How to Calculate Revenue

[akuenzi]

In another post I asked some questions as to how to set up train routes, or if there was a way to know in advance what kind of revenue a particular route would generate. Based on some of the responses I got, I started to wonder how revenue is calculated. Maybe somebody knows how it is done, but even after running a bunch of tests, I still can't come up with the formula. If someone knows what it is, please do share it!

From everything I've read in the strategy guide and seen in game play, revenue is impacted by the following eleven factors. The first eight are fairly easy to figure out, and the strategy guide gives clues as to how they impact revenue. The last three are much more difficult:

1. Initial Game Settings - We can boost revenue if we choose an easier setting.

2. Demand Level - A premium is paid for items in heavy demand. At a demand level of 0, revenue is whacked 50%. At a demand level of 9, revenue is increased 40%. And there are varying levels in between.

3. Station Improvements - Improvements can be made to stations to either increase revenue (such as building a saloon, which increases arrival passenger revenue by 10%) or preserve revenue opportunity (such as building a grain silo for a 50% reduction in pickup delays.

4. Managers - Some of the managers that can be hired will have an impact on revenue. For example, hiring George Pullman will increase passenger and station revenues by 25% and 5%, respectively.

5. Type of Economy - In a depression, the prices paid for transport services are less than in an expansion. In a depression, revenues get decreased by 20% from a normal economy, and in an expansion, they are 20% higher. And of course, there are varying levels in between.

6. Speed Record - If a train sets a speed record, the revenue for that load is increased 15%. I think this only impacts passenger and possibly mail service.

7. Certain Rail Cars - Certain rail cars impact revenue, such as including a dining car on a passenger run. The dining car will increase the route's passenger revenue by 20%.

8. Events - Scenario or campaign events can also impact revenues. Some (in the case of campaigns) may also be the choice of the player.

9. Cargo Type - We've all learned that each cargo type generates a different revenue stream. The strategy guide publishes 'base values' for each. For passengers, the base value is $60,000 per car, whereas it is only $30,000 for coal per car.

10. Time Cargo is in Transit - The time a cargo is en route to its destination impacts ultimate revenue received, particularly since each cargo type's revenue is impacted differently as to how long it's allowed to be in transit before revenue losses occur.

11. Distance Shipped - In general, the farther away something is shipped, the more revenue that will be received for it. Again, each cargo type calculates this differently.

I'm sure if I missed any factors, somebody will tell me. The first 8 are fairly easy, the last three are a booger to figure out, and I still don't know how the game calculates it. Someone smarter than me that frequents these pages may likely know the answer.

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

I came back to this post -- I think I forgot a couple vital items that impact revenue:

12. Amount of Cargo -- The more cargo shipped in a particular load, the greater the revenue.

13. Track Ownership -- A pro rata portion of revenue goes to the owner of the track over which the train travels to reach its destination. If the station of origin and destination are owned by another player, there are also station fees to pay in addition to the pro-rata track charge. I'm not sure if these are reversed from revenue or if they are treated as an expense.

14. Game Date -- The later it is in the game, the less revenue that is earned for each cell traveled.

I wonder if it is also impacted by the global settings on the map in terms of economic growth rates. This might be an indirect factor that has more of an effect on 'production' versus 'revenue,' but I'm not sure on that one. Revenue seems to grow even more complicated by the day!

[akuenzi]

In running some tests, I created a special game board that was perfectly flat, and had one length of track along the bottom. At ten cell intervals I set up new towns and stations as test destinations. At the beginning, I made a huge city so there would be plenty of passengers and mail to carry, and set up a few industries. I wanted to try to isolate the impact of the cargo type from every other influence. So, I set up several events that constantly checked the economic status to ensure it stayed with a 'normal' economy. When I started the game, I put it on the basic settings, with no revenue impact, so it wouldn't be impacted by demand levels. I also made sure there wasn't a manager on board that would impact my revenue stream. The stations had only the basic improvements in them to keep the trains running (ie, water, sand, and roundhouse), nothing that would impact revenue. For every test, I restarted the scenario so that everything had the same baseline. Only one car was hauled for each trip. The scenario was dated at 1900, and I think I used the 2-8-0 as my engine.

If anybody wants the test map, just say the word and I'll post it here.

[akuenzi]

Things I Learned:

The strategy guide has each cargo type listed out, and has three interesting columns titled 'base price, days to deliver, and ship distance.' The insert that came with the original game had a fourth item of interest, 'time & distance sensitivity.'

1. There is apparently a certain minimum distance that the game feels each cargo should be hauled before the 'base value' is earned in full. For example, the base value for passengers is apparently $60K per car. Using my test map, I had to haul passengers just over 50 cells in order for the full $60K to be earned. Of course, by then the revenue was being modified downward by other factors, so the theoretical $60K probably occurred prior to then, though I'm not sure just where. I tested also the following cargo types:

-- Oil ($40,000 base value) - earned at 30 cells of travel

-- Wool ($50,000 base value) - earned at 50 cells of travel

-- Coal ($30,000 base value) - earned at just under 30 cells of travel

-- Mail ($70,000 base value) - earned at about 55 cells of travel

-- Milk ($60,000 base value) - earned at 50 cells of travel

-- Lumber ($30,000 base value) - earned at just under 20 cells of travel

The point is, we can't just go off of the listed base values in seeing which cargo is the most valuable to haul. We have to also know how FAR we have to haul it just to earn the base value, and it obviously varies.

2. At first I thought the 'days to deliver' column said how long the player had to deliver each cargo type before penalty. As best I can determine at this point, this column only indicates how long the cargo will sit around in the station waiting for pickup. It doesn't appear to have any bearing on revenue, except in general terms (ie, the greater the days to deliver, usually this means the base cargo value is less). What I don't know is exactly how station improvements might impact this. Does the presence of a grain silo, for example, mean that grain will be held in the station for longer than the 1,600 days (50% longer) without disappearing?

3. Ship distance has four values. Passengers are represented with a 0.37, mail with 0.47, and the rest by 0.1 or 0.2. The strategy guide says that the greater the ship distance factor, the more that is received for shipping cargo long distances. As I learned on another thread, the cargo with a ship factor of 0.1 should generally NOT be hauled long distance. It's a wash at best, and sometimes you can lose money. Those with a ship factor of 0.2 can be hauled longer distances profitably, but we need to even be careful with these. The biggest impact of distance happens with mail and passengers, as most of us alread know. However, for the life of me, I can't figure out how these 'factors' impact the revenue formula.

4. I think the REAL determiner of how much revenue is received is the little column included with the insert in the original game. This is where all the car types are printed, and each is given a factor as to how sensitive they are as to time and distance. Mail is the most sensitive, and has a factor of 10. Passengers are at 8. Things like grain and logs are down around 2. And so forth. I think these factors are what impacts the coins that first appear on the train when it is loaded up. The program calculates a theoretical revenue based on the distance to the station of delivery, and it puts these coins on the train. Then, it reduces these coins all the way to the train's destination based on this time and distance factor. Being more sensitive, the mail's revenue is going to decline much more rapidly over the trip than would the revenue of grain. But just HOW it decreases (or even calculates the initial) revenue over this distance/period, I don't know.

5. While theoretical revenue decreases over the train trip to destination, where it becomes actual revenue, I'm not sure that it would ever reach zero. Initially, in following a number of trains over their route, revenue drops off steadily over the route. However, given enough time, it eventually reaches a point where it doesn't decline nearly as fast. It's almost like the decline hits a brick wall. Most of us probably wouldn't ever hit this brick wall unless we have an extremely long route with some of the earliest trains. Using the 2-8-0, I had to put its throttle on the lowest setting to get it to slow down enough to try to run revenue down to zero. And I still couldn't do it.

The longest test I did was to take a single carload of lumber 200 cells going just above 0 mph. Starting in Janyary 1900, it took until January 2011 for the cargo to arrive. That's right, over a century! And by the time it finally arrived, the cargo had a value of $19K (down from its theoretical value at the beginning of $136K). Somewhere in the middle, I noticed that when the theoretical value reached $40K, it decreased very slowly after that.

One would think that this might be consistent, but it's not. I tried this same procedure on milk. I didn't have the patience to take it to its destination, but starting with a theoretical revenue of $205K for a 200 cell run, it dropped until it reached $82K, and then declined much more slowly. The same occurred with mail. I only did it for a 100 cell run there, which had theoretical revenue of $177K. After hitting $67K, it dropped off more slowly. You can do the math and see that it doesn't drop off at the same rate between these.

6. I didn't run every test that could be run, but I did enough to get a picture as to what Jeff and Drewsey were telling me in another topic. The "0.1" cargoes earn additional revenue much more slowly over longer distances, and therefore it's best to haul them only over shorter distances. You can see the data set later, but in the case of lumber, for example, once you reach base revenue, you add additional revenue of $5K for each additional ten cells traveled. An expensive engine would easily eat up this additional revenue and produce the 'wash' he was talking about on the other page. This is contrasted with something like mail, where even after the base is reached, there is quite a bit more revenue added over ten cells ($13K between the 90 and 100th cells of a run).

7. And finally, I learned that I still don't know how to calculate revenue, though I do get somewhat of a picture as to what's going on. I'll put some data in another post below. It appears that everything is calculated on some curve using the aforementioned factors.

[akuenzi]

My humble data set, such as it is, is in a spreadsheet attached below. I did try to attach a spreadsheet, but it said I couldn't upload that type of file, so I put it in a pdf instead. I also just realized that the passenger data is skewed. In restarting the scenario each time to run the test, each train set a speed record, so the revenue data should probably be reduced 15% to take this into account.

My humble apologies if I've bored anyone to tears. I didn't do every test for every cargo, so it isn't much of a data set. Much more could have been done, but I still don't know that it would have given a conclusive answer as to how revenue is calculated.

What I'm secretly hoping is that a smarter person will come by this and say, "You dummy -- the revenue formula is: ____________" Does anyone know what it is? Jeff is sure going to need it for when he starts to program Open Railroad Tycoon. :) (It sure was easy to volunteer you for that!)

Revenue_Data_Tests.pdf

[Drewski]

Things I Learned:

The strategy guide has each cargo type listed out, and has three interesting columns titled 'base price, days to deliver, and ship distance.' The insert that came with the original game had a fourth item of interest, 'time & distance sensitivity.'

1. There is apparently a certain minimum distance that the game feels each cargo should be hauled before the 'base value' is earned in full. For example, the base value for passengers is apparently $60K per car. Using my test map, I had to haul passengers just over 50 cells in order for the full $60K to be earned. Of course, by then the revenue was being modified downward by other factors, so the theoretical $60K probably occurred prior to then, though I'm not sure just where. I tested also the following cargo types:

-- Oil ($40,000 base value) - earned at 30 cells of travel

-- Wool ($50,000 base value) - earned at 50 cells of travel

-- Coal ($30,000 base value) - earned at just under 30 cells of travel

-- Mail ($70,000 base value) - earned at about 55 cells of travel

-- Milk ($60,000 base value) - earned at 50 cells of travel

-- Lumber ($30,000 base value) - earned at just under 20 cells of travel

The point is, we can't just go off of the listed base values in seeing which cargo is the most valuable to haul. We have to also know how FAR we have to haul it just to earn the base value, and it obviously varies.

2. At first I thought the 'days to deliver' column said how long the player had to deliver each cargo type before penalty. As best I can determine at this point, this column only indicates how long the cargo will sit around in the station waiting for pickup. It doesn't appear to have any bearing on revenue, except in general terms (ie, the greater the days to deliver, usually this means the base cargo value is less). What I don't know is exactly how station improvements might impact this. Does the presence of a grain silo, for example, mean that grain will be held in the station for longer than the 1,600 days (50% longer) without disappearing?

3. Ship distance has four values. Passengers are represented with a 0.37, mail with 0.47, and the rest by 0.1 or 0.2. The strategy guide says that the greater the ship distance factor, the more that is received for shipping cargo long distances. As I learned on another thread, the cargo with a ship factor of 0.1 should generally NOT be hauled long distance. It's a wash at best, and sometimes you can lose money. Those with a ship factor of 0.2 can be hauled longer distances profitably, but we need to even be careful with these. The biggest impact of distance happens with mail and passengers, as most of us alread know. However, for the life of me, I can't figure out how these 'factors' impact the revenue formula.

4. I think the REAL determiner of how much revenue is received is the little column included with the insert in the original game. This is where all the car types are printed, and each is given a factor as to how sensitive they are as to time and distance. Mail is the most sensitive, and has a factor of 10. Passengers are at 8. Things like grain and logs are down around 2. And so forth. I think these factors are what impacts the coins that first appear on the train when it is loaded up. The program calculates a theoretical revenue based on the distance to the station of delivery, and it puts these coins on the train. Then, it reduces these coins all the way to the train's destination based on this time and distance factor. Being more sensitive, the mail's revenue is going to decline much more rapidly over the trip than would the revenue of grain. But just HOW it decreases (or even calculates the initial) revenue over this distance/period, I don't know.

5. While theoretical revenue decreases over the train trip to destination, where it becomes actual revenue, I'm not sure that it would ever reach zero. Initially, in following a number of trains over their route, revenue drops off steadily over the route. However, given enough time, it eventually reaches a point where it doesn't decline nearly as fast. It's almost like the decline hits a brick wall. Most of us probably wouldn't ever hit this brick wall unless we have an extremely long route with some of the earliest trains. Using the 2-8-0, I had to put its throttle on the lowest setting to get it to slow down enough to try to run revenue down to zero. And I still couldn't do it.

The longest test I did was to take a single carload of lumber 200 cells going just above 0 mph. Starting in Janyary 1900, it took until January 2011 for the cargo to arrive. That's right, over a century! And by the time it finally arrived, the cargo had a value of $19K (down from its theoretical value at the beginning of $136K). Somewhere in the middle, I noticed that when the theoretical value reached $40K, it decreased very slowly after that.

One would think that this might be consistent, but it's not. I tried this same procedure on milk. I didn't have the patience to take it to its destination, but starting with a theoretical revenue of $205K for a 200 cell run, it dropped until it reached $82K, and then declined much more slowly. The same occurred with mail. I only did it for a 100 cell run there, which had theoretical revenue of $177K. After hitting $67K, it dropped off more slowly. You can do the math and see that it doesn't drop off at the same rate between these.

6. I didn't run every test that could be run, but I did enough to get a picture as to what Jeff and Drewsey were telling me in another topic. The "0.1" cargoes earn additional revenue much more slowly over longer distances, and therefore it's best to haul them only over shorter distances. You can see the data set later, but in the case of lumber, for example, once you reach base revenue, you add additional revenue of $5K for each additional ten cells traveled. An expensive engine would easily eat up this additional revenue and produce the 'wash' he was talking about on the other page. This is contrasted with something like mail, where even after the base is reached, there is quite a bit more revenue added over ten cells ($13K between the 90 and 100th cells of a run).

7. And finally, I learned that I still don't know how to calculate revenue, though I do get somewhat of a picture as to what's going on. I'll put some data in another post below. It appears that everything is calculated on some curve using the aforementioned factors.

Nice work!

A few comments:-

1) That sounds just about right.

2) Yes the buildings actually have the effect of making the cargo hang around longer. An early Post Office is a must, as it in effect gives you 50% more post to deliver...and since Post tends to arrive in stations not steadily but in clumps, there's a good reason for dedicated mail trains...(mail also pays better than anything else)

3 & 4) Yes. The game calls this a ROT factor (it's in the strat guide appendix). The formula depends on economy model, economy lvl, cargo type, demand lvl, distance, ROT factor. There may be other variables. Hitting upon the exact formula by accident isn't going to happen. N.B. Goods don't rot when in a station's storage (after being dropped off for a later train)

5,6,7) It's as you said. Hauling Coal and Iron for instance, over a long distance isn't very profitable. It's better to haul them on cheap trains to a local steel mill, then ship the steel with more powerful locos to where it's needed. I think that from experience, a player can look at a map, and very roughly estimate incomes (taking a normal economy lvl as a norm).

-------------------

One of my pet peeves, is that goodwill doesn't seem to effect revenue.

Take these scenarios:- a railroad is serving cattle country (represented by one ranch on the map). But the train picking up from there, let's say also picks up grain from farms elsewhere, and takes it to a bakery. The cattle don't get to market as they should, in fact many stand around grazing in the fields. Yet the railroad possibly profits from this situation, by keeping the market price high (via the demand model). Wouldn't a rancher pay more for a better service, where his cattle were picked up on a regular basis?

Another scenario:- a triangle of three towns have a pax/mail service but it's so irregular, that no-one really knows when the trains are supposed to turn up ("When's the train arrive Stationmaster?...."Probably tuesday, maybe wednesday, might be thursday") Again the railroad keeps the demand high, but doesn't suffer from a bad service rating. Wouldn't folks be prepared to pay a little more for a reliable service?

I'm thinking of using the event manager to represent this. E.g. Goodwil = Normal no bonus, then maybe a sliding scale of -10 to +10% on various/all cargo, depending on the goodwill lvl....not sure if this is a good idea or not...would probably have to ban the goodwill boosting managers. What d'ya think?

[akuenzi]

Nice work!

Thanks! And that's about what it started to feel like -- WORK. :) I told my wife as we were getting into bed last night that I felt like I'd just turned in a college term paper.

2) Yes the buildings actually have the effect of making the cargo hang around longer. An early Post Office is a must, as it in effect gives you 50% more post to deliver...and since Post tends to arrive in stations not steadily but in clumps, there's a good reason for dedicated mail trains...(mail also pays better than anything else)

Mail does have the highest base value, but it sure does decline rapidly over a short period of time. It also isn't as plentiful as a cargo compared to some of the others. Most of the time I've kind of treated it like an afterthought, maybe adding a single car to a couple passenger cars. I didn't know it got to the station in 'clumps.' Maybe I've been missing out on something!

3 & 4) Yes. The game calls this a ROT factor (it's in the strat guide appendix). The formula depends on economy model, economy lvl, cargo type, demand lvl, distance, ROT factor. There may be other variables. Hitting upon the exact formula by accident isn't going to happen. N.B. Goods don't rot when in a station's storage (after being dropped off for a later train)

I've heard that term before (ROT factor), and I plan on checking it out some more in the strategy guide when I go home. I couldn't seem to find much on it last night, but maybe it was getting too late. I didn't know it was in the appendix, but thanks!

One of my pet peeves, is that goodwill doesn't seem to effect revenue.

Take these scenarios:- a railroad is serving cattle country (represented by one ranch on the map). But the train picking up from there, let's say also picks up grain from farms elsewhere, and takes it to a bakery. The cattle don't get to market as they should, in fact many stand around grazing in the fields. Yet the railroad possibly profits from this situation, by keeping the market price high (via the demand model). Wouldn't a rancher pay more for a better service, where his cattle were picked up on a regular basis?

Another scenario:- a triangle of three towns have a pax/mail service but it's so irregular, that no-one really knows when the trains are supposed to turn up ("When's the train arrive Stationmaster?...."Probably tuesday, maybe wednesday, might be thursday") Again the railroad keeps the demand high, but doesn't suffer from a bad service rating. Wouldn't folks be prepared to pay a little more for a reliable service?

I'm thinking of using the event manager to represent this. E.g. Goodwil = Normal no bonus, then maybe a sliding scale of -10 to +10% on various/all cargo, depending on the goodwill lvl....not sure if this is a good idea or not...would probably have to ban the goodwill boosting managers. What d'ya think?

I didn't even think about goodwill. What you're saying makes sense to me, and an event could easily incorporate it into the equation. I'm not sure you'd need to ban the goodwill boosting managers, though. In fact, having an event like this could make those managers all the more valuable to the railroad, and could add a new aspect to game play. Right now I tend to ignore the managers that only impact goodwill, because there always seems to be other managers that have a better impact to the business. If you put in this event, it would make the goodwill managers more valuable, especially if they can impact revenue in a positive way!

[Drewski]

I didn't even think about goodwill. What you're saying makes sense to me, and an event could easily incorporate it into the equation. I'm not sure you'd need to ban the goodwill boosting managers, though. In fact, having an event like this could make those managers all the more valuable to the railroad, and could add a new aspect to game play. Right now I tend to ignore the managers that only impact goodwill, because there always seems to be other managers that have a better impact to the business. If you put in this event, it would make the goodwill managers more valuable, especially if they can impact revenue in a positive way!

Good points...that vaguely crossed my mind, but I thought they might become a little too useful....maybe not....I shall peruse the matter more while doing grocery shopping in a moment (and hopefully not forget items as my mind is on gaming matters as usual) :D

[jeffryfisher]

The revenue formula is probably three or four formulas added together (even before bonus multipliers). Some parts might fall to zero at thresholds like some minimum distance and some maximum travel time. Each subformula must be protected against exploding (division by zero) at the other end as well, so any part that seems to divide by time probably divides by time+something.

There may be more than one time component: Besides travel time, the delay in pick up may matter for more than just having a load expire. Having a waiting train pick up a load as it's created may pay better than picking up the same load just before it rots. The two times (time to pick-up and time in transit) may simply be added together to make one time factor, but I wouldn't count on it. Each may be treated differently, so a fast delivery of a late pick-up might pay better or worse than a slow delivery of an instant pick-up, even if the time-sum was equal for each case. In your tests, did you control for pick-up delay?

Drop-shipments might not evaporate away, but they do seem to age (go down in value) as if on board a very slow train. If drop-shipments pile up faster than you can haul them away, you might wish that the oldest loads would rot.

I wonder if goodwill affects production rates for original sources (if more passengers show up if goodwill is good). There must be some tie between train crashes, robberies and passenger production rates. It might be goodwill, or it might be some independent factor.

[Drewski]

I wonder if goodwill affects production rates for original sources (if more passengers show up if goodwill is good). There must be some tie between train crashes, robberies and passenger production rates. It might be goodwill, or it might be some independent factor.

While this would seem logical, experience tells me it's not the case. In extreme micromanagement games (e.g. finger permanently above pause, every time a bell rings, stop the game, and check what the train needs to pick up), after a few runs, I've had perfect balance on many routes. For example, a train is scheduled to pick up 2 pax from a station, and when I pause the game on train arrival, that's time and time again exactly what is there. No more no less.

This is despite the odd crash and robbery (and of course relies on an economy staying at the same lvl for a while, and the train in question not getting blocked much or breaking down)

[jeffryfisher]

Pax production stayed level after a crashed train? That hasn't been my experience. All passengers vanish nationwide for about two years. With a large RR having hundreds of trains, one can crash about that often, effectively shutting off all pax permanently. The effect is so devastating that I simply reload from the last save whenever a train crashes (a breakdown I can swallow, but the crashes make the game unplayable with a big company).

[RSimpkinuk57]

Jeffry: Passengers stop appearing after a train crashes? I've had crashes but never seen that one.

Akuenzi, some more things to think about (by the way, some of this is from my experience with the original game; I don't know if TSC/Platinum is ever different)

.

(1) When an unloading train tells you how much each car-load has earned, this does not include all station and station-buildings revenue for passengers. I proved this when trying to work out whether buildings earn money from departing passengers, arriving passengers, or both (I did discover that, but have forgotten the answer). You need to watch how much your company cash goes up. That may mean going to the Income (and maintenance) page of the company accounts.

(2) I'd like to think that not-collected-immediately cargo (including passengers) loses no revenue if eventually unloaded at paying destination within days-to-deliver of original appearance at departure station (but could be wrong). So long as it doesn't rot away completely before collection, that is. I seem to remember that time-to-collect is less than time-to-deliver. If you do exceed time then maybe penalties are exacted at different per-day rates for days uncollected (penalty reduced by post office/silo/whatever) and days on train/at drop-station. (Obviously red-flag drops are meant to represent loaded cars sitting in a marshalling yard.)

(3) Distance for revenue calculations is straight-line from loading to unloading station, not distance travelled. Which is nasty if you need a 60-cell route round the ocean inlet or no-rights territory to connect cities only 20 cells apart.

(4) Foreign mileage deductions can be cheated. Example: A to C is 40 cells of which the first 16 are foreign track: that is 40%. Now try calling at B, 20 cells from A, where nothing unloads. You "pay" 80% of nothing to the other company. Now when you unload at C you have come from B, entirely on your own track - but the cars loaded at A still earn according to the full A-C distance (they'd better!)

(5) What does "increase by 20%" mean (for example) if there is more than one effect? Do two +20%s make a +40%, or do they make +44 (20% of 120 being 24)?

regards, Richard

[akuenzi]

Akuenzi, some more things to think about (by the way, some of this is from my experience with the original game; I don't know if TSC/Platinum is ever different)

(1) When an unloading train tells you how much each car-load has earned, this does not include all station and station-buildings revenue for passengers. I proved this when trying to work out whether buildings earn money from departing passengers, arriving passengers, or both (I did discover that, but have forgotten the answer). You need to watch how much your company cash goes up. That may mean going to the Income (and maintenance) page of the company accounts.

You're probably right. I didn't even think that far ahead. In the revenue tests I've been running, I set up a train destination while on pause, then I turn it loose, repeatedly 'right clicking' on the train as it leaves the station of origination. It soon gives out a 'theoretical' revenue which slowly rots away as the train travels. Glad to know it doesn't include the impact of station improvements, as I'm trying to isolate how revenue is calculated with distance traveled.

(2) I'd like to think that not-collected-immediately cargo (including passengers) loses no revenue if eventually unloaded at paying destination within days-to-deliver of original appearance at departure station (but could be wrong). So long as it doesn't rot away completely before collection, that is. I seem to remember that time-to-collect is less than time-to-deliver. If you do exceed time then maybe penalties are exacted at different per-day rates for days uncollected (penalty reduced by post office/silo/whatever) and days on train/at drop-station. (Obviously red-flag drops are meant to represent loaded cars sitting in a marshalling yard.)

I have yet to figure that out myself... and at the rate I'm going, I'm not sure I ever will. :) Based on what others have said, I think you're right. Once the cargo is on the train, the formula changes. Frankly, we refer to this as 'rot,' but I'm not sure that's the correct term. It would seem to be a revenue adjustment. The faster we get it to the destination, the more revenue we make. If we take too long to get it there, our revenue is penalized. In real life, if we ship something 'overnight,' we pay a premium, say to FedEx. If we drop it in the mail, it will take longer to arrive and the fee for it is lower. In both cases, the item shipped gets there intact (ie, not 'rotted'), but the fee structure is different. But maybe it does depend on the cargo, depending on the era. In the early days, maybe there was a lot of 'rot' or spoilage when produce was shipped, until this was mitigated with refrigerated cars.

(3) Distance for revenue calculations is straight-line from loading to unloading station, not distance travelled. Which is nasty if you need a 60-cell route round the ocean inlet or no-rights territory to connect cities only 20 cells apart.

Yeah, I hear you there. I'll have to run some tests on this at some point. All of the tests I've done on revenue so far are in a simple straight line. I haven't done any on the 'diagonal' to see how it works.

(4) Foreign mileage deductions can be cheated. Example: A to C is 40 cells of which the first 16 are foreign track: that is 40%. Now try calling at B, 20 cells from A, where nothing unloads. You "pay" 80% of nothing to the other company. Now when you unload at C you have come from B, entirely on your own track - but the cars loaded at A still earn according to the full A-C distance (they'd better!)

Thanks for the tip -- I never thought to do that! Then again, I haven't faced that situation either. Maybe the day will come when I'm forced to use a competitor's tracks, but up to now, I'd just build my own track so I didn't have to run on their two-bit railroad. Or even better, I'd take them over, and then force them to run trains on my track, so I get the revenue while they pay the cost. :)

(5) What does "increase by 20%" mean (for example) if there is more than one effect? Do two +20%s make a +40%, or do they make +44 (20% of 120 being 24)?

regards, Richard

I don't know the answer to that, either. Maybe it's as Jeff stated above, with the revenue calculation likely being several formulas. I'm sure the formulas for it would appear relatively simple, or at least understandable, if we knew what they were. But they aren't... and the amount of head scratching I've done trying to figure this out is making me go bald all the more quickly!

[RSimpkinuk57]

When people here write about "rot", I've always taken that to include not just loss of revenue but also the way that, if you leave cargo uncollected for long enough, it won't be there to collect any more (most noticeable with erratically-arriving mail). ..Richard

[jeffryfisher]

I think that rot is just one specific thing, not multiple things. I sometimes wonder if it is an acronym, maybe "return on time". I still am unsure about how rot and "days to deliver" work together. I think that rot affects revenue for slow vs fast trains, while "days to deliver" is actually "days to pickup" before a load vanishes from a station.

[akuenzi]

(3) Distance for revenue calculations is straight-line from loading to unloading station, not distance travelled. Which is nasty if you need a 60-cell route round the ocean inlet or no-rights territory to connect cities only 20 cells apart.

I did some tests to see how revenue on the 'diagonal' is calculated versus straight stretches. I know it will vary depending on the specific cell locations of the two stations, but I made one small test with a '45 degree angle' line. In other words, the rail line went up one cell, over one cell, and so forth. Like everyone has observed, more revenue is earned when traveling diagonally across a cell versus straight, but I wanted to see if the Pythagorean theorem played a part. Maybe others already had this figured out, but I learned today that the Pythagorean theorem most definitely is involved in calculating the distance (can't imagine what else they would have used :), but at least now my curiosity has been satisfied), with one tiny catch: The distance in cells, plus one, needs to be rounded down to the nearest whole number. I'm not sure about the 'plus one' part, but I'm never exactly sure just where a train has traveled one whole cell versus half a cell.

Here's an example. Let's suppose that a train is delivering product from station A to station B, and that station B is nine cells up, and nine cells to the right, relative to station A. Using the pythagorean theorem, the distance of the triangle's hypotenuse (plus one) is 13.73 cells of travel. The game looks at this as 13 cells rather than 14 cells. The amount of revenue earned for this run was identical to a straight 13 cell run.

It seems curious to me that it would do this. I did this test with only one commodity (lumber) and at one point in time (1820), but it was consistent through the run, and I tried it for 30 separate stations on this route. The revenue for the rounded-down hypotenuse length was identical to the same number of cells on an ordinary straight run.

Have I just confused everyone? I'm not explaining this very well, but maybe that's because I'm an accountant and not a mathemetician. Suffice it to say, revenue on the 'diagonal' works off the Pythagorean theorem! I would post my data, but this site doesn't seem to allow postings of Excel spreadsheets.

[akuenzi]

In post # 3 above, I mentioned, "There is apparently a certain minimum distance that the game feels each cargo should be hauled before the 'base value' is earned in full." I didn't realize at the time that it's not that simple... much to my consternation... and much to my now carpal-tunnel-infested wrists after running literally hundreds of tests. For whatever reason, I chose lumber for my tests. Not sure why -- maybe it's because I live in the Pacific Northwest. I now have to add one more factor to the calculation of revenue, which is simply game date. Evidently there is more revenue to be earned hauling cargo to a station 20 cells away in 1800 versus 1900 versus 2000. In addition, the tests showed me that the amount of revenue added for each additional cell travelled is inconsistent. There might be $2K added for each cell at a certain point, or maybe $5K, or some other number. There is a noticable pattern, and it does give some clue as to the 'cutoff point' beyond which further transport of the cargo might not be wise. Some have mentioned here that certain commodities they don't haul long distances. The data I'm seeing bears this out, and for lumber, the magic number is around 21 cells of travel. After that point, the revenue piled on much more slowly, particularly in the later years.

If anyone is interested, I'll post two files. The first is a graph that shows revenue earned for transporting lumber 21 cells. In 1800 you would earn $70K for this trip. In 1900 you'd be at $35K. The good news is that it drops off much more slowly than that thereafter, and you're still north of $30K in 2000. The second file is the raw data, which was a lot of 'fun' to come by. I made every attempt to shut off anything and everything that would skew the underlying data engine. In other words, no station improvements or anything like that. The data is based on 'theoretical' revenue. This means that I'd start the locomotive off heading X cells away. As it was loading and pulling out of the station, I'd right click the engine on the main screen to see what 'coins' (or revenue) it would theoretically earn. This was the value I put in the spreadsheet. The coins would decrease en route, of course, but this gives the baseline. Then I'd hit 'restart' and do it again. This way every test had the same starting point.

A word on the raw data. I'm not sure exactly how to calculate the number of cells travelled. You'll see a 0 and 1 cell at the top, and both have a value of zero. Of course, the train can't earn any revenue just sitting there, but watching it on the screen, it appears to need to travel at least 2 cells in order to earn anything at all. I think it starts in the middle of a cell in one station and ends in the middle of a cell in the second station for the minimum distance. Maybe that's just 1 cell. Well, if you don't like it, then just shoot me. :) All data is in thousands, and I did every test in 5 year increments. And brother does my finger hurt from pressing the right mouse button so many times! The tests for all years go through about 30 cells of travel, except for the one in 1800, which I took out to 60 cells. It's about 1,335 data points. Unfortunately, the site doesn't appear to allow Excel spreadsheets, so I had to dump all this into a pdf file as well. The graph that is attached is based on line # 21 in the data set.

I won't post my test map, but if anyone wants to play with it, just say the word and I'll dump it on here.

Where to go from here? I wish I could determine the formula from all this. Maybe there is a way to do it if I play with it long enough, but I'm likely not smart enough to do it. I'm running some predictive tests on other commodities using lumber data with mixed results. We'll see what happens. If I had half a brain, I'd quit all this and simply play the game.

-----------------------------------------------------------------------------------------------------------------

Jeffry -- Thanks for the suggestion. I'm sure you've made it in a couple other areas and I just forgot. I redid the upload, so now you can have the raw Excel file with the data for lumber. It also includes a couple hidden columns which had the tests I did for the revenue calculations 'on the diagonal.'

Theoretical_Revenue_21_Cells.pdf

Lumber_Data.zip

[jeffryfisher]

the site doesn't appear to allow Excel spreadsheets, so I had to dump all this into a pdf file...

You may also dump it into a zip file, which would be easier for others to extract. In fact, you can dump multiple files into a single zip file for simpler upload/download ops.

Good work!

[leechboi]

Anyway I can get my hands on that PDF? When I try to download it I get a 0 byte size pdf file.