[OTTD] FISH....FISH FISH FISH FISH (work in progress)

[andythenorth]

I've redesigned the set, some ships are gone. Not saying more right now. Mount Blaze thing may or may not reappear in future.

No promises, but I think a 0.1 release is pretty close.

[andythenorth]

Squid Ate FISH! RC 1 now released.

http://www.tt-forums.net/viewtopic.php?f=67&t=71080

[andythenorth]

Anyone transporting Alcohol, Milk or Water cargos with Squid ships? Which ships?

(Figuring out what's a good gameplay capacity for these cargos).

[leifbk]

Anyone transporting Alcohol, Milk or Water cargos with Squid ships? Which ships?

(Figuring out what's a good gameplay capacity for these cargos).

I'm using "Little Eddie" (aka. the Eddystone tanker) for alcohol. I don't usually boost grain and fruit production (keeping farm supplies for the forests), so the size is just right for 100-200 units a month.

[leifbk]

Regarding running costs versus size of the ships, something seems to be way off. I just compared the values for tonnage, cost, and running costs for the Little Cumbrae, the Altamira, and the Frisco Bay:

Code: Select all

Little Cumbrae       140    £12,187      £3,062
Altamira             360    £28,437      £4,725
Frisco Bay           810    £73,125     £11,725

Upgrading from Little Cumbrae to Altamira may be a good idea, at least when you're running on open water, and you've got more than a handful of ships on a route. But while the Frisco Bay is about 2.25 times bigger than the Altamira, the running costs are 2.5 times higher per year

In real life, the reason for running big ships is that the cost per unit of cargo is much lower than for small ones. The fuel for a bigger ship may cost more than for a smaller one, but the most important factor is probably the crew size. You don't need twice as many persons to run a 200,000 tonner than a 100,000 tonner. The equation for passenger ships is obviously different, as a relatively large staff of service personnel is needed aboard those.

In OpenTTD there are several effects acting against large ships:

A ship does not make money while sitting at the dock loading. Small ships load quickly. A big ship can take months to fill up. And then there's the "Time since last visit" variable, at least when you're using the default "grim and mean" station rating algorithm. Small ships with frequent visits keep the rating high. Delaying the arrival of a big ship in order to reduce loading time results in a catastrophic fall in the station rating. And then there's the potential overhead when eg. 5 ships are too many, and 4 are too few.

This morning, I replaced 5 Frisco Bay ships with 19 Little Cumbraes on a route carrying wood from a forest going gung-ho to an iron work. A part of the route is along a river/canal, so the Cumbrae will run faster. Combined with the fact that an entire shipload can be ready when the Cumbrae arrives and then load instantly, only 2/3 of the tonnage is needed. The profit will be higher because of the fewer days in transit. The running costs are almost exactly the same:

5 * £11,725 = £58,625
19 * £3,062 = £58,178

This is the only source of wood for the iron work. Thus, the metal production will go way up when the wood arrives much more frequently.

So, what's the point of big ships, apart from their impressive looks?

[andythenorth]

In short, two simple answers.

1. yes - large ships are arguably pointless in OpenTTD. Because ships have no infrastructure constraints, the optimum ship size would be 8t (for industry reasons), travelling at the maximum ship speed (around 79mph). This would make a highly efficient, but not very interesting gameplay though.

2. yes - Squid Ate FISH needs cost adjustment before 1.0.0. Some of the ships have been shuffled around for capacity, speed etc, and the costs need tweaked accordingly. They're intended to be roughly balanced against Iron Horse, but balancing cost against other transport types is notoriously hard.

Meanwhile, would you post any other ship costs that look strange? I'll tweak the costs soon.

[Eddy Arfik]

I find the large ships are very useful when I have a lot of oil rigs or fishing grounds close together, I can send a few ships to pick up from all the sites, with conditional order to return to shore when full, instead of sending a horde of small ships to each pickup point individually. Also I run 1500 ton trains, so ships with similar capacity are great for transfers

[leifbk]

Oh yes, the large ships are useful in special, very high volume situations, where you need to transport several hundred tons a month. I don't think andythenorth would have bothered creating them otherwise But in general, with normal FIRS industry output, they're just a waste of money. If a ship takes more than a month to fill up, you should definitely consider smaller ships. Or, if possible, go for other transport means such as road vehicles.

For my own part, I like the "small is beautiful" concept. In the current Squid incarnation I miss the tiny Meteor, and I hope that it will return one day. I also loved the Saint Marie barge tug; it was perfect for hauling ore, coal, and metal on inland waterways. The 100-200 ton ship class definitely has its use, but is all but gone after 1970 now.

[Eddy Arfik]

Pretty much everything is high volume in my games, my brother and I compete for top spot on delivered cargo graph, it's not uncommon for us to have secondary industries producing over 8000 ton/month. I do miss some of the smaller ships from earlier versions though, a river through the rainforest doesn't quite look right without log tugs.

[Wahazar]

Squid Ate FISH needs cost adjustment before 1.0.0. Some of the ships have been shuffled around for capacity, speed etc, and the costs need tweaked accordingly.

I propose to use formula based on ship mass/power/v_max to determine ship running costs.
It should help with tweaking running costs inside a given set.
For example, try to use following formula:

Code: Select all

crew_number * CREW_RC + mass * (CONST_1 + v_max * v_max) / CONST_2

where CREW_RC, CONST_1 and CONST_2 are constants.
I'm using following particular constants: CREW_RC = 10, CONST_1 = 1296 and CONST_2 = 2592, but probably you will find another numbers satisfactory.
Additionally, for large/fast ships, you can use switch for dynamic change of running costs (depending on velocity).

Roughly adjusting for others vehicle sets can be done via base costs.
For example, I'm playing with ships running costs increased 2x to compensate free infrastructure.

[leifbk]

Meanwhile, would you post any other ship costs that look strange? I'll tweak the costs soon.

I've set up a small table for all the general-purpose ships, leaving out the special ships and passenger ships. The columns are the name of the ship, cargo units, speed in km/h, buying cost, annual cost, and finally mass times speed (a simple "performance" measure) over annual cost. I should maybe have added buying cost over lifetime to the annual cost, but it's probably of little consequence.

Code: Select all

Little Cumbrae       140    29      £12,187      £3,062    1.32
Altamira             360    29      £28,437      £4,725    2.21
Frisco Bay           810    29      £73,125     £11,725    2.00

Marstein             270    34      £22,343      £4,375    2.10
Cadiz                450    32      £40,625      £8,575    1.68
Santorini           1120    32      £86,328     £15,050    2.38
Maspalomas          1440    42     £101,562     £19,337    3.12

Eddystone            330    32      £28,437      £6,475    1.63
Hopetown             770    32      £65,000     £13,212    1.86
Yokohama            1660    42     £132,031     £22,312    3.12

Whitgift              65    29       £4,062      £1,925    0.97
Nieuwpoort           744    48      £62,968     £15,137    2.36

As you can see, there's little coherence in the numbers. The final column, performance per annual cost, is in my opinion the most interesting. The higher the value of this number, the more profit you can expect to get from the ship. It seems like a number around 1 is appropriate for the smallest ships, growing to around 3 for the largest. The big disadvantages of the large ships discussed in my earlier post will in practice tend to equalise the profits, so you may want to go for an even bigger difference. Although I suspect that when you need a Maspalomas, you couldn't care less about the cost of running it

The Cadiz stands out as a hopeless under-performer in this table, and I don't get why you'll need it when you've got the Marstein which can do the job even faster.

BTW, McZapkie's formula looks interesting, but I haven't studied it closely yet.

[andythenorth]

Current formula is here

https://dev.openttdcoop.org/projects/fi ... ip.py#L120

'fixed_run_cost_factor' was originally intended to represent number of crew.
'fuel_run_cost_factor' was originally intended to match relative costs of diesel, steam etc

In practice, stuff like that just doesn't matter for gameplay, and Squid needs this simplified, and the values reset.

Iron Horse has a formula, but just one property on the vehicle that can dibble the default formula result up or down for balance. The realistic formulas don't tend to result in satisfying gameplay. Also it's plausible that every vehicle ends up dibbled, meaning the formula would be pointless and the cost property might as well just be set directly on the vehicle

Table looks useful, ta. I'm going to reset and normalise values against that.

[PikkaBird]

Hoverzeppelins!

(I know it's the full cabin version you particularly want, it's coming soon (tm)).

[STD]

Look simply fabulous . Looking forward to the rest of the ship models .

[leifbk]

There's too little difference between the Marstein and the Cadiz; I just replaced 13 Marsteins with 10 Cadizes. Normally you'll want to halve the number when you upgrade, so I suggest increasing the capacity of the Cadiz from 450 to 600 tons. With the same annual costs as in the table above, that will increase its "performance per annual cost" from 1.68 to 2.24 which looks much better.

[Digitalfox]

I just watched a documentary about http://wingship.com/ !

Speed's up to 180KM/h and starting at 50 up to 150 passengers!

Not only it has been built, it's already on testing on waters.

This would make a great addition to the set

What do you think Andy?

[andythenorth]

Ekranoplan is often asked for but game limits ships to 79mph or so. At that speed it's gameplay-identical to a hoverzellepin.

[Digitalfox]

Ekranoplan is often asked for but game limits ships to 79mph or so. At that speed it's gameplay-identical to a hoverzellepin.

Yeah I'm aware of ship speed limit http://www.tt-forums.net/viewtopic.php?f=32&t=71173

But maybe by including it, put just a little more pressure

[Wahazar]

Is it possible to enable euro-roster in official on-line version?
I'm still using buggy alpha version on my server.

[andythenorth]

Is it possible to enable euro-roster in official on-line version?

No, it's not close to done yet