Strategies for reducing congestion

[GuineaPig]

I just thought that one of the fun things I try to do with this game is use elaborate track arrangements to reduce congestion and keep trains going. The real problem, map size, can't be changed (your average train has a length of 50-100 km), but some maps are tighter than others.

I was just wondering what other people's strategies were to reduce congestion. I use the grid exploit as much as is feasible, bypasses, un-electrified shortcuts on electrified main lines for diesel freight trains, several terminus stations in big cities, etc. Are there any obvious (or not-so-obvious) strategies I'm missing?

[Gwizz]

They are the ones I use.

I do try to reduce the time the trains stay in the stations by buying the related support buildings, Telephone, etc.

I use straight track leaving a station so the trains can get up to speed a fast as possible.

On the map I'm playing now, (Heartland 6) I program a lot of AI trains and have them use their own stations to deliver cargo to my industries.

If I didn't my trains would wait on the AI trains to unload.

[Andrew]

I try to get straightest track possible on flattest land.

I also try to make sure trains usually take right turns as much as possible (and build relevant right turn "off ramp" tracks). With as few trains taking left turns as possible, you don't end up with as many trains trying to cross over each other which can cause points of blockage.

[Gwizz]

I have used two single tracks a straight mainline route and then used a crossover to change the right hand rule to a left hand rule or the other way around. This slows the trains a little because of the turns but doesn't stop a train.

Another time, and Just for the fun of it, I built a 10 track yard. I placed trains (stopped) in the yard as I built it. Some loaded with cars.

I placed stations by number at different positions around the yard. (for example: track 1 west and track 1 East, stations)

Then I stated each train in turn individually changing stations until the train was in a different life like positions.

It looked like a real yard in action.

I had planned to program the yard with trains passing past each other through the yard.

The main problem was the locos didn't back up or turn around in a realistic way.

So, I used way points on balloon tracks at the edge of the yard which turned the locos in the correct direction.

I put some stations at the center of balloon tracks on the turn around. Way Points worked best for just turning.

For a small layout I might have an oil well at one turn around station

and a diesel refinery at another and a port that accepted diesel at still another.

I also built a bridge over a cut with a lower track and an upper track over the bridge, crossing over the cell corners.

The trains passed normally, one above and one below.

What I could not solve was the upper train's flags went sown the steep slope and up the other.

This slowed the train like on a grade. This also indicated that a loco is controlled by its flags.

[GuineaPig]

I have used two single tracks a straight mainline route and then used a crossover to change the right hand rule to a left hand rule or the other way around. This slows the trains a little because of the turns but doesn't stop a train.

Can you explain what you mean by this? I don't understand the layout.

[Gwizz]

The two single tracks cross the cells at the sides. At the crossover. the two tracks cross over the corners of two cells. This crossing puts the trains (still going in the same direction) but the train that was running on right hand track, is now running on the left hand track. By doing this. neither train has to stop for the other train at the crossover.

[GuineaPig]

Ah, I see what you're saying. I guess that would work really well as a freight/medium pax main line, and then have another, straighter, long pax/mail line.

[jeffryfisher]

Tips from the top of my head:

* Arrange crossings to be diagonal at points. This is doubly important when crossing a foreign RR.

* Build lots of track. Instead of having only one transcontinental main line that will become congested in the middle (or wherever busy branches merge), build several parallel routes. Your passengers aren't picky about where they go, so NE pax can be exchanged with NW, and SE w SW, and middle with middle etc.

* Move slow freights on their own tracks to their own stations. You can "back door" industries to produce fast freight in the stations on (or connected to) your mainline. If unconnected-to-other-track building is allowed on a map, then freight can be run on entirely separate networks so slow freights can never wander onto express lines and get run over.

* Trains can be hosed when trying to turn around on busy track; (because they get hit from both directions). Therefore, don't run your main lines through cities. As soon as you can afford to, plan your mainlines to miss mid-route cities completely. Resist the temptation to use mid-route cities for the sand/water/roundhouses on your mainline. Spend some extra money to buy a few extra stations for the roundhouses etc along the mainline. Put each big-city station on its own branch, even if it is parallel and only two squares away.

* After crossing two busy lines at a diagonal point so they won't interact, you'll probably want a ramp between them so some trains can transition from one to the other. If at all possible, make the ramp longer than your longest train. If a train gets its head into one line while its tail is still in the other, it can block (or be blocked by) trains on both lines. If both lines are busy, the train can be stuck for years (or decades).

* When selecting locomotives, give credence to acceleration. When a train has been stopped, its acceleration determines both the penalty for stopping and whether it can get back up to speed before causing another collision.

* When selecting locomotives and planning consists, try to engineer all of the trains in a net so that they move at the same speed.

* Avoid long trains on crowded track. The effect of length is maybe cubed. Weight reduces acceleration and speed while length increases passing distance and ramp requirements.

* Don't buy unnecessary trains. Not only should you avoid the siren song of carrying every load of freight within reach, but you should avoid buying new trains just because you see pax/mail piling up in your big cities. If you have a grind spot causing trains to stop, the accumulating pax/mail may be a symptom of congestion itself, and not a demand for more trains. Buying more trains in that situation will just add to the problem.

Long ago, when I was a newb, I spent several game years feeding new trains into a wasp-waisted net as I celebrated the bountiful pax and mail in my cities. Then I began to wonder why the pax and mail were still piling up. I found a mid-continent choke-point, and by the time I had untangled the mess over ten game years later, I had pulled over 70 trains from the wreckage.

[GuineaPig]

A couple more of my strategies:

- I find that for regular trunk lines (the ones that are busiest) selecting locos with higher speed isn't as important as selecting ones with good acceleration

- I'm actually OK with running main lines through cities; I build a bypass for freight and waypoint them around so that they don't clog up the urban area. If there's room, I'll also build a turning loop or two so that trains don't have to turn around in the station; they can continue going straight, then loop back (preferably with a right hand turn) onto the main line

- By the time I have enough cash, I try and run most of my long-distance pax/mail on dedicated, flattened lines that run as straight as possible

- Building multiple stations in big cities is key. I run them like how Paris, London, etc. run in real life: I have different terminals for different directions of travel

I was experimenting on the England map. I built radiating outwards from London: one station in the south had a through-track that handled all north-east and southern local traffic. Another station in the west handled all west-bound local and freight traffic. A third one in the north was the terminus for my north-south high speed line.

I built two principal lines: London-Oxford-Birmingham-Manchester/Liverpool, and London-Leicester-Leeds-Newcastle-Edinburgh-Glasgow. There were bypasses around some of the major cities, most notably Birmingham which had previously been incredibly congested and needed two stations, one for the high-speed line. Branch lines connected London to Exeter and Southampton, and Edinburgh to Inverness. Then I built a high speed line that connected London-Newcastle-Edinburgh, with a branch to Birmingham (dedicated stations for each).

Long story short, I was able to handle 100+ trains with very little congestion. The real difference this time was just planning ahead for rights of way and routes for the trunk lines. Also, the geography of the map helped too, because I was able to run the high speed line in between the two axes of the other trunk lines without interference or a need for a link across the high speed line where it wasn't diagonal.

[HeWhoHunts]

A few rules

Multiple Stations in the same city: A city can have multiple stations at it, for example at a busy city I will often make one station for freight and another for passengers/mail.

Quadruple track / multiple routes: this seems to be one that escapes a lot of newbies but setting up a second layer of track next to your initial / main line can drastically cut down on traffic, just make sure you use...

Waypoints: use them! They are fantastic for diverting trains away from congested lines, trains automatically take the shortest route possible, even if there is another route available which is only a tiny fraction bit longer, however setting a waypoint will ensure that trains get off the busy mainline and still get to their destinations at roughly the same time.

Priority: priority can speed things up too, in most of my games I have all my freight trains set to "slow" my regular passenger/mail trains to "normal" and my transcontinental passenger/dinner cart trains set to "express". (I tend to play before 1900 so passengers are worth the most, this exact example might not apply to modern set ups but you get the idea, setting priority can cut down on congestion)

Branching: Branching off the mainline to stations is a big one, although there are times, especially at the beginning, when making direct lines is vital. However when you're up and rolling try to keep stations off the main railways.

[jeffryfisher]

Multiple Stations in the same city: A city can have multiple stations at it, for example at a busy city I will often make one station for freight and another for passengers/mail.

This is especially important if a city has multiple industries that demand the same input. If you want your grain to be baked into bread, then you don't want a giant terminal sending it to the distillery across town. Therefore you may want multiple freight stations, the extras being tiny depots that target specific clusters of industries. The output can still be picked up at a main-line station.

Other reasons for multiple stations:

* Heavy Lifting: Deliver iron and/or coal at sea level and haul away steel from the top of the continental plateau that you'd rather not climb.

* Extra demand: The game's demand model is broken. Instead of demand being figured for buildings and then seen by stations (the way loads are), each station figures demand independently, without regard for deliveries that may be flooding neighboring stations. If you really need the income, then you may spread your deliveries simply by building more "doors" to the same factory/city.

* Avoid crossing other tracks: If some of your coal is mined in the south, and other coal is mined in the north, then you may avoid crossing your E-W main-line by having separate north & south terminals. Just be careful to select the right one.

Waypoints: use them! They are fantastic for diverting trains away from congested lines, trains automatically take the shortest route possible, even if there is another route available which is only a tiny fraction bit longer, however setting a waypoint will ensure that trains get off the busy mainline and still get to their destinations at roughly the same time.

Even worse is when two routes are exactly equal. It appears to me as if trains will always default to veering LEFT, causing them to cross oncoming traffic, probably twice (the second being when the routes come back together).

Waypoints are useful when applicable. Unfortunately, at least at the screen resolution I have, much of my track is either hidden behind station-stars or too tangled to select reliably (even when holding the control key). The route selection map needs a few levels of zoom so that fine and ultra-fine waypoint selections can be made, but we will probably never get another patch to this old game (I can't even get the copyright owner to talk to me to find out if the source code even exists anymore).

[Gwizz]

The source code was used in some other games. It may still be. So it may be a while before it becomes shareable.

[jeffryfisher]

The source code was used in some other games. It may still be. So it may be a while before it becomes shareable.

It may still be possible to fork development from the v1.56 patch (but only if the owners will negotiate, and right now they won't even do me the courtesy of saying "no, not interested").

[HeWhoHunts]

This is especially important if a city has multiple industries that demand the same input. If you want your grain to be baked into bread, then you don't want a giant terminal sending it to the distillery across town. Therefore you may want multiple freight stations, the extras being tiny depots that target specific clusters of industries. The output can still be picked up at a main-line station.

Other reasons for multiple stations:

* Heavy Lifting: Deliver iron and/or coal at sea level and haul away steel from the top of the continental plateau that you'd rather not climb.

* Extra demand: The game's demand model is broken. Instead of demand being figured for buildings and then seen by stations (the way loads are), each station figures demand independently, without regard for deliveries that may be flooding neighboring stations. If you really need the income, then you may spread your deliveries simply by building more "doors" to the same factory/city.

* Avoid crossing other tracks: If some of your coal is mined in the south, and other coal is mined in the north, then you may avoid crossing your E-W main-line by having separate north & south terminals. Just be careful to select the right one.

Excellent point.

I once had 3 stations at a city for a steel to goods operation, one station really far out of the way for the coal/Iron drop off, another for the steel pick up, and a 3rd a little ways away for the tool and die factory. (this particular map had one of the only Tool and die factories and steel mills in the same city) I also had a 4th station for regular passenger/mail operations.

[akuenzi]

... but we will probably never get another patch to this old game (I can't even get the copyright owner to talk to me to find out if the source code even exists anymore).

Jeff,

I once exchanged a couple emails with the original programmer of the game, Phil Steinmeyer. I had a question on the train acceleration formula. He actually replied, which was cool even though he wasn't able to answer my question. Do you suppose he might know if the source code is still around? And even if it was, like someone else pointed out, it might not be obtainable if it was used in other games.

So here's a question -- is there anyone that visits this forum that knows how to program? I spent a lot of time when I was in college playing Transport Tycoon, and I've been amazed at the longevity of that game. Now my kids even play that game. Or rather, not the original game, but 'Open' Transport Tycoon, which I understand was reverse-engineer-programmed and built from scratch by fans of the original game. I'm also amazed at just how vibrant of a community remains for that game, even though it is older still than Railroad Tycoon II. I just checked one of its websites (http://www.openttd.org/en/), and they have a recent post in November 2011 with yet a newer build of the game. It's obvious that a whole bunch of folks still care about that game and have kept it going.

I'm sure you're all familiar with that game, and the legacy that has followed it. The question I have is, could the same thing be done by those of us that like Railroad Tycoon II? If I knew how to program computers, I'd already be working on it. But I don't. I'm just a dumb accountant. :) There are a few things I could contribute perhaps in the financial model, but I don't know how to program. Does anyone else that trips through here have both the willingness and the skill... and the sense of organization to pull something like this off, in a collaborative effort? If Mr. Steinmeyer answered a couple emails, I wonder what he'd do with an appeal that asked for help in trying to program a new one, where we could do what Open TTD did, adding new things that everyone likes, or fixing old problems that everyone points out. I'm sure it's not that any one of us want to steal anything or breach copyrights. Goodness knows, every time a new railroad game comes out, it's not long before it's purchased and in my collection. The simple problem is that we like a game that is no longer maintained. It's practically abandonware.

Thoughts anyone?

[jeffryfisher]

I once exchanged a couple emails with the original programmer of the game, Phil Steinmeyer.

So have I, within the last month even. I was pleased to get a reply, but he couldn't answer many of my questions either.

Do you suppose he might know if the source code is still around?

I asked about the very thing. All he could tell me was that he sold out his company to Take2 Software. Sadly, Take2 won't answer my inquiries.

And even if it was, like someone else pointed out, it might not be obtainable if it was used in other games.

Deals can be struck, depending on what one wants to do. Since I want to build a US History teaching tool rather than competing against the latest, most beautiful graphics, I could get a pass... if I could get an answer. Right now, Take2 will not even tell me that they hold a copyright. Take2 won't even deign to answer my voice mail (yes, I got tired of lost email and called on the phone).

is there anyone that visits this forum that knows how to program?

That's me, and that's why I've been trying to get at the source. If Take2 would bundle the last platinum patch in a GNU license, then they could probably claim more in tax write-offs than if they sold what's now about 10-yr-old source code.

...'Open' Transport Tycoon, which I understand was reverse-engineer-programmed and built from scratch by fans of the original game.... could the same thing be done by those of us that like Railroad Tycoon II?

There are aspects of the game (like graphics, both images and manipulation) that I would not want to touch with a ten-foot pole. I just want to patch a few bugs, add some data features (like filtering the train list to see "problems"; replacing many engines at one go; adding a few functions to the scenario editor; that sort of thing), and improve some AI logic. I'd settle for simply fixing the bugs (e.g. halving grain production in one territory works on the one territory, but if the effect is "temporary", its eventual expiration boosts grain production in ALL territories).

We would really really want someone like Phil to "discover" an old backup of the v1.56 patch to use as a starting point. Unfortunately, Phil didn't rise to my suggestion that he could do so.

If Mr. Steinmeyer answered a couple emails, I wonder what he'd do with an appeal that asked for help in trying to program a new one

He gave me the impression that he had washed his hands of RRT2 and didn't like to be bothered about it. It was courteous of him to tell me where his copyright interest had gone. I don't intend to bug him anymore.

[akuenzi]

That's me, and that's why I've been trying to get at the source. If Take2 would bundle the last platinum patch in a GNU license, then they could probably claim more in tax write-offs than if they sold what's now about 10-yr-old source code.

There are aspects of the game (like graphics, both images and manipulation) that I would not want to touch with a ten-foot pole. I just want to patch a few bugs, add some data features (like filtering the train list to see "problems"; replacing many engines at one go; adding a few functions to the scenario editor; that sort of thing), and improve some AI logic. I'd settle for simply fixing the bugs (e.g. halving grain production in one territory works on the one territory, but if the effect is "temporary", its eventual expiration boosts grain production in ALL territories).

If you'll pardon an admittedly naive question, how hard would the game be to reverse engineer? If the source code isn't available, is that the only option? Or, is it really realistically possible that you could get your hands on it by cutting some arrangement with the current copyright holder? It sure would be nice to start wtih something! Do you think it would help to have a couple others appeal to their sense of generosity? :)

I agree with you -- I wouldn't want to change the graphics or interface either, but I imagine we all have our own 'wish list' of things we'd change about the game. One thing I'd love to change is the passage of time in the game, or to at least gear it to the size of the map... and that's only a start. I'd love to customize a few other things, too!

[Gwizz]

Phil took the RRT2 game engine to its' limit. RT3 had a new game engine and Take2 pushed the development team to their limit, never letting RRT3 reach its full potential. RRT3 went out the door before it was finished.

The team spent many months developing RRT3.

I used the RRT2 game engine to build an educational tool for business and the stock market classes.

I tested the tool at the high school level and was told by the superintendent that it was too much like a game.

I never had access to the code. Poptop made the adjustments I needed.

But, not enough to take the look of a game out of the proposed educational tool.

I was told the code was used in a number of games and for that reason the code had to take priority to my needs. When I tried to change something the game engine would break and Poptop would end up fixing it for me.

I would like to have a part in a remake of a better RRT2.

As out of date as they are, I still have my programing tools. But then, RRT2 is out of date, so maybe that is not such a big deal

[CrownVic95]

Phil took the RRT2 game engine to its' limit. RT3 had a new game engine and Take2 pushed the development team to their limit, never letting RRT3 reach its full potential. RRT3 went out the door before it was finished.

The team spent many months developing RRT3.

I used the RRT2 game engine to build an educational tool for business and the stock market classes.

I tested the tool at the high school level and was told by the superintendent that it was too much like a game.

I never had access to the code. Poptop made the adjustments I needed.

But, not enough to take the look of a game out of the proposed educational tool.

I was told the code was used in a number of games and for that reason the code had to take priority to my needs. When I tried to change something the game engine would break and Poptop would end up fixing it for me.

I would like to have a part in a remake of a better RRT2.

As out of date as they are, I still have my programing tools. But then, RRT2 is out of date, so maybe that is not such a big deal

So would I, as RRT3 was not the answer.

I've even thought about setting out to write my own new game, but haven't yet come close to gathering the ambition to actually begin such a major undertaking. I wrote a bunch of games for the Commodore64 back in the 80s, but am solidly rusty today. I would be old and grey before....wait a minute....I'm already old and grey!

Ideally, I'd love to see a RRT II/RRT3 hybrid modded to fix the bugs in both in the process.

[HeWhoHunts]

One things from RRT3 I would love to see is the ability to build your own factories/plants.

The ability to group trains and some kind of "train search" feature would help immensely, also an improved way point feature (so you don't need 500/500 vision to see all the tracks) would not go astray.

I never thought about a scaling time feature based on map size but that would kinda make sense... also bigger maps are always welcomed :D

[jeffryfisher]

If you'll pardon an admittedly naive question, how hard would the game be to reverse engineer?

That depends on whether you have someone on board who has built something like it before. To me, the moving graphics are heap-big magic. Heck, just the art work looks impossible. Routing looks difficult but solvable. My strengths would be in data structures and functional logic.

If the source code isn't available, is that the only option?

Essentially yes, we would build an imitation of an old game and then give it a new name to avoid trademark infringement. To do so, we would need a lot of labor, including some precious specialized skills, and we would have no money to attract those resources.

I tried such a project once before (when the original Airport Tycoon looked so promising but had such annoying bugs). The other participants were eager to generate feature lists, but their enthusiasm drained away when I started educating them on object-oriented design and software project management. They started to see how much real WORK would be involved, and the project stalled.

Or, is it really realistically possible that you could get your hands on it by cutting some arrangement with the current copyright holder?

It becomes more realistic every year as long as the code hasn't been lost. Unfortunately, Take2 won't even talk. I have read receipts from copyright@take2games.com, but no reply in about two years of calling and writing. Because of their (I think rude) silence, I don't even know if the source code still exists or if they still own the copyright.

Do you think it would help to have a couple others appeal to their sense of generosity?

Not at all. Since they're not even generous enough to hit the reply button to a businesslike email and tell me they do or do not own the copyright, I'm sure they're not generous enough to even go through the motions required to release a GNU license unless there's an incentive. A healthy tax break might do it, but an emotional appeal is unlikely to move them.

I imagine we all have our own 'wish list' of things we'd change about the game. One thing I'd love to change is the passage of time in the game, or to at least gear it to the size of the map... and that's only a start. I'd love to customize a few other things, too!

I think there's a wish list thread around here somewhere. Find it and resurrect it with your own ideas.

[jeffryfisher]

I used the RRT2 game engine to build an educational tool for business and the stock market classes.

Do you have any tech docs describing the inner working of the economic model or stock market model? One of my motives for seeing the source isn't just to change the code, but to answer some nagging questions about undocumented features, especially pertaining to scenario editing.

For instance: Exactly which parameters can be set for one territory or one company? Exactly how does mapping one territory to another change the way effects work on it? Looking at the algorithms and data structures would show me how each type of change (production, revenue etc) was being handled, and then I would not be blind-sided by unexpected results when I play my map.

For instance, if I grant train and track rights to one company in one territory, then that company gets those rights in every "sub" territory that maps to it. However, if I boost troop revenue in the same territory, then its sub-territories do not get the boost; I must affect them separately.

To make matters more confusing, some effects seem to turn on whether the right-click is mapped or whether a border is displayed. These subtleties are not documented in the manual for the map editor, so I am left to poke and prod at what is essentially a black box. If I had been organized, then I would have collected my discoveries together in a secondary editor manual. Unfortunately, my discoveries are scattered through out many of my rants on this forum.

[akuenzi]

In terms of reverse engineering the game, it's obvious that the Transport Tycoon folks were successful with what they created. Maybe they have a larger pool of folks to draw upon in terms of programming talent, or maybe TTD is a simpler structure to program. Maybe both. I went onto Open TTD's forum and did a search for 'Railroad Tycoon,' and it comes up quite a number of times. Undoubtedly there are 'lessons learned' that those guys went through that would be applicable here. Of course doing this is 'real work,' and the person/people involved would need to have a true passion (not to mention time!) for this, or else it won't ever get done. I wonder if anyone over on Open TTD would be interested in taking on this challenge? Of course, I'm just 'talking out loud' here, and maybe dreaming delusionally. :) If I had programming skill, I'd already be working on this. When I was a kid, I spent hours upon hours making my own board games. What works on paper doesn't quite translate to the electronic realm, unfortunately. Think it would be worth a few inquiries over on the Open TTD board to see if anyone would bite?

[Gwizz]

Do you use the ctrl key to help you place a way point more exactly?

[HeWhoHunts]

Would it be possible to purchase the source code?