Help weaning off micromanaging

[rkagerer]

I'm looking for tips to improve my game on Railroad Tycoon 2.  Especially strategies for effectively managing 25+ trains.

I've found I'm quite successful at the beginning when I aggressively manage each trip for my handful of trains.  My strategy tends to be: Pick up something from current station, take it to someplace far away where it's in demand.  I try to favor destinations where there are valuable goods waiting to be returned for backhaul or carried someplace else.

I wind up scheduling "one-way routes" for my trains.  The game doesn't give you an easy way to this, but I instruct the train to HOLD at its final destination by making the last consist be nothing but a single cargo car which I know the destination won't have, and setting the red light, making it wait for the never-available load.  For consistency I prefer to pick one resource to use for this purpose.  e.g. In my last game most of my stations weren't near milk-producers, so I could effectively ignore it as a resource.  When my fleet grew to more than a handful of engines, I would periodically scan the list for single-milkers "waiting for cargo" and assign them their next duty.  This gets unwieldy once you have like 25+ trains.

I've tried to experiment with more "set-and-forget" approaches.  e.g. Ferrying passengers back and forth between two big, reasonably distant cities can be a reliable income-generating route that can be left alone for a while.  But inevitably the profitability on even that runs out (especially in a recession) meaning you have to go back and tinker - e.g. reduce cargo cars, or reassign to new routes.

I'm having trouble finding the right balance where you're delivering goods in line with the rate they're produced / consumed.  Then a recession or boom hits and you need to rejig everything.  I've found the yellow "half-load" light helpful for keeping trains running at the beginning of a recession, but beyond that feel like it's just a band-aid.  Once poor times go on long enough those less profitable routes need to be addressed.  And in boom times it's a mad rush to add more trains and I find myself wanting to buy "single-use, disposable" engines to clear out the excess - since the revenue from the first trip dwarfs the cost of the engine and unlike waiting for an existing car to arrive the new engine appears instantly where you need it.

In my last game I tried setting up a circular "regional" route where a few set-and-forget engines delivered nearby iron and coal to a steel producer.  With enough tweaking I managed to keep my factory pretty well fed, but I found the transport network wasn't very profitable on its own.  I tried to be clever and have the trains back-haul the final goods, but that quickly became pointless as the prices were driven down.  The sometimes-mixed loads between stations was also complex to manage.

Lately I've been thinking of trying "hub-and-spoke" regional systems that collect cargo and pile it up at a node (e.g. small station with no demand for anything), then setting up long-haulers from those regional nodes (or running the bigger loads by hand to wherever is the most lucrative demand).  Has anyone tried anything like that?

What strategies and tactics do you use for migrating from a small, micro-managed outfit to a fully functional fleet where you're not having to pay so much attention to individual trains?

• Do you have lots of trains with dedicated routes, and let them just sit and wait for cargo until it's ready?
• Do you tend to use short trains to haul individual loads back and forth, or long trains to dump a pile of stuff where prices are high?
• Do you carry mixed loads or dedicate trains to a single resource?
• Do your routes visit several stations accumulating goods along the way, or are they more point-to-point?  If the former, how do you avoid dropping goods prematurely when passing through stations which demand them but don't offer good prices or aren't where you want to deliver to?  (Do you have to route around such stations?)
• Do you relay cargo through "depot" nodes?

Also some specific questions:

• How does redirecting a train in-flight impact profitability?  (e.g. Passengers don't mind if you take them someplace other than the advertised destination?  I could have sworn one time I saw my gold bar for the load go UP after redirecting the train to a CLOSER location - I assume the new destination gave a better calculated distance/time tradeoff or something??)
• What happens to a car's value if you drop it off without selling it (assuming you have the necessary building e.g. warehouse, refrigerator)?  Does it slowly rot or is the value frozen and the goods just disappear at some point?  Does the game track each individual car's value as a function of age (reduces it) and total distance traveled (presumably increases it)?

Any other advice?

[jeffryfisher]

Wow, lots there. First: Look through our old threads.

Second: Never delete an engine. The game has bugs in it that sometimes don't handle deletions well.

In my games, I micro-manage only until I have all double-track. Then I change over to twice-per-year scanning for problems (like empty trains waiting for cargo).

My routes are all set-and-forget (until I see a problem). During recessions I set a few to only half-full, but I also put up with a lot of idle trains. The game has another "feature" that brings the houses within a city into sync with one another, so they tend to produce lots of passengers and mail one month, and then go dry for months (sometimes years). If you see this happen (especially several decades / economic cycles into a map), then you just have to leave trains waiting for the next flood.

I tend to under-serve most routes a little bit so that my trains stay full and making money during average economic times. Except during the initial phase of a game, I resist the urge to buy all of the trains that a boom can fill.

I like my pax and mail to roll fast, so I keep them short. Before the 4-6-2 Pacific, my express trains are only 2 cars. After that, they are 3. Only during boom times with later, more powerful engines do I make some longer (and after the boom, I shorten them back to 3).

When cargo is left at a depot, it continues to age down to its minimum value. I no longer depot pax or mail.

For industrial loads in the complex economy, I mostly have each train work a production sequence (e.g. Sugar -> Food -> [import] -> etc). At the bottom end though, I'll have some heavy, slow trains just pile on something like coal and/or ore to create a ready supply of steel for other, faster trains to transform into goods and autos as needed.

I do set some routes to accumulate cargo through multiple stops (e.g. three wood sources on the way to a mill). These are usually rural, so (on most maps) there's little or no chance for an industry to pop up and intercept my cargo. However, I have been known to do this with villages that supply pax/mail without demanding them (e.g. climb the steep grade out of San Francisco with only a single carload and then add a car to be filled in Reno / Elko / Winnemucca). When those villages grow into towns, such routes are suddenly buggered.

 

[rkagerer]

Thanks @FedaYkin for that extensive advice, it's super helpful.

What's a "normal" number of trains for you to have midgame?

I'm playing the Great Divide and am up to 60 now.  I've gotten a lot better at managing them (rather, managing them less and just letting them run).  Not stressing anymore about cleaning out all the waiting passengers was a game-changer.  Now I aim to keep the backlog at around 2-5 in cities on my main lines, and don't worry about adding capacity 'till it creeps above that.  I've spread out my passenger routes more strategically to avoid exhausting supply at individual cities - especially those lucrative ones at the edges of the map.  I mainly reserve those passengers for the longest-distance connections, then set up routes between my other cities (matching up cities by size and ensuring routes are reasonably long).  I do keep a few shorter-distance routes around to supplement cashflow while waiting for the long-distance cash cows to arrive (e.g. between Toronto, Montreal and Quebec).

For the moment I'm reloading if the economy crashes, to give myself time to learn how to nurture a more automated game.  I've also avoided venturing too much into other types of cargo for now, until all these passenger routes are running smoothly.   It's 1875 on this map and Vancouver is connected.  There's a big bottleneck through the pass from Quebec to Fredericton due to track laid before I learned about the "grading tool" ;-).  All in all the game is starting to get more fun again.

Is there a hotkey to expand / collapse the train list?  (I find that more helpful than the other list, which doesn't fit as many entries on the screen).

I can't imagine what it would be like trying to manage hundreds of trains through these basic listboxes.  Some filtering options would be nice.  I actually made a spreadsheet to keep track of my routes (yeah I'm feeling like a nerd).  Has anyone made any tools to pull data out of savegames and present it in a more workable format (kind of like CivAssist)?

Edited October 25, 2018 by rkagerer

[jeffryfisher]

9 hours ago, rkagerer said:

I can't imagine what it would be like trying to manage hundreds of trains through these basic listboxes. Some filtering options would be nice.

That, and once filtered and sorted, I wish I had a way to select a bunch of trains and replace all their engines in one fell swoop.

There are some GUI-automation programs (e.g. auto-hotkey) out there that can tie a hotkey to one or more mouse clicks. We had a thread about them a couple years ago. I almost got an engine-replacement script to work, but it had no intelligence. It would just grind through a fixed number of trains replacing engines. Given the RT2 bug that hoses any train whose engine is replaced while in a station, I gave it up. YMMV.

[rkagerer]

Cool.  I wrote a hotseat turn "automator" for Civ3 way back when.  It would detect when you end your turn, then zip and ship the savegame to the next player via email.  On the other end it would spin up a VM running the game and make Civ load the file, so a player could simply double-click the email attachment and immediately start playing their turn.  It worked by hooking the game window and watching for small rectangles of pixels (at known coordinates) to figure out when dialogs were shown, certain text was present, etc.  I took full screenshots of the game then used Paint to lasoo a snippet and grab coordinates for whatever image I was interested in detecting.  It worked for Civ because the graphics (at least the ones I was interested in) are pretty "static".  Not sure if it would for this game but could be worth a shot.  Happy to share the code if you want it - I think it was done back in my VB days.  In fact I uploaded it to a now-defunct fan site and a guy contacted me interested in getting the source and maintaining it (since I didn't have time to myself).  Not sure if they ever ran with it and refined it further.  Let me know if you want to have a peek.  IIRC the code documentation is gentle - even includes diagrams.

Back to this game... What's up with the "never delete a train" and "never upgrade the engine while in a station" rules?  Could you point me to some more detail on those bugs?  Do they exist in the Platinum version with the 1.56 patch?  (Is there some unofficial 1.57 patch does it help?)

Edited October 25, 2018 by rkagerer
Ask about train deletion / patches

[jeffryfisher]

18 hours ago, rkagerer said:

What's up with the "never delete a train" and "never upgrade the engine while in a station" rules?  Could you point me to some more detail on those bugs?  Do they exist in the Platinum version with the 1.56 patch?  (Is there some unofficial 1.57 patch does it help?)

I've only played v1.56 (and self-modded derivatives thereof), so all my notes come from that version.

The game definitely has a bug in the map editor where deleting an event causes neighboring events to get their trigger and/or effects crossed up. NEVER do that.

I have a note-to-self from maybe ten years ago warning myself to never delete trains, only replace engines. It comes from so long ago that I can't remember what caused me to write that. Maybe it was just that deletions would renumber all higher-numbered trains so that my to-do notes on a game would become confusing (things like "fix T283's route after next delivery" would no longer make sense after it became T279). On my US History map, I typically run 400+ trains from 1880 to present and beyond, so train renumbering alone is enough reason not to delete any.

Replacing an engine on a train waiting in a station (or while reversing over a switch) will result in overlapping engine / tender / cars. The train may then become a runaway, speeding along its route without stopping at stations or exchanging cargo. The only escape is to either replace the engine en route (costs $) or give the train only one station with an empty consist (and even then, a folded steam-engine + tender might not be straightened out).

The home-grown 1.57 patch fixes some broken internal data. For example, I think it fills in some missing demand recovery parameters for factory inputs that would become zero-demand after first delivery and then never recover. We know how to use hexadecimal editors to fiddle with bytes in many of the data tables within the RT2 EXE file. However, we can't change program logic, so logic errors remain beyond our grasp. If I knew more about decompilation and DLLs, maybe I could write some replacement functions (as someone did for a couple bugs in SimCity 4), but that's a little bit beyond my tool kit.

[MaglevForever]

Micro- can give best profits, I played for years that way. But, finally I managed to relax my play a bit more and dabble in automation.

 

Simplification was how I went about it. For express traffic I look for pairs of cities.  These should be of about equal size and as far apart as practically possible, connected by as straight track as practically possible with no grades on the turns. I will go for max carriages, mix of mail and passengers. Wait to fill 4. For 4-house towns it will be wait for 3 out of 5.

 

For freight, it's better not to mix cargoes. Highest revenue is achieved by dumping a full trainload of one cargo type.

 

At this moment I would summarize 3 overall strategies for set-and-forget freight.

 

1. Use it just to provide volume to push city growth. Long distance express revenue is so much higher. Best use in the 19th century. Rising running costs and rot factor make this less practical for a modern game.

 

2. Industrialize. Buy the factories, and run as short and cheap service, both resource in and finished good out, with the maximum of volume. Focus on the industries. Industrial profits are perhaps a little less, but more stable over an economic cycle.

 

3. Work the chains as Jeffry suggested. The money maker is hauling finished goods a moderate distance. Manage those end-product demands as first priority. You are in control of supply via how many resources you hook up. Resource hauls should be as short as possible, minimize costs. This is the place to use wait-till-full trains. Then when converted you have a full trainload of say Grain -> Food. Often I will repeat the resource section, but alternate destinations. Food -> City A followed by Food -> City B. This gives time for demand to recover.

 

The typical map has heaps of resources that are just for eye candy. Automation is accepting this is ok and for #3 to ignore even more of them. I have a plan for distribution ahead of time before connecting a new resource.

 

A rare few cargoes*, including Food can work as long distance. But said journeys in the 20th century are best if not dead-headed on the way back. That means combining with a different chain. This makes the route more complex, but after a bit of practice it can be setup without headache.

 

*There is some confusion in the documentation about distance factor.  https://forum.dune2k.com/topic/23923-cargo-data-for-v156-ripped-from-exe-file/?do=findComment&comment=395629

At the moment the figures I trust most are in the data included with the download of Jeffry's US History map.

 

PS. Sorry to say I haven't found time and/or worked out how to do the check I mentioned in the link.

[jeffryfisher]

Also beware of "efficiency" dilemmas: You might think you are perfecting your routes by micro-managing pickups, but if you spam your mainlines with new trains, you could lose more to traffic jams than you gained in your stations.

Finally, I enjoy seeing a map evolve over many decades as cities grow and tech advances. If I micro-manage, then that takes too long. If I create too many trains, then the engine-replacement years are too painful. So I give up some business opportunities in order to make the game more entertaining. I know, it's an act of will for anyone with OCD, but I make it happen because I know the payoff.