Is there an economy mod that would not reward distance?

[epictetus]

A - "Hey, I brought you a bucket of eggs. Interested?"
B - "Yes, I'm always accepting eggs!"
A - "My friend always produces eggs! I got these here in 5 minutes, can I have twice as much money?"
B - "Uhh, no. As long as they're here within 24 hours, you get the regular rate."

A - "But they came all the way from the other side of town!"
B - "That's nice. Regular rate."


If you find that exchange to be too realistic for your TTD playing style, then there are some irreconcilable differences between us

[Eddy Arfik]

A- here's that shipment of coal you ordered
B- it's 2 weeks late, we're only paying you half
A- why? Coal doesn't perish in transport
B- no, but we ran out 3 days ago, the whole city has lost power, they are going to sue us
A- but it came all the way across the country
B- deliver on time, or pay the penalty

Still think delivery time isn't important?

[epictetus]

A- here's that shipment of coal you ordered
B- it's 2 weeks late, we're only paying you half - we ran out 3 days ago, the whole city has lost power, they are going to sue us

There isn't an actual contract. Presumably, all the powerplants that aren't being supplied are having just as much trouble providing power. When industries aren't being served, it's assumed that there is some sort of invisible market force keeping them alive (even if you permit industries to close in the settings).

A- but it came all the way across the country
B- deliver on time, or pay the penalty

To which I would add: "If you can't deliver on time, perhaps you shouldn't ship it from so far away. Our other suppliers don't seem to have problems."

If you consider that FIRS actually codes this into the supply chain of building materials and supplies then the incentive for regular delivery is already accounted for. In the context of OpenTTD, penalties for slow delivery only make sense if rewards for fast delivery do. Since rapid delivery of coal is mostly illogical, the penalty for the opposite is equally so.

Edit: Passengers and mail are two exceptions because the intrinsic value of the cargo is based on origin and destination. Any curve or ratio for time and distance would therefore render some types of transportation more advantageous than others (from both a passenger's perspective and a logistics perspective).

[leifbk]

A- here's that shipment of coal you ordered
B- it's 2 weeks late, we're only paying you half
A- why? Coal doesn't perish in transport
B- no, but we ran out 3 days ago, the whole city has lost power, they are going to sue us
A- but it came all the way across the country
B- deliver on time, or pay the penalty

Still think delivery time isn't important?

That example is not about the delivery time, it's about timely delivery.

In real world logistics, you'll get paid for delivery on a scheduled date. A shipment coming all the way from China to Europe may take months. Your transport company makes a deal with the customer to deliver on a certain date, or maybe within a week of that date. If the transport company is unable to deliver within the agreed time window, all hell will break loose.

In the FIRS economy there's a similar aspect, as some industries require two or three different cargoes delivered within a month to yield full production. If for instance a load of coal hasn't been delivered to the steel mill within the last month, you may lose a third of the production from a shipload of scrap metal coming from the opposite corner of the map. (Or rather, you'll lose the opportunity to transport a third of the produced metal to the metal foundry, and you'll lose the opportunity to transport some manufacturing supplies and goods from the metal foundry. You'll even miss the opportunity to transport some goods from the site receiving the manufacturing supplies from the metal foundry.)

That's a very instructive approach to the concept of timely delivery.

[epictetus]

A- here's that shipment of coal you ordered
B- it's 2 weeks late, we're only paying you half
A- why? Coal doesn't perish in transport
B- no, but we ran out 3 days ago, the whole city has lost power, they are going to sue us
A- but it came all the way across the country
B- deliver on time, or pay the penalty

Still think delivery time isn't important?

That example is not about the delivery time, it's about timely delivery.

Well said. And timely delivery is precisely what doesn't make sense for most cargo (within reason, of course - oceanic shipping shows the upper limits of how long a trip could take and still be viable). For perishable things, it should be much more harsh if it is to mean anything. And passengers/mail are a special case, of course. I think my first post got a bit lost since it's at the end of page 2 and I posted my example conversation immediately after.

[leifbk]

Well said. And timely delivery is precisely what doesn't make sense for most cargo (within reason, of course - oceanic shipping shows the upper limits of how long a trip could take and still be viable).

Oh yes, timely delivery does actually make a lot of sense for most cargoes. In modern logistics they're operating with minimal inventories ("stockpiles"), and rely on the transport company being able to deliver on the agreed date. Read about "Just in time" on Wikipedia.

[epictetus]

Well said. And timely delivery is precisely what doesn't make sense for most cargo (within reason, of course - oceanic shipping shows the upper limits of how long a trip could take and still be viable).

Oh yes, timely delivery does actually make a lot of sense for most cargoes. In modern logistics they're operating with minimal inventories ("stockpiles"), and rely on the transport company being able to deliver on the agreed date. Read about "Just in time" on Wikipedia.

Yes, but we're not dealing with the real world and all its complexities. I understand all about JIT logistics, but that's beyond the scope of what the game can model. It's something that you interpret the game to be simulating, which it is not.

The philosophy of JIT is simple: the storage of unused inventory is a waste of resources. JIT inventory systems expose hidden cost of keeping inventory, and are therefore not a simple solution for a company to adopt it.

How does OTTD model inventory? It doesn't. What would a JIT Inventory-based argument for timely delivery require? Inventory costs.

Regular delivery is modelled through FIRS.
Stockpiles are already modeled using station ratings.

You're clutching onto distance. Let it go.

[leifbk]

You're clutching onto distance. Let it go.

I think you're misunderstanding me. In RL, time in transit (ie. speed of the transport medium) is inessential for most cargoes. What counts is that the cargo is delivered on schedule. That's why shipping around the world makes economical sense, while the Trans-Siberian railway (mostly) doesn't. In the TT world it's the other way around.

[leifbk]

Yes, but we're not dealing with the real world and all its complexities. I understand all about JIT logistics, but that's beyond the scope of what the game can model. It's something that you interpret the game to be simulating, which it is not.

Yes it does. The FIRS economy rewards small and frequent supplies of multiple cargoes. It's ideal for delivery by road vehicles and small ships.

«During the birth of JIT, multiple daily deliveries were often made by bicycle. Increased scale has required a move to vans and trucks (lorries).»

[epictetus]

I think you're misunderstanding me. In RL, time in transit (ie. speed of the transport medium) is inessential for most cargoes. What counts is that the cargo is delivered on schedule. That's why shipping around the world makes economical sense, while the Trans-Siberian railway (mostly) doesn't. In the TT world it's the other way around.

I am not disagreeing at all. But in the context of what OTTD currently does and what it could do, including distance travelled unnecessarily complicates things and obfuscates the role of the player which is, arguably, to connect producers and consumers. The path of least resistance (the cheapest route) for resources is for them to go to nearby factories (or whatever).

There is no schedule for cargo to be delivered by in OTTD. The game suggests that this is what it models, but FIRS more accurately satisfies that. Removing distance has no direct effect on FIRS' model, it merely removes the unnecessary complexity.

WRT shipping vs. the Trans-Siberian: one makes more sense over another for a variety of good reasons and timely delivery and responsive supply would be an obvious factor. In TT it is the opposite because of the drawbacks to including distance in the reward scheme. The excellent example given previously that showed how a coal mine could "earn" 10,000 per tonne versus 60 or 500 and so on which demonstrates the inherent imbalance.

If the coal mine has a limited supply of coal and the market rate for that is fairly fixed, those constraints define the conditions around which an efficient system can be built[/b]. Should I use trucks or trains? I'll start with one and see how it pays off. Hmm... too far to be profitable. Maybe if I take it to the mainline a train and drop it off nearer.

And so on.

The 10,000$/tonne coal mine contradicts basic economics and that's why something is wrong.

[epictetus]

Scenario 1
If I built a powerstation at the far corner and took coal from the opposite corner and transported it by aircraft and made boat loads of cash, I would not consider that to be an economy, simulated or otherwise. It would be a gimmick and a cheap shot by exploiting the reward system.

Scenario 2
If you remove distance from the reward system, this is rendered impossible to sustain. Even if you tried to deliver every month, for years, as consistently as possible. The powerstation (which you, as a logistics company, built ...) would be booming in the middle of no where and you'd be bankrupt very quickly.

If you saw a mine and no factory nearby, it might be a good idea to build a factory nearby because the town would buy the goods. Profits all around, hooray! The high price of industry construction might preclude this, though. This is where strategy develops - as a means of optimising a solution. Profits might actually mean something.

Do you see what I'm getting at?

[leifbk]

The 10,000$/tonne coal mine contradicts basic economics and that's why something is wrong.

Of course it's totally wrong. But the game isn't really about making tons of money either. Personally I consider the 10,000$/ton coal mine as a childish exploit. As a grown-up, I rather prefer to build an efficient network, and tend to avoid the easy-money routes. Such as it is, the game is very open-ended and allows a large spectrum of personal playing styles, whether the basic economic model is broken or not. And I don't think we can expect that model to change.

[epictetus]

Ah. So could it not be sort of dealt with by making a NewGRF? There's the Altered Costs one that flattens out coal but the other industries would need some more work:




I'd like to see an even steeper drop off for fruit. Stable to a certain point and then a steep drop to zero.
Am I missing something? It seems too easy to do it this way.

[Panando]

Removing the untimely delivery penalty actually makes the problem worse. Without that penalty, then delivering coal twice as far will plainly and simply double the money you get. You would need to remove both the distance bonus and the untimely delivery penalty in order to avoid 10,000$/tonne coal.

The simplest solution which is already implemented is to use vehicle sets which only allow you to deliver coal at low speed. What this means is you are forced to eat the untimely delivery penalty when delivering stuff like coal long distance. There are a few reasons why this might be found objectionable but I think the main one is that it's just not a very adaptable system. And for instance, you still get paid a lot more for delivering coal by train at 80km/h than at 40km/h assuming the distances are long enough to make the time penalty bite.

One of the fundamental game balance problems with the current system is the (positive) double-whammy from speed. If a train runs twice as fast, it can deliver twice as much cargo, and gets paid significantly more per load delivered. Actually, just delivering twice as much cargo would be bonus enough from speed (at least when running costs aren't at an av8 level). For this reason, a simple flat rate for delivering stuff like coal might actually be quite well balanced. Certainly, if anyone wanted to try an experimental gamescript or patch, the conceptually simplest solution should be tried first. It would bias the profitably towards the shortest routes, but that is not really a problem, and is kind of actually what people want in this thread - that there should be incentive to deliver to the nearest consumer, rather than the furthest one!

[epictetus]

Removing the untimely delivery penalty actually makes the problem worse. Without that penalty, then delivering coal twice as far will plainly and simply double the money you get. You would need to remove both the distance bonus and the untimely delivery penalty in order to avoid 10,000$/tonne coal.

The simplest solution which is already implemented is to use vehicle sets which only allow you to deliver coal at low speed. What this means is you are forced to eat the untimely delivery penalty when delivering stuff like coal long distance. There are a few reasons why this might be found objectionable but I think the main one is that it's just not a very adaptable system.

I forgot that the rate is based on 20 tiles. I knew I had messed up

I'll be using slow trains from now on then. Thank you for clarifying this. I'll have to imagine that for any given industry-resource pair, they've already established that one is willing to do business with other and for precisely as many units as I happen to transport!

[Eddy Arfik]

Ah. So could it not be sort of dealt with by making a NewGRF? There's the Altered Costs one that flattens out coal but the other industries would need some more work:

It's entirely possible to do via NewGRF, in fact that would probably be the easiest way to do this, only problem is finding someone who thinks it's worth spending the time to write one.

[leifbk]

Removing the untimely delivery penalty actually makes the problem worse. Without that penalty, then delivering coal twice as far will plainly and simply double the money you get. You would need to remove both the distance bonus and the untimely delivery penalty in order to avoid 10,000$/tonne coal.

What I'm thinking of wrt "timely/untimely delivery" is delivery on schedule, ie. that delivery happens on an previously agreed date or time window. It has no direct connection with the actual time in transit or the speed of the transport vehicle.

The simplest solution which is already implemented is to use vehicle sets which only allow you to deliver coal at low speed.

That may probably be regarded as a "Bad Feature" because it takes choice away from the player. But of course it's the player who decides to use that vehicle set rather than another where all vehicles run at full speed.

If I understood andythenorth correctly, a game script might pick random (or not so random) routes and make contract offers. Constructing payment rules overriding the defaults and maybe schedules for delivery dates seems possible.

[andythenorth]

What problem is being solved by these suggestions?

Still I see no understandable explanation of what the perceived problem is. For example (because it was a recently mentioned case), what is the problem with $10,000/tonne coal?

[epictetus]

Still I see no understandable explanation of what the perceived problem is. For example (because it was a recently mentioned case), what is the problem with $10,000/tonne coal?

I don't understand where the extra money to pay for this coal is coming from or why the advantage should exist and therefore it feels either like:
a) cheating to build such routes (bullet train from corner to corner) or
b) like I'm being foolish for building a shorter line.

That's just me and I've hogged this thread enough, so perhaps someone else has a better explanation?

[Wahazar]

What I'm thinking of wrt "timely/untimely delivery" is delivery on schedule, ie. that delivery happens on an previously agreed date or time window. It has no direct connection with the actual time in transit or the speed of the transport vehicle.

It seems to be complicated to implement and play with (for example, all those subsidies are worthless) , but it remind me my idea with different payment model,
where instead of time-in-transit penalty, delay penalty is counted if transit time is greater than average transit time specific for certain cargo,
over given shortest (MD) route.
Average transit time would be counted using maximal velocity of available vehicles (available for given cargo + locomotives), of course vehicles without limits does not count.
Of course such average speed is updated and depend on newgrfs loaded.