Idea: Improve acceptance of goods in terms of realism

[Dave]

Which comes back to a principle that is well known: pushing realism in the game for the sake of realism is BORING!

Speak for yourself pal - you strive to improve the game to be better to play.

I strive to improve to game to be more realistic.

[Eddi]

Which comes back to a principle that is well known: pushing realism in the game for the sake of realism is BORING!

Speak for yourself pal - you strive to improve the game to be better to play.

I strive to improve to game to be more realistic.

which is why he is a dev and you are not, maybe?

[Dave]

Which comes back to a principle that is well known: pushing realism in the game for the sake of realism is BORING!

Speak for yourself pal - you strive to improve the game to be better to play.

I strive to improve to game to be more realistic.

which is why he is a dev and you are not, maybe?

That is more likely to have something to do with his excellent knowledge of the coding language...

Whereas mine doesn't exist.

Still - the difference is that he's not there to be realistic - he's there to re-create a classic, timeless game.

I'm sure belugas understands that we all play differently. If he builds a game on which my ideas can be implemented via other means, then we all win don't we?

[belugas]

I'm not against realism in the game. Let's make it clear.
I'm against the notion of adding realism just because it is lacking.

An example? A long time ago, a dev removed the capability for planes to be refitted to a lot of cargo, just because, for him, it was not realistic that a plane can be refitted to carry, let say coal.
Sure, it's not. But it's fun to have a fast mean of transportation for coal (or else).
In this case, realism killed a very nice feature of original game. True, it can be "re-introduced" by means of grf, but you see the point.
I understand that people like realism. But adding realism that will affect gameplay in such a way is not really wanted, at least in my book.

Plus, sometimes, those realistically oriented suggestions might interfere with the grf system. And that becomes a problem... But that is another matter.

Edit: By the way: andythenorth: you really are on to something there. I think you've added something very interesting to the debate, at least (once again) from my point of view.

[PikkaBird]

The conceptually simple (but not simple to implement) solution is that industries place orders (long term contracts) for input cargos with other industries.

Aww, andy, you stole my thunder.

I definitely think a contract-based system for industry cargos (ie, not passengers or mail) would be the way to go, if one was rewriting the payment system from scratch. However, to keep some gameplay options alive, the contract should be between the accepting industry and the player, not between industries.

So a factory, for example, would offer a contract: we want x units of cargo delivered this quarter, and we'll pay $x. If a player accepts the contract, they can source the cargo from anywhere they like. Perhaps players could also propose contracts to industries; perhaps contracts could be negotiable (amount supplied, length of contract, payment). There would have to be some kind of economic AI in the industries (and in AI players) which determines what an agreeable contract is, based on available supplies, industrial demand, etc. Non-fullfilment of contract would result in non-payment, or eventually the cancellation of contracts and industries unwilling to do business with you.

Of course, one drawback of this system is that income would no longer be directly linked to vehicles, so some kind of new vehicle rating system would also be required.

[andythenorth]

Edit: By the way: andythenorth: you really are on to something there. I think you've added something very interesting to the debate, at least (once again) from my point of view.

Thanks, I'll post something more coherent later. Working now

[andythenorth]

Cargo payment thoughts

Lets see if this is more coherent than my post on the previous page of this thread. I will repeat the previous post somewhat, but it may make more sense.

-- Part 1 --

In real life - with apologies to you Belugas, but I think you'll permit this - the cost to the customer for any delivered cargo involves two components:
- cost of the physical items
- transportation cost
To aid simplicity I am ignoring other costs such as tax, foreign exchange, packaging, customs fees etc.

In the game world of Transport Tycoon, player income is derived from the second of these components - transportation. The game does not attempt to provide an economic model that includes the first component - cost of the physical items.

The game does not have an economy based upon market value of goods. The game is concerned primarily with income and expenditure in respect of transportation, or related to transportation, including costs of construction, advertising, bribery, sabotage or purchase of competitors, and the special case of funding (but not owning) new industries.

The model for calculating income from transportation is fundamentally sound and correct for the game world. Player's revenue is derived from the distance cargo is transported, with a varying degree of time-sensitivity. This is:
- accurate in respect to the real world (useful, but not essential)
- simple to understand
- pleasing in terms of game play
- convincing, in that it fits an intuitive economic model for how value should be calculated.

Any proposal to use a different model for calculating income from transportation will simply fail, because the current model is the most correct. (sticky anyone?)

-- Part 2 --

The game world permits players to earn more money by choosing to carry cargo from a source far away, when there is an equivalent source nearby to the the destination.

That is confusing or irritating to many players because it goes against the intuitive sense that industries shouldn't pay more for an input cargo than is necessary. I believe this is a reasonable expectation. As humans we seem inclined to make instinctive calculations of value and economic efficiency.

When we're making great games, we can ignore 'real life', but we shouldn't ignore the natural tendencies of the human mind.

Because the current model permits un-satisfying behaviour, suggestions to modify cargo payment rates are frequently made on this forum.

These solutions tend to conflate the transportation fee for a cargo with the cost of the physical items. As I have argued above, these are properly treated as two separate components.

The game offers no model for the cost of the physical items, so there is nothing that can be modified in this respect. We could provide a market economy, but this might be both an onerous amount of coding, and undesirable - I have further thoughts on that, but want to keep this post focussed.

Because there is nothing to see in respect of the cost of the physical items, players (reasonably) look to what they can see, and make suggestions to adjust the cargo payment rates. For these players, the desired change would be to have industries 'prefer' cargos from nearby sources.

This 'preference' cannot be achieved by simply modifying payment rates. Such modifications would fundamentally damage the underlying economic model that *is* represented by the game - that of a transport company earning revenue for the distance a unit of cargo is transported (with a time sensitivity modifier).

What is needed is a solution that addresses the *true problem*: some players think (reasonably) that industries should 'prefer' cargos from nearby sources.

Therefore we should simply make industries 'prefer' cargos from nearby sources.

How would this best be achieved - both conceptually, and in the implementation?

[Dimme]

A very interesting post Andy! I really like the ideas you have for the FIRS set, it is obvious that you have thought things thorough.

Any proposal to use a different model for calculating income from transportation will simply fail, because the current model is the most correct.

Nice observation. Correcting the cargo payment rates may change things a little, but will not make the game much better.

I'd like to point out that for passengers and mail, cargo destinations changes the game in a very interesting direction. Some changes can still be made to that model, but the player now has to simulate a real transport company. Especially if players are allowed to freely connect to each others' stations or in some other way compeat to give the best transportation options, I think that is the way to go.

For other cargo, however, cargo destinations is not a 'natural-to-mind'(yes, realistic) model.

I belive good solutions may be found simply by changing the way cargo destinations distribute these cargoes. Say, if a company wants to transport coal to a far power plant, and another wants to transport to a close plant, most of the cargo chooses the short route.

One thing I'd like to point out is that the cargoes should not be distributed between all accepting industries, only up to the, say, three most preferred.

The model may of course be made more sophisticated. One could take into account if the far power plant has any nearby source of coal or not, or if it close to a power consuming industry and so on.

Ideally, large access should be given to NewGRF writers, so Andy can make his own cargo distribution preferences for the FIRS industry set

[AndersI]

Say, if a company wants to transport coal to a far power plant, and another wants to transport to a close plant, most of the cargo chooses the short route.

I think you are using the wrong measure here. Distance shouldn't mean so much as 'transported volume per unit time per unit cost'. As a producer, I wouldn't care where the goods go (as long as I get paid for it), the transport should be as efficient as possible, with a cost as low as possible.

This would mean that a high volume, reasonable cost, long distance route should get more goods than the local five tile horse carriage route.

Not as easy to calculate, though...

[Dimme]

Say, if a company wants to transport coal to a far power plant, and another wants to transport to a close plant, most of the cargo chooses the short route.

I think you are using the wrong measure here. Distance shouldn't mean so much as 'transported volume per unit time per unit cost'. As a producer, I wouldn't care where the goods go (as long as I get paid for it), the transport should be as efficient as possible, with a cost as low as possible.

This would mean that a high volume, reasonable cost, long distance route should get more goods than the local five tile horse carriage route.

Not as easy to calculate, though...

It is quite complex, I agree. Distance does matter, since the cost of transport depends on distance. Volume also matters greatly. Many other things may be taken into consideration. My main point is that if cargo can choose between several industries, it should choose the one that is best for seller and buyer, not for the transportation company.

What I would like to see is that transport companies compete to make industries happy

[audigex]

I think you're on the right tracks ('scuse me) - very few industries in [warning: example situation] London would pay £100 per crate from a factory in Edinburgh if there's one in Essex that can transport it for £20.

Although at the same time... perhaps we need to take into account the amount of goods that are made in India/China?

[Dimme]

perhaps we need to take into account the amount of goods that are made in India/China?

Then you can make a scenario where most of the goods producing industries are in one side of the map

I think a proper calculation of goods destinations would need to consider the position and production of all industries on the map in some way.

[jasperwillem]

It just sounds like Railroad Tycoon II / III, a town station has an economic limit of acceptence based on industries in range of this station (2 houses take more carge as 1, same with factories), and if you need 1 STEEL + 2 FRUIT to make goods, they are both stored at the factory and it does only produce if you have the 1/2 combo.

Then in TRAFFIC GIANT of JOWOOD it works with destinations for a SIM. A sim goes to work (1 movement from A to B, and back B to A)

Both ways have a good basic idea... but its not near the proximaty idea... So I hope this would be applyable...

Edit; read about happyness of factories... this is used in MMORPG's before... but a SIM with satisfaction is horible... just like humans in real life...

[SirXavius]

I agree on your proposed changes, andy. If i hear you correctly, companies will REQUEST cargo based on proximity. Excellent idea, but i would also give the player choices. For example, the three closest suppliers, or maybe one close supplier, a medium, and a distant supplier. The main reason you would do this, is because the player may have already built rail lines or other paths, and it may cost less startup for one delivery as opposed to another request.

Even more, how would this affect the placement of subsidies? Perhaps subsidies can have a multiplier based on these requests (rather than the flat multiplier in the Advanced Settings).

[ever]

How would this best be achieved - both conceptually, and in the implementation?

This might seem radical or it might not but here it goes.

Allow the player to charge for demand. That is, allow the player to determine cargo payment rates.

Now this would work differently for industry and passengers.

For industries it would work something like this:

First thing first get rid of station ratings. Instead of that in the Station GUI for a station that accepts a certain good, the player will be able to set the amount of money to receive in exchange for the good. The higher you charge the less will be demanded.

In a monopoly situation this will be all about hitting demand elasticity. i.e finding the price at which P*Q is maximized.

Now every industry type can have its own demand curve for cargoes and can be easily calculated. For example

P = -3Q + 10,000 = A generic equation for an incredibly simple demand curve.
Q = (10,000 - P) /3

Here the player will set P at lets say $20 and Q will equal 3326. Thus 3326 units (per month that is) of the Cargo will be demanded at price $20 and assuming the player can deliver all of it he will receive $66,533.

Now there are two "problems" or objections one can have with this approach.

The first objection would be why would anyone ever go over the price hitting elasticity? I mean everyone will have figured it out on openttdcoop and posted spreadsheets online so it will be just like having static cargo payments again.

Well lets see.

In the above given example the intercept on the Q axis is at 3334 units of cargo. This means the midpoint is at Q = 1667 and at this point P = $4999 Which yields a Revenue of P*Q = $8,333,333 (Sorry about the weird numbers.)
This is the largest amount you can get with a single one of these hypothetical demanders.

So why would you set a price higher than $6667? Well If you're delivering below the demanded amount of units a month then its to your advantage to set the price higher even if it means that the hypothetical maximum you could earn at that price is lower.

So if you're delivering 300 units of cargo it makes the most sense for you to keep raising the price till only 300 units are demanded. In this case that will be $9100/unit earning you $2,730,000 a month compared to charging the elastic price and getting $1,499,700

Of course you have all the incentive in the world to get cargo delivered up until that magic number of 1667 units a month as this gives you maximum revenue. This will often mean being forced to deliver from longer distances as one nearby source isn't enough to hit the quantity at elasticity. Whether or not the higher revenue will be able to generate a higher profit depends on the costs of delivery. This will be up to the player to calculate. However the incentive of higher revenues the more he delivers will

A)Make the player want to deliver from longer distances.
B)Make the player prefer nearby industries to those far away

The Second objection is why would anyone want to go under the price at elasticity? Why would anyone ever want to charge less than $4999. You will notice that even at $4990 Q goes up only to 1670 and total revenue is only $8,333,300 which is $33 less than that at the higher price!

Well the answer to this objection is competition.

With this sort of model competition becomes really dog eat dog. Simply put if one player is making a 20% profit margin on a particular route then another player has the incentive to replicate that route exactly and just charge less. Sure he won't be getting a monopoly price but at least he'll be making profits. This will bid prices down severely till the quantity demanded cannot be satisfied by the current state of supply. (disclaimer: the demand curve would actually be hyperbolical not linear, I just did linear for ease of maths)

This is the incentive to charge less than the monopoly price. The non-monopoly price imo, this would also make competitive play a lot more fun

Now a third even more tenuous objection might be "But the player shouldn't be able to influence the deals between suppliers and demanders"

This objection can only be asked if you don't understand exactly what I'm proposing. Ok, here it is assumed that all suppliers are in perfect competition. That is all coal producers charge $300/unit of coal. So when the demander is deciding where to get the coal from or how much to get the only variations in price are the cost of transport you attribute to them. So the only difference between coal from coal mine X and coal from coal mine Y is the amount $300 + $the_amount_you_charge.

Now if anyone has actually bothered reading this I think you'd agree that this is the best way conceptually to achieve

What is needed is a solution that addresses the *true problem*: some players think (reasonably) that industries should 'prefer' cargos from nearby sources.

Therefore we should simply make industries 'prefer' cargos from nearby sources.

That.

And if you don't agree tell me and tell me why. I'll be glad to hear

[Xander]

That just seems very math-y and exceptionally dull. Does it really matter so much how much money you make? The great joy about (O)TTD is that it doesn't get bogged down with running a business like most other games do. It just lets you build a track from Coal Mine to Power Plant, move some coal, and provided someone isn't doing it better than you, you make money.

That bit in bold is pretty crucial. This is Transport Tycoon, not Business Manager 1998. It's about moving Item A to Location B as well as you possibly can, which is what station ratings measure. The mechanics reward the player who moves the most the furthest, which is exactly what Transport is all about, not who balances the economic scales the best.

[ever]

Well the maths was only to help you understand the concept I don't imagine any player will actually bother too much with it.

It was just to demonstrate:

A) The higher the price the less demanded, the lower the more demanded.
B) How this gives the player incentive to transport things from close by and far away, as opposed only some defined maximum profit distance for a given speed.

The question was how to solve the problem of very unintuitive unprofitable short route connections.

I solved that problem.

I don't think that not liking the business side of the game is a valid argument against it. I mean you can always just play with cheats and make the most efficient transport system you want.

But then this discussion isn't really for you.

[Xander]

The question was how to solve the problem of very unintuitive unprofitable short route connections.

I solved that problem.

Transport Tycoon (Deluxe) has Never rewarded the player who moves cargo 3 tiles. The whole point of the game is/was about moving things as far as you can, as quick as you can, as cheap as you can. TTD was not about setting the fees and balancing the economy. This wasn't a problem, it was a major chunk of how the game works!

Chris Sawyer is clearly an exceptional programmer and it would've been easily within him to make the cargo rates a variable and reward or punish players for charging too much or not enough. But he didn't, and neither should the OttD team. From their own About page:

It attempts to mimic the original game as closely as possible while extending it with new features.

Changing the model of how money is made is not, IMO, mimicing the original nor is it a feature. It is a full blown re-write and the creation of a whole new game. If I want to play a game where you win or lose based on your business ability, I'd play Capitalism. If that's what you want, I recommend you do the same.

[andythenorth]

The whole point of the game is/was about moving things as far as you can, as quick as you can, as cheap as you can.

Hmm, yes, no disagreement there. But the original game was a radically smaller map than many of us choose to now use. I spent what seemed like most of 1996 playing original Transport Tycoon, and it was a rather different game to the flavoursome beast we now have in the form of OpenTTD. (For a start, most of the time was spent placing and tweaking signals on rail routes, which is an entirely story).

For various reasons I am interested in tweaks to the economic model that also *preserve* the essential method by which a transport company gets paid (a charge per ton / mile, modulated by the time sensitivity of the cargo). My motivation for this is that I find it damaging to the integrity of the game world that industries accept unnecessarily high (bizarrely high) payment rates for cargo.

(insert 'in my view' throughout this, it's highly subjective, I'm also potentially rambling here).

There are many things in games which don't and should not reflect reality. However the cargo acceptance model goes way beyond a plausible suspension of disbelief. Industries do not act as discerning customers, but simply accept whatever we, the transport company, choose to deliver to them, at whatever transport cost.

(Digression) What we have currently is seriously limited as a 'game'. For a player, the basic rewards of a 'game' are usually:
1. figuring out what the challenge is ("I get it" = a dopamine hit).
2. increasing skill level ("I'm good at this" = a dopamine hit).

In OpenTTD these rewards are met by:
1. understanding that you are extremely likely to win as you long as you start the game by shipping coal from a mine to a power station over a relatively flat route.
2. getting good at managing routes to meet the various game goals.
Beyond that there's really not to the 'game'. However because OpenTTD constantly evolves, and there are also regular newgrf releases, it remains rewarding to try out new features, newgrfs, or different tactics, or different self-imposed restrictions. However it seems there are a lot of people treating the 'game' as either:
- a low-hassle, space-saving and low-cost simulator of a model trainset (pleasure comes from variety and combination of vehicles, and setting up semi-realistic routes).
- a network optimisation simulator (pleasure comes from how many maglevs can fit through a junction in one minute).

These are both fine and valid things to do, but not really much of a 'game'. I'd like a bit more challenge, something on the level of Railroad Tycoon 3, where the supply and demand economy kept players on their toes, but I am trying to figure out how to do it in OpenTTD *without* requiring a supply and demand economy (RT 3 was far more a trading/industrial game than a train game), and *without* changing the cargo payment calculation model (which is absolutely correct as it stands).

I asked for possible theoretical implementations of a method for industries to 'prefer' cargo from nearby sources (see reasons in posts on page 2 or 3 of this thread). I'm not sure this will even help solve the 'more challenge' problem, but I *do* think it would put an end to the frequent and occasionally tiresome suggestions to change the cargo payment model, and let us move on to something more useful. For me this would be something that moved the game back towards being a more compelling transport game and less like a network optimisation simulator or a trainset. I think representing the actual needs of the customer (industries) somehow would be useful to set part of the challenge. OTOH, this might all get solved with cargodes/cargodist patches!

Ever has suggested a solution to the question I posed. I'm not sure I understand it, or that it's the right solution, but it's welcome that someone has taken time to think about it. So thanks to Ever for that.

Incidentally (and this is no argument with you Xander - more thinking aloud) - if the game is simply about moving things a long way at top speed, I could code a 2048 ton capacity ship with an insane top speed, and a purchase and running costs of £1. I could also code an industry newgrf with warehouses guaranteed to both produce and accept 2048 of goods per month. Put those on islands at either end of a giant, mostly water map, run insanely fast ships, and you'll easily win yes? But would it be fun?

cheers,

Andy

[AndersI]

There is already one source of 'demand' in the TTD world: Subsidies. Perhaps that concept could be augmented to allow for a more demand-driven game, but still very much TTD-style? No deep thoughts here, just an idea that struck me:

- Generate many more 'subsidies' with a factor of 1.0 (we don't want to earn a lot of money), let them have a much longer life (10-50 years?).

- Confiscate the profit on non-subsidized routes (but let us have the loss if there is one). Or perhaps confiscate X% of the profit?

This way you have to 'listen to the demand'. You could still build a non-requested route by speculation that it will get a demand later, but you wouldn't earn anything from it (or very little).

The subsidies generator could be built as sophisticated as you want, preferring shorter routes, giving higher factors for shorter routes, etc.