Pak128 Airplanes Balancing

[Roboron]

Playing online with other players has made me realize how unbalanced airplanes are. We are on 1937 and a player is swimming in absurd amounts of money by running tons of small airplanes between major cities.



Of course playing well should be rewarded with profit, but profit from airplanes is just not balanced, (more so having into account that they barely incur into infrastructure costs). For comparison, profit margin for my boat service is ~13%, which is ok, as boats also do not need much infrastructure investment.



This aircraft is making double the money of its operational costs, and you can run many of these with little infrastructure. If this can happen in 1937 already, this can only get worse...



(The fact you can make runways so small also deserve a discussion on its own)

Can we come up with a better formula for airplane balancing in Pak128?

I think we should aim for a slower maximum margin for airplanes, specially since they make a lot of money for every vehicle. Something like ~15% if you are doing things optimal.

[prissi]

Honestly, from my point of view pak128 is very unbalanced, not only planes. Trucks are tiny but have almost not running costs. Track maintenance is so high, you almost cannot build trains at the start.

In later years, there were way too cheap ships (running costs 1ct), although I hope I caught them by now. Planes should cost 80% of their earnings when filled at the end of their speed limit and infrastructure maintenance should be higher than tracks.

One could try to use the balancing sheet from pak64 for the planes, which seems challenging but still doable. And one probably wants to have some of the scripts working to have the calculation automized. Then a similar script could run over the ships for each good category.

Actually there is the nice tool of An_dz:
https://forum.simutrans.com/index.php/topic,18192.0.html

I run it and remove the non-passenger/mail planes and did a quick script for cost calculation. Attached is the Excel although the interpolation of the speed_bous speed could be improved.

Unfortunately the export option is broken right now, so one would need to manually update some planes and see if these give reasonable values.

Ok, export works now two, here is a modified data file with also some more retire dates.

[makie]

I have used the translator for export .dat in csv files.
The pakset_info can do this.

Also the pakset_info can do suggestion for prices (example: rcost_makie)
This are also written to a downloadable file. Look at the end of the table.

I have update the .dat files with this simple program.
https://sourceforge.net/p/simutrans/code/HEAD/tree/tools/translator/script/parameter_update.php

After installing php call: php parameter_update.php
The path of the directory with the vehicle with the .dat is inside the coding.

Rules for calculation of pak128.german decriptet in:
https://pak128-german.de/inc/formeln.html
sorry is german language

This was the way i do the balancing in pak128.german

[prissi]

Since pak128 uses a high speedbonus, the pak128.german formula might not work that well, especially for planes. Anyway, fixed the excel sheet to use the interpolated speedbonus of the intro year.

EDIT: Updated XLS

[Roboron]

Ok, export works now two, here is a modified data file with also some more retire dates.

What did you do to get the export working? I have been trying to run datSheet, but first it failed because of missing libraries, so I had to compile it (I had to write my own CMakeLists.txt because Andz didn't let any makefile). Ok, now I have a self-compiled version but the program fails with:

Code Select Expand

File saving warning FDATOUT2:Could not create file for writing for object ...
Did you encounter this error too?

[prissi]

I made a fork on Github since linzip++ was not found only libzippp. That work when compiled under MSVC

[Andarix]

The very large winnings on airplanes are also likely to come from the speed bonus. If the speed significantly exceeds the current bonus speed, then very high winnings arise.

Other costs give me more to think about.

- the difference between locomotive prices and ship prices
- the flat monorail has higher maintenance than the elevated one
- maintenance of ports is 5 times that of canal ports

to name just a few


These are the reasons why I don't like pak128. The freedom of play is very limited. At least if you want to play with bankruptcy.

[prissi]

The airplanes make very much money because they travel so many tiles quickly. The money per tile for a fully loaded airplane is even lower than for ships.

pak128 is not well balanced, it is very much geared towards cars. Track maintenance is very expensive, rolling stock is very cheap. A complete overhaul would be indeed needed.

Anyway, I submitted new balancing for passenger and mail planes, passenger planes make minus when less than 50% load and mail when less than 40% (sinc email transport is more asymmetrical), all assuming little taxiing and no circling.

[Andarix]

That's the point. Set the speeds higher for Air in the speedbonus.tab and you will have significantly lower winnings without changing an aircraft.

[prissi]

The aircraft were not balanced evenly anyway, the revenue changed a lot between a factor of 2 and 5 (at most extreme). Now one could indeed alter the speedbonus a little for further challenge.

[Andarix]

I assume that the bonus speeds were changed but the operating costs of the aircraft were not checked.

Because the Boeings in 1960, despite being 100% loaded in both directions, are barely making a profit or are even in the red.

I can't judge to what extent this has to do with the slow game speed.

[prissi]

The speed bonus increases steeply in the 60ies, hence profitable aircraft may soon fall out of profit. The script is quite simple in this regard, depending on the exact year the speed may be set too tight.

Which airplane you are thinking of? 707?

[Andarix]

you can connect roborons server game

It's strange, now profits are going up.

Before, the server responded very slowly. Is it possible that something goes wrong during the calculation if there is an overload?