Engine Age and Replacement

[jeffryfisher]

After a few months away playing Minecraft, I am shifting some attention back to long-time favorites like RR Tycoon. It has been quiet around here without my agitation, so I'll try to stir the pot.

My question of the day (or maybe week): At what age do you like to replace your engines for the sake of age alone (i.e. even with no new tech)?

How long might you stretch an aging batch of engines to catch a new type you know is about to be introduced? Do you "run the numbers", or do you "have a feel"? Does the tedium of replacing engines for a large number of trains affect your decision?

Footnote: There's another thread around here about GUI-scripting automated engine replacement, so the tedium of large numbers can be overcome.

[jeffryfisher]

Answering my own questions:

I have a rule of thumb that says replace engines at about 15 years. I'll replace them sooner if a newer, faster engine comes along, and I'll delay engine replacement / new trains for about 5 years when expecting a "significant" new engine type. I'll sometimes look at the annual maintenance cost difference between new and old engines and calculate whether a new engine can pay for itself, but I usually end up keeping aging engines longer than they're worth simply because it's such a pain to walk through 400 trains replacing some and skipping others.

I'm trying to develop some engine-batch methods that will facilitate wholesale scripting. I've written a script that will replace 'X' engines starting at engine #Y.

And yes, the wish-list thread includes my request for advanced train list filters and select-all / replace -selected functions.

[outofmage]

Happy to see your posts again after so long, jeff.

 

I replace the engines on a planned pace. I usually take a peek in the Editor to see which engines are available for this map before I play, pick my favorite ones to cover the playing years. Then I look at the time span each engine would serve and 'cut' each engine's serving period into reasonable pieces depends on their reliability and costs. 

 

For example, in North America map I currently play on. I plan to play from 1830 to at least 2000. These are engines I picked and their serving periods.

passenger trains:

1831-1837 John Bull        no rep.

1837-1848 Prussian        1842  rep.

1848-1868 America-C     1858  rep.  

1868-1902   8-wheel        1880 rep.  1892 rep.

1902-1935   Atlantic       1918 rep.

1935-1990   GG1         1952 rep.    1970 rep.

1990-forever   E111         every 15 years

 

fright trains:

1877-1905   consolidation   1886  rep.  1895 rep.

1905-1919   Camelback       no rep.

1919-1935   Mikado       no rep.

1935-1990   GG1         1952 rep.   1970 rep.

1990-forever   E111         every 15 years

 

I plan these ahead and think about pros and cons against each choices during a period, then I can do a replacement plan. I have trouble choosing between consolidation and Mogul for freight trains in 1895 replacement, I believe the reliability of consolidation can make up the speed disadvantage to Mogul. Over all, I have a strong lean to reliability because I anticipate a big game and the tracks must be busy, so I cannot stand a broken train in the way. I also choose Atlantic over Pacific between 1918-1935 for the same reason.

[jeffryfisher]

Ah, now there are some differences in play style. I am way too lazy to replace my Prussians in the middle of a 10-year period. Even though they're unreliable, they're slow enough that their breakdown chance is still low. Also, that early in the game, I am expanding rapidly, so my track far outreaches the population of towns and cities (and therefore the pax and mail being carried). With uncrowded tracks, breakdowns are just one train's loss of speed, not a major catastrophe.

I also tend to build parallel mainlines to relieve congestion so that even later traffic can usually tolerate (recover from) a temporary blockage. Therefore, I jump on those Pacifics when they become available and learn to live with the breakdowns. I still use Atlantics in that time period, but only for freight on level track (the Atlantics have a quick acceleration, but they grind to a halt on hills -- almost as bad as breaking down 100% of the time).

Full disclosure: My US History map has events allowing me to spend piles of extra money to improve my company's safety (periods of doubling track maintenance costs to make improvements). Therefore, the Pacifics are only irritating, not completely useless.

In the end-game, my US History map enables an event pathway to Eurostars etc (even though they appear to be "off" in the editor), so they haul my pax while E111's usually haul freight. The radical difference in speed presents difficult decisions. For some "fast" freight, I have made profit hauling with Eurostars. For bulk freight, I try as much as possible to build dedicated track where slow trains won't be overrun by fast, higher-priority ones. This strategy presents its own challenges:

1) Squeezing extra rails into sometimes tight geography (a fun puzzle that I usually enjoy) and

2) Tricking fast and slow trains into not comingling on each other's intended tracks. This puzzle is often frustrating because "leaks" don't always appear right away, and they're not always obvious on a big map. I've experimented with using diesels for bulk freight and using electric/non-electric track for separation, but it's hard to walk away from the E111 for my heavy trains. That's why my wishlist thread includes a plea for high-definition waypoint placement (the ability to create and adjust a train's waypoints while viewing the main map fully zoomed in with the grid on).

I've thought of building small extra stations on difficult-to-select ramps so that I'd get a star to pick on the route-map. I'd put up with a brief unnecessary stop in order to avoid worse. Unfortunately, the situations needing such extreme measures usually don't have room for a station with 3 track segments.

[MaglevForever]

Yeah, I have been taking a break from RTII for a while also. Did either of you guys play Cascadia yet?

 

Honestly, I am too much of a control freak to enjoy playing very large maps like you guys play. I made it through US History by playing as Mexico since I could leave the game on auto-pilot most of the time, but am still stalled on my latest attempt. I am a control freak with my railway and feel that the auto-pilot style is so wasteful of true revenue potential. I guess I don't have the patience or management-oriented mind to balance all the thousands of things that go on at once. Well done, that you guys can do it and enjoy it!

 

In light of this I don't have a set plan for when to replace engines. Other than replacements to upgrade to better locos I don't often do mass replacements. When I do I will be watching the maintenance costs of a train that runs hard and also watching for breakdowns, both percentages when working hard and approximate frequency. So to answer your question, I don't have a calculated plan, more of a "feeling" approach.

 

If I have a high traffic corridor I will sometimes replace engines that use it far earlier than those on the rest of my network. This is especially true of the Pacific which I use. I also have the opinion that, at least the way I play, it is a better choice than the Atlantic. I also tend to run mainly the same type of loco for passengers and freight as in shorter maps I don't have the cash or patience to make dedicated tracks. This is the simple option and enables me to leave most of my trains on normal priority.

[noctilucus]

After several years without RT II, I've finally picked up the platinum edition...

 

I usually replace my trains based on "the numbers", looking mostly at the chances of breakdown and overall speed. And of course how much cash I have on hand, which seems to be quite the bottleneck in most Second Century campaigns.

[jeffryfisher]

Did either of you guys play Cascadia yet?

I've downloaded it but not played it. Is it special?

 

Honestly, I am too much of a control freak

As am I until my company has something like 50-100 trains. In fact, at the very beginning, I'll build single track and watch over my half-dozen trains, doubling track just in time to save each collision.

However, by the time I have multiple arrivals each month, if I've been playing right, then my routes are well balanced with a slight surplus of pax and mail to keep my express trains full most of the time. From then on, I do not even bother looking up the loads available for each arriving train. Instead, I run off a half year at a time, scrolling through the train list each January and June to look for trains waiting for full loads.

That "release" frees me to think about other, more strategic aspects of the game. I guess I am just not smart enough to maintain a chain of strategic reasoning that is being interrupted 10 times per game-month by train arrivals.

 

Other than replacements to upgrade to better locos I don't often do mass replacements.

Likewise, most of my mass replacements are upgrades. However, I hate allowing engines to age to 25+ years, so when there are larger gaps between new best-engine types, I'll pick a year half-way to turn most of them over.

One of the bits of tedium that I forgot to mention before: If one replaces an engine on a train that is waiting for cargo or reversing direction ("folded"), then a bug in the program will give you a jammed up train with cars and/or coal tender overlapping. I have also seen trains with big gaps between cars. Some of those trains then run through their assigned stops without exchanging cargo, which means that they cease to earn any revenue.

All that is a long way of saying that a bug in the game forces me to eyeball every train before I replace its engine. I must skip those that are vulnerable, writing down their train numbers and revisiting them later.

[noctilucus]

I never knew about that bug so either I've been lucky or I've not been paying attention to a few trains no longer earning revenue (which doesn't always jump out on longer routes)

 

I remember there are a few cases where I'd let my trains run for more than 25 years simply because no better model becomes available in that period. One example is running non-electric tracks with the GP18 which outperforms most newer trains for the next 20 years in terms of reliability, speed (especially at grades) and yet reasonable running costs. Somehow those 25 year old engines still run with a 1-2% breakdown chance...

[MaglevForever]

I've downloaded it but not played it. Is it special?

To me it is. One of my all-time favorite maps. I think it is a good challenge with an advanced military supply chain using the ports, management of fuel costs, track laying in the mountainous terrain, and nice events (including engine introduction events, supplying/building a dam that really fills with water and haulage based track unit allocation). This map really favors an efficient style and that's my thing I guess. Anyway, I asked because you said you were planning to try it.

 

When I see inefficiency on auto-pilot mode (whether it is a couple of months or half-yearly like you said) I tend to be annoyed by it and feel that I should do something about it. Trying to fix it causes me to chase my tail as I then don't have the foresight (brain overload) to integrate new parts such as a new route without going through all my trains. I think you are on to something in not hauling every last load of express. That way your long-haul routes survive economic cycles better without "stacking". Once you get a bunch of trains stacking it is very annoying to me also that the first one to station doesn't always load first. This can quickly ruin any attempt at spacing them when you buy them. (Do you space trains when you buy, or do you think that isn't a worthwhile enterprise in a large game?) I have tried to buy fewer trains, but in boom times there are so many passengers that gather just taunting to me buy some more trains. :( You can see where that ends up in my stalled play of your map. Features from the wishlist such as reorganizing trains in the list such as by route and forcing the first train into a station to load first while waiting for loads would help. As you said fixing those bugs with replacement would be great too.

[MaglevForever]

I remember there are a few cases where I'd let my trains run for more than 25 years simply because no better model becomes available in that period. One example is running non-electric tracks with the GP18 which outperforms most newer trains for the next 20 years in terms of reliability, speed (especially at grades) and yet reasonable running costs. Somehow those 25 year old engines still run with a 1-2% breakdown chance...

You know grade-climbing ability and to some extent acceleration level must help this engine as well as the GG1 in terms of reliability (I have often forgot to replace it before it becomes unavailable as I don't see a need and then it's gone). Because corners seem to be handled like slight grades these engines will "work" less turning all those corners during their lifetime. Lower acceleration also makes them "work" less hard both accelerating normally (leaving stations, etc) and accelerating out of corners, which I believe is equivalent to accelerating up a slight grade.

[noctilucus]

You know grade-climbing ability and to some extent acceleration level must help this engine as well as the GG1 in terms of reliability (I have often forgot to replace it before it becomes unavailable as I don't see a need and then it's gone). Because corners seem to be handled like slight grades these engines will "work" less turning all those corners during their lifetime. Lower acceleration also makes them "work" less hard both accelerating normally (leaving stations, etc) and accelerating out of corners, which I believe is equivalent to accelerating up a slight grade.

 

Actually I never knew that so much detail and variables went into reliability before coming across this forum :)

 

[jeffryfisher]

I think it is a good challenge with an advanced military supply chain using the ports, management of fuel costs, track laying in the mountainous terrain, and nice events (including engine introduction events, supplying/building a dam that really fills with water and haulage based track unit allocation). This map really favors an efficient style and that's my thing I guess. Anyway, I asked because you said you were planning to try it.

I've never played with limited track-building. I like being to build whatever I can afford, even tearing down lumpy track to replace it with more level track. I'm still on the fence about trying it, but it's on my list to investigate.

 

(Do you space trains when you buy

I don't know what you mean, so I probably don't. I buy most of my engines on January 1. Sometimes I am so eager to use a new engine type that I will replace a bunch on new-engine introduction day (June 1).

[MaglevForever]

Haulage-based track allocation is the best type in my opinion. This rewards you for your efforts. Cascadia has plenty of products that you can haul to get more track units, I never found myself waiting very long to make a connection which I could afford. I think that it is an excellent idea from the creator to get the player to haul a wider range of the freight available, instead of going for only the needed and really lucrative stuff.

 

Spacing was what I was trying to do with transcontinental runs on US History. Depending on the length of the run I was buying a set of new trains for either end of a passenger run every few months to try to get a constant stream of traffic on that route. The intention was to try to get regular service from a city and hopefully avoid the problems with loading priority in the stations. The fluctuations in passenger supply with the economy can mess this up. Sometimes I was scared it would happen, other times it really did. Stopping some of the trains is probably the best option, but I also found that with such long routes trains that filled in boom times often arrived in recession and I lost a lot of revenue. Did that make it clearer?

 

@noctilucus there are many tips here. Some of the stuff is proven, other stuff like this is based off some facts + some observation/deduction. A good tip is the crossing-at-the corners "overpass" trick. This helps to relieve traffic jams. Basically, turn the grid on with the G key then lay a length of diagonal track. Where the diagonal track passes through a corner between the cells on the grid it is possible to lay another length of diagonal track in the other direction that also passes through that same corner. This will look like an X crossing on the map. This trick enables trains on one track to pass through those on the other track without stopping.

[jeffryfisher]

Track allocation:

Ah, what Cascadia does with track allocation, US History was doing by offering construction discounts (cheaper track building after you create a surplus in lumber or steel). The idea was to get players to care about some of the slow but economically essential freights (and layer an entertaining puzzle on top of the game to keep experienced players from becoming bored).

Now that I think about it, a small annual track allocation supplemented by bonuses for hauling track ingredients would create an excellent driver for slow heavy industry. However, I would want to find a way to refund track credits whenever a company tears up track (for regrading or route revision). Tracks not destroyed by disasters or war are generally reusable.

BTW, How do AI companies fare with limited track allocation? Does their construction become even worse?

Spacing:

I don't bother with it since traffic will mess it up anyway. I count on wait-to-fill orders to spread out trains that have become bunched.
 

Diagonal Overpasses:

 

The trick with diagonals is absolutely essential to the game, at least on a large map where a three car train stretches maybe eighty miles in length and can take a week or more to pass through an intersection. It may have been a programming oversight, but it compensates slightly for the game's scaling problems on continental maps.

[MaglevForever]

BTW, How do AI companies fare with limited track allocation? Does their construction become even worse?

The AI on this map are pre-built companies that own industries. These industries are involved in the goals. They never build any track, so no worries there.

 

By the way, this map has a larger scale than most. It represents the state of Oregon on a medium sized map. The engines have even been given performance boosts to compensate for the scale.

[letsdance]

in reply to the OT: i look at the numbers, but i don't calculate it exactly. it depends alot on reliability: the reliable engines can work alot longer, up to 30 years. the unreliable ones need earlier replacements, 15 years sounds good for those. i am lazy. therefore i like reliable engines alot =)

[immpy]

I am the micro-manager type, and I do stop the game play far too often for other's tastes, but that is me. I prefer a more reliable engine over one that may be a little faster, but replacement is more of a "feel" thing, and I amaze myself sometimes with how long I hang onto "old" engines. Depending on the consists, sometimes a slower engine just makes all of the runs easier to execute with full (5 or 6 cars) consists. Speedier trains would necessitate less cars per train and more runs, with less demand value at destinations. I like everything high, Greedy aren't I? It usually takes a rash of breakdowns or the change in car tonnages to make me go "hey, I need to replace some engines....and then I do. Highest age I can recal seeing for one of my engines was 32 yrs....prolly too long.

[Batchman]

I load RRT2 up every couple of years, usually, and start playing through it again. In the past, I usually played through about scenario 10 in the campaign, then felt I had had enough for another couple of years.

 

This most recent time through is the first time I have played through the entire campaign, and now have gone through scenarios 1-10 a second time at the intermediate level.

 

The odd thing is, this time around I have been in massive speed mode ... the question of replacing trains has not come up this time around, because I keep finishing every scenario in 5-10 years. (I now complete the first scenario in 2-3 years.) I think the longest scenario for me (of the ones I have done many times) is Knitting with Iron. It may take me an entrie 10 years to manage to earn enough money to buy the required access to Bavaria. But the last time through, even it may have only taken me about 7 years.

 

But I used to consider replacing some of the weaker, earlier trains about every 5 years if breakdown chances were getting too high, or a much better engine came along.

[jeffryfisher]

It's time to elevate your difficulty level ;)

[Batchman]

I'm working my way through the entire campaign on the medium level, then I'll turn around and do it on hard ... while checking to see what badly coded events I can fix in the process. ;-)

[MaglevForever]

You are right, the original campaign is really easy. I believe the steam version has the second century campaign included? The last scenarios in that, while a bit of a suspension of belief, should give some challenge to your management skills. There are no express gold mines to be had on those. You could also try some custom maps such as Cascadia mentioned above. There are lots of custom maps available from this site.

Link to "The Terminal": http://theterminal.dune2k.com/.

[RobS]

I happened to come across the issue of train replacement in the manual yesterday and as I remember it said 15-20 years would be a good lifespan for a loc.

I run mine longer, typically 20-25 for steams, 25-30 for diesels and 30-35 for electrics. In reality though, both the steams and diesels typically get upgraded to better models earlier. Once I get to the E111 though, it's every 30-35 years, depending on when I happen to remember to scroll through the list of trains. By that time, money usually is no object. I figure that's about how long they run in real life.

[jeffryfisher]

The reason I replace locos more often than dictated by reliability is because of maintenance cost. Take a look at train details to see two full years side by side. See how much more you pay for maint as an engine gets older. At 35, an engine could be chewing a quarter of its list price every year. Compare that to the annual maint of a new one.

[RobS]

9 hours ago, jeffryfisher said:

The reason I replace locos more often than dictated by reliability is because of maintenance cost. Take a look at train details to see two full years side by side. See how much more you pay for maint as an engine gets older. At 35, an engine could be chewing a quarter of its list price every year. Compare that to the annual maint of a new one.

Of course, maintenance costs is really the only reason to replace a loc (assuming a player does not want to change to a different loc for some reason). If maintenance costs never increased, there would be no incentive to ever replace a loc.

However, comparing the maintenance costs of an older loc to that of a new one is not the only consideration. Another important consideration, and many times an even more important one than comparing old versus new maintenance costs, is the price of a new loc.

Granted, the game usually makes the decision for me. That is, if I have the money to replace my older locs, I will replace them. If I don't have the money, I'm usually looking at going bankrupt. So the decision often isn't mine to make, but there are fairly rare occasions where I'm in a position to have to chose between spending money buying new locs, or spending money doing something else  (usually expanding my network to prevent the AI opponents from grabbing nearby cities).

My point is, are you really saving money in the long run by replacing locs that are only 10-15 years old, considering the price of a new loc? Again, the correct answer probably lies in what type of scenario you're playing, although frankly, looking at the victory conditions of most scenarios and their relatively short duration, I'm doubtful whether loc maintenance costs plays a large role in winning or losing. I could very well be wrong. What I do know is that once I reach a certain point in sandbox mode, where that cashier's bells rings almost continuously, maintenance costs seem to mean very little. Of course, if I'm running 200 trains all over North America and were to let them all age to 50-75 years, I would probably start to feel the pinch, but I'm guessing the constant breakdowns that would happen would pretty much preclude such a scenario in any case.

I should mention here that I never play scenarios. I only play sandbox mode, usually until I've bought out all other companies and own the entire map. Even if it started out as a scenario, I never pay attention to the victory conditions and simply "Continue Playing". I do play the advanced settings, at maximum difficulty, and usually end up around 203% difficulty. That's my "sweet spot" which I enjoy playing.

Edited April 22, 2016 by RobS

[jeffryfisher]

You're right that the purchase price of a replacement engine figures into the replacement decision. There's probably an ideal ratio between price and maintenance savings that somebody could calculate independent of reliability. I estimate it as "payback period". If replacing the engine will save me its purchase price in maintenance before it reaches the same age as the engine I'm replacing, then it's economical. By that calculus, engines should usually be replaced about every 10 years. I hate the tedium of replacing hundreds of engines, so I tend toward 15 years. I also look ahead at when up-coming new engines will appear so that I'm always due to replace engines at about the same time as a major improvement appears.

Saying that maint is the only consideration is incorrect. Reliability is a valid reason to replace engines even if no better engine has appeared and even if maintenance remained constant. I think that reliability is an under-appreciated aspect of the game since its effects are so hard to quantify (how much money did I lose because a train was delayed, and all trains behind it were blocked, and a traffic jam ensued...?). You can only ignore reliability if you play with accidents turned off.