The International Simutrans Forum


Author Topic: Long distance revenue and average speed  (Read 1221 times)

0 Members and 1 Guest are viewing this topic.

Offline fbfree

  • *
  • Posts: 34
Long distance revenue and average speed
« on: October 23, 2011, 05:06:23 PM »
I've found a few oddities on long distance routes where more than a month passes between stops and less than one convoy passes per month:

1) The average speed history for the line is generally lower than the average speed of the convoys, and can be significantly lower than the speed of any of the convoys (even down to zero average speed).
2) Effect 1 is reduced as line frequency increases beyond 1 convoy/month.
3) The revenue on lines experiencing effect 1 is unusually low (or can even be zero when the distance between stops is large).  I expect that an average speed well below the speedbonus level is responsible.
4) Revenue on long distance routes returns to its expected behaviour when all the convoys run lineless.
5) Revenue/passenger seems to be correlated with the previous month's average speed of the line, not the current month's average speed.  (I'm not 100% certain that this effect isn't due to correlations between stop distance and the previous month's average speed rather than being a symptom.)

I've uploaded, to demonstrate these effects.

The (40) South-East line is a long distance line that should be profitable, but suffers from receiving no revenue for passengers and mail arriving or departing from Oakchester Dock.  When running all these convoys lineless, each convoy earns, and should earn, approximately $15,000 when arriving at Oakchester.
Note that I am running line (40) very close to a revenue cutoff; below 17km/h over a 300km run, passengers will provided no revenue, and mail will provide no revenue soon after.  The convoys are running between 19 and 20km/h.

The (15) Upper Peninsula line also experiences these effects.  However, since it has more closely spaced stops (still > 1 month apart for some route legs), the loss of revenue to due an incorrectly calculated speedbonus is not as disasterous as it is for line (40).  In an earlier save game where this line had only 4 convoys running, the average speed would occasionally drop to zero.

I'm using experimental 10.2, pak-brit-exp 0.8.2 with a modified speed bonus configuration and increased boat running costs.


Offline jamespetts

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 20194
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Long distance revenue and average speed
« Reply #1 on: October 23, 2011, 10:54:22 PM »

thank you for reporting that. I think that I have found and fixed the problem (on the 10.x branch), which was related to average speed recording as zero when lines had not had any vehicles arriving at any stops during a month. This is now fixed by copying the previous month's value. (Incidentally, it is intended that the previous months' value is used in revenue computations, as the current month's value might have been set by just one or two journeys, and might not be a true average). This fix should be available in the next release.

As to the speed bonus settings - might I ask why you used those particular numbers for the water? If you have found by testing that those balance better than the existing numbers, it might be worthwhile using those for the next release of Pak128.Britain-Ex.