Kicking ideas around...

[Gumboots]

I was just idly pondering scenario ideas. Seems to be a side effect of looking at all the existing ones while re-coding the site's pages.

Couple of things occurred to me. One is the classic AI stupidity when laying track. Easiest way around that would be to use a no-access territory. If the AI can't get access to the side of a mountain, it won't be able to lay stupid track straight up at insane grades. You might be able to use a no-access territory to guide an AI along the bottom of a winding valley, as long as the route wasn't too complex. This is probably worth testing as it could improve results in a lot of scenarios.

Another thing was I was thinking about Brazil, and in particular the Sao Paulo Railway. This was an odd one in that it was only a short line (47 miles total) but at one stage it was the richest railway in the world. The reason it was so rich is that it was the only (sane) way of getting cargo down from the highlands to the port of Santos, and back the other way, so everyone had to use it to haul their stuff whether they liked it or not.

The section up the Serra do Mar was too steep for locomotives in the 1860's, being built at roughly 10% grade all the way. This was because the whole thing was built on a tight budget, and trying for more normal grades would have made the line impossibly expensive to build, so they went with steep grades and used cables running from fixed engines to haul the cargo cars up and down. IOW, no locomotives. Locomotives did the flat section from Santos to the bottom of the escarpment, then from the top to Sao Paulo, but were detached for the steep bits. It worked really well.

This would be easy to model in RT3, because there's no locomotive that's easier to model than a locomotive which doesn't exist. Which would be a nice change from obsessively detailing the little mongrels. The escarpment could have pulling power reduced by event by a suitable factor (99% should kill just about everything) then the invisible non-existent locos designed for that section could have pulling power and free weight set to a suitably high level to compensate for the event (ie: FW 333 and PP 333) with top speed set to something like 10 mph.

That way you'd have to run the non-locos up the escarpment, because everything else would be stuck at 1 mph, and you wouldn't run the non-locos anywhere else because they wouldn't have enough top speed. You could still run fully loaded normal locos down the escarpment and get away with it*, but then they'd be stuck on the coast, so you'd need to buy new ones up in the highlands again. That should be a disincentive to "cheating", although no doubt someone would still do it at times if they really needed to.

*Because on a really steep downhill run the boost effect from the descending grade would be enough to get train speed up to useful levels, even with the event greatly reducing pulling power.

[RulerofRails]

Some maps did use this strat. One I remember is Wolvy's Orient express remake. But I think the idea didn't really catch on because it's impossible to block access to hidden territories, which is an aesthetic tradeoff.

I don't recall any reports of solid testing on AI behavior. I do have a theory or two. The enforced reserved cells in the city must have something to do with AI planning. If you notice they only enter or leave on the x or y axis. If there is a river/ocean or manually placed building blocking those cells, that is enough to confuse him completely.


Nice idea for the cable "non-locos". Intermediate stops could be a little troublesome to model in RT3, depending somewhat on the map scale used. Might have to use some custom buildings or warehouses, or perhaps 1.06 ship-at-a-loss to keep "normal traffic" flowing.

[Gumboots]

Yup, the "Maltese cross" reserved cells are definitely to cater for the AI. No other reason to enforce them.

Figures someone would have tried guiding the AI with no-access territories before. One use case which occurred to me is the company at Santiago in low grade's Chile scenario. The one that always jumps the river and goes straight up the side of that small mountain. If the mountain itself was a no-access territory the AI would have to think up something else stupid to do. The boundary could be hidden with a few trees and buildings if necessary. The AI might go back inland from the coast (Valparaiso, I think) if it couldn't go rock climbing.

Nice idea for the cable "non-locos". Intermediate stops could be a little troublesome to model in RT3, depending somewhat on the map scale used. Might have to use some custom buildings or warehouses, or perhaps 1.06 ship-at-a-loss to keep "normal traffic" flowing.

Would need to be a decent scale. If it was too small you could just put a large station top and bottom and that would cover the cargo anyway. Traffic flow should be ok, even if price differences were minimal, as long as there was a bit of space between catchments. The steep grades should help with differentials. But yeah, warehouses or whatever might come into play too. I think it's worth checking out sometime anyway.

[Gumboots]

Been thinking lately that I might be bonkers enough to do another scenario, and perhaps Romania would be worth doing. At the moment there is no Romanian map as such (although Romania does appear in some other good scenarios) and I think it has scope for some good games. It's a country with an interesting history.

I might even do it in 1.06, since with the Orient Express running through Romania it would make sense to have express goals, and they are immune to being cheated with 1.06's haul-at-a-loss.

Another possibility is Bolivia, which again is a country that should have good scope for shenanigans, and is not yet in the archives in its own right.

[RulerofRails]

It's a late reply. I was going to, but then forgot. After binging on TM to get Warrington Wire done, I haven't done much in RT3. Romania would be interesting for sure. I'm not a history buff, so I'm sure I would learn a few things.

Accurately simulating the Orient Express seems like a challenge. I'm cautiously confident that for gameplay it might better to not represent it physically, but have scripted events centered around it (determine the cost it has to pay for passing over your tracks, a trade-off for low vs. high, time pressures to make repairs on the line in case of disaster, some events relating to the classic "murder" theme where you can make some choices about what your railway police do, etc..

Long distance passenger travel is quite hard to represent accurately in my experience. Local is well represented by the game, but pumping production will mostly pump local traffic, not really what you are after. If you really want to physically represent the "Orient Express" train you might be better off making the map modular, so for example use Troops renamed to something else such as "rich tourists." Then you could add "Troops" supply/demand to some special "houses" (or other city buildings, Fancy Hotel?), which will be only placed along the actual route. Make sure to have a lot of them at both map end-of-lines. Maybe you want to put the large concentration in the neighboring countries so you can work in the border delays somehow? Although this method would in the best case only get a steady "flow along the route." Tourists will take many stops at the sights, not that realistic.

[Gumboots]

Accurately simulating the Orient Express seems like a challenge. I'm cautiously confident that for gameplay it might better to not represent it physically...

I'm not sure how you would represent it "physically". Unless you are talking about livery for carriages, etc (which would be fun, and quite easy, but obviously would apply to all carriages in the game).

...but have scripted events centered around it (determine the cost it has to pay for passing over your tracks, a trade-off for low vs. high, time pressures to make repairs on the line in case of disaster, some events relating to the classic "murder" theme where you can make some choices about what your railway police do, etc.

Yes, generally that sort of thing, although I would probably skip the murder theme, since that was set in the 1930's and not in Romania.

Long distance passenger travel is quite hard to represent accurately in my experience. Local is well represented by the game, but pumping production will mostly pump local traffic, not really what you are after.

Yup. Although it may be possible to do something by pumping production at well-spaced major cities along the route. I wouldn't be getting too fussy abut trying to emulate an actual "Orient Express". I'd just treat it like other scenarios have done: emphasise pax traffic and add whatever else to make things interesting.

For example, Romania was one of the first countries in the world to have a significant oil industry. The streets of Bucharest were lit by kerosene lamps in the 1850's, which was pretty advanced for those days. There was a lot of exploitation of oil by foreign investors (with very little of the profit staying in Romania) which means scope for tycooning and for dirty deals done dirt cheap.

Constanta is another interesting point. Until the late 1800's it was only a tiny fishing village. The port and the city around it came about due to competition with the alternative OE route through Serbia/Bulgaria. Romania was trying to keep a large chunk of the traffic, so had to invest (a lot) in facilities and in express ferries ("the white swans of the Black Sea") to handle the last leg to Istanbul.

[Gumboots]

I have figured out something about my lack of motivation to play RT3.

The classic RT3 scenario goes: Start with next to nothing > pig around for ages, desperately trying to get something happening > stuff starts happening.

I can't be bothered with the first stages anymore. I want to be entertained, right from the start. I get bored with the "challenge" of getting through the slow initial stages. I know it's possible, but I just don't find it interesting anymore. I think the game needs scenarios that are still challenging, but in a way that is interesting right from the start.

At the moment I'm not sure exactly what such scenarios would look like.