Thank you - any information gratefully received.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

jamespetts

See here for an updated list of planned coding projects for Experimental, to which I have added a project about realistic mail and passenger numbers as well as realistic city populations, as I think that the three need to be considered together.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

jamespetts

A further thought experiment on mail volumes. Let us imagine a citizen of one of our Simutrans cities to see how many trips that he makes and how much mail that he produces on an average day. Let us imagine him in a generic mid-20th century setting, before the advent of e-mail, telex or fax. Let us say that he is a solicitor. He goes to work in the morning. That is one trip. He has a productive day in the office, and manages to write ten letters. He also sends a large parcel with a bundle of papers for a forthcoming case to counsel's chambers. He then goes home. At home, he pays his water bill (one more letter), writes to his cousin Jeremy (another letter), and sends a package to his great aunt Frida containing a birthday present. All in all, he has made two trips and generated twelve letters and two parcels. That would be enough to fill, at best, 1/8th of a mail bag, if the papers that he sent to counsel's chambers were on a weighty matter. That is a ratio of 2 trips to 1/8th of a mailbag, or 1/16th of a mailbag per trip.

But that is not an average day. That was a mail heavy day. Let us imagine the next day. The next day our solicitor is a busy chap. He gets to the office in the morning (one trip), and manages to knock off two short letters before having to make another trip for a conference in counsel's chambers to discuss the case in which he sent the papers the preceding day.  He is there for several hours. When he returns (another trip), he has time to write and send two short letters before going home (a fourth trip). When he gets home, he is too tired to write any letters, and goes straight to bed after dinner. That is four small letters for four trips. Assuming, ungenerously, that a mail bag can hold 100 letters, that is 1/100th of a mailbag per trip. The average of those two numbers comes out at 58. That would suggest, on that analysis, that there should be 58 passenger trips for every bag of mail generated. Even if we were to take a more generous view of mail than that and round down to, say, 1:32, that would still be a very different ratio of mail to passengers than we now have, at 1:3.

Do these figures seem to make sense on the face of them? Are there any statistics anywhere about the number of mailbags transported versus the number of passengers transported on some railway in the UK somewhere?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

The solicitor was not a very average person in the first half of the 20 century.


A table from the the Robbins Report (1963) shows in 1902 only 1 in 100 19 year olds would still receive eduction. Thus having a higher education. Expect that only those who were in school with 14 and older were writing more than a letter a year at all.

Since we mostly don't care for rural population in our simulation, you'd can divide those numbers by the percentage of people living in towns. (thus increasing it, sinc you can expect nos significant number of rural population to stay in school after 14)

ӔO

My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

AP

Don't forget,
Quote from: jamespetts on November 29, 2012, 01:12:57 AM
A further thought experiment on mail volumes. Let us imagine a citizen of one of our Simutrans cities to see how many trips that he makes and how much mail that he produces on an average day. Let us imagine him in a generic mid-20th century setting, before the advent of e-mail, telex or fax. Let us say that he is a solicitor.

Does not this solicitor's office have a typing pool? A team of women typing all his letters, thus increasing his output. Truly though solicitor is a special case, it's a highly educated job - letters have to be written in a more considered fashion, not merely churned out.

Compare this with companies where a lot of business mail comes from. E.g. one of my parents, in their much younger days, worked for the regional office of a utility company.This company would send out either monthly or quarterly bills to every household in the county. And there were seperate offices for Electric, Gas, Water, Sewerage, etc. And that, of course is merely billing, before any mail to make such a business actually keep working.

I suspect if a utility company employee wasn't sending out a good few dozen letters a day they probably were underperforming... and of course, every household was writing a cheque and posting it back.

kierongreen

If you want to simulate mail more realistically it might be worth having sorting offices (or for some time periods travelling post offices). Having spent one christmas at a sorting office can say the amount of effort required to ensure post gets to where it is meant to is huge. Could be done by forcing mail interchange stops to have special extension buildings. Oh and mail in the UK at present is transported around on 'Yorks' (small trolleys pushed often by hand) maximum weight 250kg, which are usually full of 48 stacked trays of mail each weighing up to 4kg I think.

paichtis

Quote from: kierongreen on November 29, 2012, 09:00:21 AM
If you want to simulate mail more realistically it might be worth having sorting offices (or for some time periods travelling post offices). Having spent one christmas at a sorting office can say the amount of effort required to ensure post gets to where it is meant to is huge. Could be done by forcing mail interchange stops to have special extension buildings. Oh and mail in the UK at present is transported around on 'Yorks' (small trolleys pushed often by hand) maximum weight 250kg, which are usually full of 48 stacked trays of mail each weighing up to 4kg I think.
Aren't mail stops kinda already sorting offices given the scale of the game ?

wlindley

This brings up a question and an observation:





jamespetts

Quote from: wlindley on November 29, 2012, 03:00:10 PM
This brings up a question and an observation:

  • Question: What vehicle parameters exactly does "becomes a traveling post office" change?

TPOs in the present version of Experimental increase the revenue for carrying mail if the distance travelled is great enough. [/list]
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

kierongreen

But in real life mail was sorted on travelling post offices while the train was in motion. Would mean a reduced loading time to simulate not having to sort before departure.

jamespetts

Hmm - the loading time does not simulate the sorting: it simulates the actual mechanical putting of mail bags onto the convoy. Loading time doesn't start until the convoy arrives, whereas sorting time can start as soon as the bags of mail arrive at the transfer stop.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

TygerFish

James - have you had any more input on implementing the mail time sensitivity?  I'd expect that it would be a positive thing.  Mail still should be much more common than passengers in early eras, but it almost seems like there's *too* much right now.  Even with a mail time tolerance 5x or 10x that of passenger time tolerance, that might be enough to simulate latent mail demand (e.g., overnight express service).

Another thing to consider with mail traffic is network capacity, especially in early eras.  I started a 0.9 source test game in 1800 and with a handful of cities connected, the mail traffic was almost overwhelming my infrastructure.  If canals limited traffic the same way roads do (as proposed for long-term) it would be completely overwhelming.  Another possibility to fix that would be some more high-capacity mail vehicles, which I think I'll try out either way (by cloning some of the freight boats to carry mail).

jamespetts

#25
According to my calculations here, it might well be the case that the passenger factor (which is inherently tied to mail generation) does not need altering at all, but that the number of passengers needs to be controlled instead by adjusting the proportion of local to long distance trips and the journey time tolerances.

On the assumption that recalibrating the number of passengers does not require an overall reduction in the passenger factor, the real issue for mail is the proportion of mail generated per passenger, since the idea for mail is that it is not subject to the journey time tolerance constraint. In game terms, the idea is for mail to be more lucrative and less fussy than passengers, but to be far less in amount with the result that, although in the early game when using only roads, mail is the most profitable traffic to carry, the market for mail will quickly become saturated, and later networks with fixed infrastructure, such as canals and railways, will not be able to sustain mail traffic alone, simultaneously giving players the opportunity for an easy early profitable start and forcing players to expand to other modes with more intricate engineering required before too long.

I should like people's views on this. Is there anything to be said for changing the mail:passenger ratio now from 1:3 to 1:32 or even 1:48 until I can implement a more sophisticated system of having mail demand vary over time?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

This seems like the easiest fix to achieve good results. I think we should try 1:32 and see how well that works.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

greenling

I Think that better it the price that you get for a Bag mail to reduce.
And to make the Devlievertime for mail longer.
I self alive it that´s the Postman first cam after 11:00 o clock.
Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

asaphxiix

sounds like a good plan, working with the numbers for now. I might have suggested a lower ratio (a bit more mail than 1:32, that is).

jamespetts

#29
In the passenger calibration thread, I wrote that we need to be working to a figure of about 1,250 potential passenger trips per head of population per year, constrained by journey time tolerances.

Turning to some actual data about mail, in 1992, there were about 54,000,000 letters posted a day according to "A History of Britain's Post" by Nance Fyson (ISBN 0-946003-99-8). Google tells us that the UK population was about 57.5 million at the time. That gives an average of 0.93 letters per person per day or 343.02 letters per person per year.

The current weight of a mail bag in the code is 50kg, but that is obviously far too big (the size of a small person), as mail bags were often easily carried by hand. This needs to be reduced by more than fourfold: according to this website of the people who make the current mail bags for the Royal Mail, their limit is 11kg. On the basis of that weight, about 550 first class letters per bag could be carried (based on the maximum permissible weight for a first class letter).

If we divide 343.02 by 550, we get 0.62, meaning that each head of population should generate 0.62 bags of mail at 11kg per year. If we assume a waking day of 16 hours (as we are doing for the passenger generation, different to the 13 hour day suggested above), each head of population would generate 0.000106164 bags of mail per game hour. A small town of 2,000 would generate 0.212328767 bags of mail per game hour (or take five game hours, nearly a game month at 250m/tile and 21 bits per month or 125m/tile and 22 bits per month) to generate one bag of mail. On another view, it would require a passenger to mail ratio of 2,016:1 if we assume 1,250 potential passenger trips per head of population per year.

On that basis, a town such as Slough with a population of around 140,000 would produce 14.86 bags of mail per game hour. This suggests a need rather to increase the size of towns, largely by increasing the level of larger buildings - this is in any case something that has been identified as necessary for passenger generation.

Currently, the mail carrying vehicles in the game are calibrated on the same basis as passenger carrying vehicles, especially as to their capacity. Unfortunately, finding real capacity data for mail carrying vehicles is not easy: the foremost Google search results for 'capacity "bags of mail"' are web pages about Transport Tycoon vehicles. There are some pages that give the mail capacities of sunken ocean liners, such as the RMS Empress of Ireland, which was carrying 318 bags of mail at the time of its sinking in 1914, or the Titanic, which according to this website, had 3,364 bags of mail and between 700 and 800 parcels when lost on her maiden voyage in 1912. This source suggests that the capsules of the London Pneumatic Dispatch Company could carry 35 bags of mail each. Sadly, these are unusual vehicles and it is not clear how these compare in size to ordinary mail vehicles, except that it is clear from the article that their capacity was higher than that of carts.

More useful information appears here about the Postal Special, the train that featured in the famous 1936 documentary the "Night Mail" (and the poem of the same name). There, it states that the whole train carried about 3,000 bags of mail. Page 276 of the same text records that each "R. P. O. car" would contain between 250-280 bags (suggesting that the train was up to 12 carriages long). Page 275 records that this was the largest of all special mail trains, running from London to West Scotland at nigh, although dropping off and collecting mail at many destinations and transfer stations along the way.

I wonder whether the modern size of 11kg per bag might even be too much for our bags. Older pictures seem to show smaller bags, not filled as much as the modern bags are when at their top filling line. It might be quite rare for postal bags to reach their maximum capacity. Even if we assume half that, at 5.5Kg/bag, that still gives us, however, only a very small amount of mail generation.

Thoughts on this situation would be much appreciated.

Edit: A somewhat unusual piece of research on mail bag weights: according to this somewhat angry forum thread, "Mr. Grumpy", who is apparently a postal worker, after his tirade against the original poster, states that he delivers 12 bags per day averaging 130Kg, or 10.83Kg/bag.

According to a plaque in an American city, a large seed company was receiving about 100 bags of mail per day in 1917.


According to this source about the now defunct underground narrow gauge mail only railway in London:

Quote
Jane Rendell: The Post Office railway runs under here. It was started in 1913, but halted by the First World
War, the tunnels were only completed by 1926, and opened in 1927. Thirty four new trains were added in
1981. The railway is six miles long and runs from Paddington to Whitechapel, with six sorting offices, plus
the main sorting office at Mount Pleasance. The trains are 27 foot long and consist of 4 containers, each
carrying 15 bags of letters in 6 bag panels. Trains run in each direction every four minutes at peak times,
moving 50,000 bags of mail per day. The GPO Headquarters is under the King Edward Building, where there
is a maze of tunnels in a figure of eight shape. These include two twin blind tunnels under Angel Street. Here
the shaft connections to the deep level post office tunnels stop abruptly. There were to be extensions to the
south east to Elephant & Castle/Waterloo and to Victoria.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

I think the mail to passenger ratio sounds well off - in particular companies and the busines of covernment of an empire will have generated vast amounds of mail. Considering just households for mail is not necessarily going to give the right answer.

--

Where is the passenger calibration thread, I can't find it...

Consider Travel to Work Areas - where 75% of the population work within that area (essentially catchment areas for peoples everyday journeys).
In 2011 the UK had 243.
In 1991 the UK had 314
In 1981 the UK had 334.

In the c18th , almost everyone worked within a short distance of their home - so essentially every village and town was its own TTWA - the uk will have had tens of thousands of them.

Is Simutrans 'supposed' to be simulating the journeys to work - within the TTWA -  is that where the 1250 trips pppy comes from? Or just the occasional journeys further afield? Those trips will be very short, and as the c18th progresses to the c20th and people commute further to work, do we expect the number of passenger journeys to remain flat rather than increase?


jamespetts

According to these statistics linked above, there 117 items of mail generated per head of population from about 1910 onwards (this had increased steadily since 1839) annually. This is broadly consistent with the reported figure of 343.02 in 1992.

The thread on calibrating the passenger factor is here. The source of the data for the number of trips per person per year is here. The figure given there is 1,100 - the figure of 1,250 is based on the premise that the number of actual passenger journeys in Simutrans will be less than the number of generated passenger packets because not all will be able to reach their destination within their journey time tolerance.

Certainly, trips to work must count: they are a very high proportion of all passenger journeys. The possibility of passengers walking to their destination is already accounted for in Simutrans, but only for really quite short distances and in quite a crude way.

For a more detailed discussion on a long-term plan to try to simulate the development of commuting more precisely, see this thread and the subsequent discussions (which suggest that the private car part of what is proposed may well not be viable).

The change in the volume of passenger numbers over the years is intended to be represented by the journey time tolerance feature, which, as will be seen in the passenger calibration thread, needs to be enhanced in order to give a realistic spread of tolerances.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

#32
The maximum capacity for trucks and vans can be easily worked out, but their maximum volume might be a bit more difficult.

Just consider maximum permissible combined weight, which is 38t combined for UK, then subtract the vehicle's weight.

Curb weight/Dry weight (typical)
Car/estate wagon prior to 80's: 600kg to 1200kg
Car/estate wagon from 80's up: 1000kg to 1500kg
Van: 1200kg to 2600kg
Box van/cube Truck: 2500kg to 8000kg
Large truck: 5000kg to 14000kg
Semi truck: 6800kg to 9100kg
Semi-truck (trailer): 5000kg to 8200kg


Loaded weight http://www.dft.gov.uk/vosa/repository/A%20Simplified%20Guide%20to%20Lorry%20Types%20and%20Weights.pdf
Car/estate wagon: 1500kg to 2000kg (usually no more than 600kg, or +40% of its own weight with passengers due to chasis limit)
Van: 2000kg to 3500kg (usually no more than 800kg with passengers due to chasis limit. Also 3500kg limit to light goods vehicle)
Box van/cube truck: up to 7500kg (limited by 7500kg for small HGV)
Large truck: up to 18000kg (limited 18000kg limit for dual axle HGV)
Semi-truck: up to 38000kg (limited by 38000kg limit for quad axle semi HGV)
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

jamespetts

Interesting information - thank you!

I wonder whether, in the light of the above, we need to return to an idea floated, and for the time being discarded, above: replacing "bags" of mail with "bundles" of mail. Each bundle could represent 10 or 50 letters or 1/10th or 1/50th of a bag and be 1kg each. We would have to adjust the revenue accordingly.

The reason for doing this is the difficulty of dealing with smaller units: one "bag" of mail may be filled by many different postboxes. If an entire small town/village produces on "bag" of mail or so per game month, how do we handle trickles of internal mail, or the differentiation of its location? "Bags", I think, are insufficiently fine grained to deal with the small quantities of mail that smaller towns produce/consume.

If we revise the figures on this basis, a 50 letter bundle (which works best with the minimum weight of 1kg) would give us 6.2 bundles of mail per head of population per year, or 0.001061644 per head of population per game hour. A small town of 2,000 would produce 2.12 bundles of mail per hour, or 13.58 per game month at 250m/tile and 21bpm or 125m/tile and 22bpm.

A 10 letter bundle would multiply the above figures by 5: 31 bundles of mail per head of population per year, 0.005308219 per head of population per game hour, 10.62 per game hour from a town of 2,000 and 67.95 per game month (at the above settings) for the same town. The weight breaks down with a 10 letter bundle, however, as it would have to be <1kg, which would not be recognised.

All the mail vehicle capacities would have to be adjusted accordingly, multiplying by 10 (from the base number of bags) for the 50 letter bundle or 50 for the 10 letter bundle. So, for example, a 60ft mail railway carriage from the 1930s, which carries around 250 bags of mail, would have to carry 2,500 bundles of mail at 50 letters or 12,500 bundles at 10 letters.

At 50 letter bundles, the ratio of passengers to mail becomes 1:202 mail to passenger ratio, or a 1:40 ratio with 10 letters per bundle. This is based on the 1992 figures: the 1920 figures are about 1/3rd of that, or 1:606.

In all events, the logistics of mail would be quite different to the logistics of passengers. Mail has no alternative destination, no journey time tolerance and cares not for comfort; mail behaves much like passengers behave in Standard, save that, when the numbers are adjusted, there will be much less of it. The "network effect" will be much greater for mail for this reason: the connexion of more nodes on one's network will increase the amount of mail to be carried exponentially. Trunk routes will require very large mail capacities (many large lorries or a few long trains), whereas small rural areas will suffice with a little bicycle (we could perhaps do with a motorcycle for later times). Mail should be more profitable per vehicle-full than passengers (perhaps by about 1.2x), but it should be much harder to get full vehicles in the first place, and there should be fewer of them.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

That sounds about right.

There are only 15 Class 325 trainsets that only need to run at night, which is a very small fleet.
Since each car only carries 12000kg of mail, that would be 48t per set and 720t total for the fleet.

The only catch with lower weight mail would be how to have storage buildings with that capacity, without breaking the player's bank.

even storing 12000 separate pieces of mail would require a huge station extension as it is now.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart: