Balance and realism issues with RT3 v1.06 locomotives

[low_grade]

um, got the beta Lirio zip... giving me difficulties extracting it...

[Gumboots]

Use 7zip for that. WinRAR wont handle it. http://www.7-zip.org/

[low_grade]

Thanks! This reminds me of some work I did back in 2009...

Campbell 4-4-0 for NA in 1944, with tweaks to American and Duke and Ten Wheeler also included in this thread's downloads. (I'll keep Lirio's updates of those engines for now...)
viewtopic.php?f=67&t=2002

Baldwin 2-6-2 for NA, W in 1901 (great reskin project for somebody!) Also modified Atlantic, Pacific, and Shinkansen if you want to try those out with some tweaks...
viewtopic.php?f=67&t=2005

GWR Hercules 0-6-0 for E in 1842 (a Baldwin 0-6-0 for Europe...)
viewtopic.php?f=67&t=2001

[Gumboots]

Missed those. I'll check them out. Also, if you feel like tackling some reskinning I'm getting quite good at that, so am happy to help with tips or with tracking down which parts do what. I probably wouldn't want to actually work on the thing (have my own projects going at the moment) but am fine with making things a bit easier for others if I can.

I've found this loco balancing is tricky. Take the Berkshire vs H10. Changing each engine's stats is easy, but IMHO the real question is what effect this will have on your company's bottom line. To get a genuine increase in profit over a period of time seems to require more radical differences in stats than I initially thought. Once the locos are running in the game there are things like random availability of cargo, random stopping in random locations depending on cargo value when trains meet, random breakdowns, etc that tend to mess with the basic differences between two locos on the same run.

This may mean that modest differences in stats are effectively illusory when it comes to benefits. You may have a nice "progression curve" of stats but the results the locos deliver, in terms of profit, may not be noticeably different. Reliability is a factor here too, since if a newer engine has better hauling ability but a higher chance of breakdowns (mumble grumble Mallard mumble grumble Atlantic mumble bleh) the time lost to breakdowns/line congestion could easily outweigh any perceived benefit from the better stats.

[AdmiralHalsey]

since if a newer engine has better hauling ability but a higher chance of breakdowns (mumble grumble Mallard mumble grumble Atlantic mumble bleh) the time lost to breakdowns/line congestion could easily outweigh any perceived benefit from the better stats.

God I hate using those loco's. I don't care if the Atlantic has better acceleration the amount of times they breakdown is insane. I normally wait for the Pacific if I need to upgrade my long haul expresses.

[Gumboots]

Tell me about it. Atlantics aren't too bad if you service the crap out of them (ie: keep oil always at 60% or higher) and replace them fairly young (around 8 years). The Mallard is just plain annoying. Can't stand the silly thing. Breaks down anytime just for fun, and totally wilts as soon as it sees a 1% grade ten yards long. Also looks stupid to me.

[RulerofRails]

I agree about the Atlantic and Mallard. Very rarely have I used them. Did use a couple of Mallards on my play through the PopTop Italy scenario when going for my no-industry all-steam personal goals. On very flat routes only, maybe I am wrong, but they seemed a better option than the Northern in that case.

Reliability is a major concern. It can outweigh a lot of other concerns. But I would ask, when is a new engine enough of an improvement in performance to merit replacing the old one(assuming it is not ready for the scrap yard)? And when this is so doesn't it skew the engines toward no realism?

Fast acceleration is good for light passenger trains, but often for me to get a network running at high speed I have to contend with fewer cars per train, higher costs and adding double track.

I wish that scenarios had more engine events in them, like randomly the game would give you an option to hire an engineer who would work on a new engine maybe every ten years or so. From a financial standpoint there are plenty of engines to choose from especially in the 1920s and 1930s with all the ones that have been made. Making favorite engines better than the others is great, it will just make the "most cost-effective" engine change. I don't have anything against any of the engines, just wish there was a way even the poorer engines would be more necessary. There are many engines in the 20th century that I seldom use when playing.

[Gumboots]

That engine event idea isn't a bad one, and would be quite easy to code.

[AdmiralHalsey]

That engine event idea isn't a bad one, and would be quite easy to code.

You could also just add some by hand using the scenario editor.

[low_grade]

I know back when I was hex-editing, my goal was this: to give me a reason to use more kinds of engines for different purposes. I was tired of games where 90% of my engines were one model, and then the A2 or something becomes available and I replace ALL my engines immediately. To me, the problem was the superstars. The 2-D-2, for example.

So I worked up a spreadsheet that took all variables into account, and rated each engine in like 18 different circumstances (flat, medium, or steep grade vs long or short runs vs passenger, freight, or mixed.) I tweaked engines to make them the best at something for a period of time, to give me a reason to use them. No engine should be the best at everything at any time. My ideal was a "smooth curve" of progression in the various areas of expertise, with better engines coming along every 8-10 years. But not so much better that you replace what you have, keep those old locos running til they're ragged!

[Gumboots]

That's what Lirio was aiming for too. But, IMO the problems are:

1/You also have to take reliability into account. Worse reliability will counteract better other stats, because it wont haul more if it breaks down more often.

2/ As I mentioned before, changing the stats to get a "smooth progression" is easy, but getting an actual advantage in profit seems to be trickier and to demand more radical changes in stats. The only slightly distinct engines you get from a "smooth progression" are probably going to be indistinguishable in terms of what they do for your bottom line.

Only way I can see of doing it is running multiple engine vs engine comparisons for long periods (10 years minimum) and checking for any change in net profit. It gets pretty tedious.

And IMO running old locos until they're ragged often doesn't make sense. Maintenance costs go up around 10% for every year of life. That means after 10 years your maintenance costs are roughly double that of a new loco of the same class, and after 20 years they are 4 times as much. Old locos cost heaps to run, and also have a much higher frequency of breakdown, which means they haul less and they clog up your line for other locos when they break down.

[Lirio]

Thanks for the feedback guys! I'm glad the my modifications seem to generally be a positive experience. I agree that the default Atlantic and Mallard are lacking, that's why I raised their reliabilities to above average. I also made the Atlantic faster and the Mallard a bit slower but harder pulling. I think this makes the both of them a lot more worthwhile. I haven't been playing around with RT3 much lately, but I have cleaned up the prior beta of a few bugs and issues it had. I'll improve the Shinkansen's reliability as suggested by Sabratha, and possibly also the end date as well. Should have it out within the next couple of days.

My initial feeling is that the changed version of the Mikado is possibly a bit too good. In the OP of this thread you (Lirio) said "On that note, both the H10 and Northern have pulling capabilities that I find somewhat out of line with the other engines, and indeed their actual capabilities in reality. They were impressive, sure, but not quite that impressive".

However, in this update pack the pulling power of the Mikado has been substantially increased. I realise that this is down to trying to get it to make sense in the context of the other existing engines, but my feeling is that perhaps it would benefit from a bit more adjustment. Then again, the Berkshire probably could too.

It had originally been my intention to decrease the Mikado's pulling power and increase its free weight, but I wound up leaving the free weight alone and increasing the pulling power. I tried to keep notes of why I've made the various changes but very quickly proved far too tedious, so I gave up and figured any change worth making would have an obvious rationale after a bit of thought. I guess I was wrong, because I actually don't have much of an idea as to why I did it that way. The balance must have worked out better that way.

And IMO running old locos until they're ragged often doesn't make sense. Maintenance costs go up around 10% for every year of life. That means after 10 years your maintenance costs are roughly double that of a new loco of the same class, and after 20 years they are 4 times as much. Old locos cost heaps to run, and also have a much higher frequency of breakdown, which means they haul less and they clog up your line for other locos when they break down.

Yeah the 10% increase per year to maintenance costs is pretty brutal. I'd reduce it to 5%, or even 2%, but unfortunately I think it's hard-coded into the game. The only other way for old locomotives to make financial sense is if the rise in maintenance costs for newer locomotives is made a lot sharper, but I'm not sure it's possible to pull that off without screwing up the game economy. I'm rather not inclined to try.

[Gumboots]

Yeah it's hard coded into the game engine so there's nothing we can do about it. I've given it some thought, and it seems to me the only real way of making reliable engines worth using long term is to give them exceptionally low initial maintenance costs. This makes a certain amount of sense anyway, because usually a more reliable engine would cost less to maintain (well, you'd think so anyway).

[Gumboots]

Hey I was just looking at some stats. This reminds me of Mikados and Berkshires. You might think that the A1 Berkshire I made should be nobbled some so that it gives a "smooth progression" from the game's Mikado and so "people wont be tempted to replace them all" and all that. However, in reality, Berkshires were not a "smooth progression" from the average Mike. They ate Mikes for breakfast. That was the whole point of them. If they'd only been an infintesimal improvement nobody would have taken any notice. The reason they were successful is because they cost hardly any more to build than a Mike (just a couple more wheels and a bigger firebox) and they provided masses more usable grunt combined with lower fuel costs per ton/mile of stuff hauled. So, not only did they haul faster, they did it cheaper too.

When I was modelling the Berkshire I tested it against the Mike in an attempt to model (as far as is possible within RT3, without going mental) the actual differences between the two and to check that the Berkshire actually did deliver somewhat better (not hugely better) bottom line returns over a period of years. This required stats that may look quite radical compared to the Mike, but it seems to give results that accord with roughly how the engines were matched in reality.

Despite the rather pronounced difference in stats between your re-balanced Mikado and my version of the Berkshire, what I found when testing over a decade or two (basically, until one locomotive crashed) was that on a water level route the Berkshire didn't provided any noticeably improved returns, while on a graded route that averaged about 2% (mostly against the train one way, with the other) the returns from the Berkshire ended up averaging about 10% more than the Mike. So, if you had a network full of Mikes and you replaced all your water level routes with Berks, you wouldn't end up any better off. If all your routes were graded you might be 10% better off, purely in terms of profit per locomotive (ie: not considering track maintenance, company overhead or whatever else).

[Gumboots]

Hey Lirio, you might want to check out the "1.06.1 patch" thread. I've just been doing some number crunching with regard to possible changes to loco stats. After doing that, I'm pretty convinced that what you and others seem to want is not what the original devs intended. That doesn't mean you can't get what you want, but may explain why the game as made doesn't deliver it.

People seem to want a "smooth progression" of loco stats and the ability to keep reliable engines running for long periods of time. As far as I can tell, the original game coding is more or less aiming at the opposite. It seems designed to encourage fairly rapid turnover of locomotives, with random jumps in their cost and abilities, and with some included specifically as duds, or as traps for the unwary. The stats for running costs are all over the place until you get to the fictional super trains, at which point you have presumably arrived at a plateau of awesome awesomeness where your trains all look like spaceships (which must mean they are awesome) and nothing matters any more. IOW, it appears to be heavily targeted towards the under 12 age group. Not sure what that says about us diehards.

Anyway, given that, it's not surprising that some people find the original coding unsatisfactory. Substantial changes may be in order.

[Hawk]

I agree to drifting towards a little more realism.

[bombardiere]

I am not able to open that Lirio's zip pack. I have tried to that in two computers and with default Win7/10 zip opener, WinRAR and openRAR. No joy. Could it be repack so that could be opened without difficulties? I do not want to install a new program for just one unpack.

[Hawk]

Try 7-Zip. It's free and opens zip and rar files, as well as 7z files, plus a bunch of other stuff.
Once you use it you'll probably never look back or need WinRAR or WinZip again. I only have 7-Zip installed and no longer need either of the other two.

What's the name of Lirio's file?

[low_grade]

Yes, 7-zip.

[sbaros]

Alpha mask support isn't available in most image editing software. Far as i can tell it's pretty much only GIMP and Photoshop

Well, I am not familiar with very advanced graphics concepts, but as far as I can read in my CorelDRAW Graphics Suite Help, Corel seems to support "Alpha mask":

To add the current mask to an alpha channel Back to Top
1. In the Channels docker, click an alpha channel.
If the Channels docker is not open, click Window- >Dockers ->Channels.
2. Click the Save to current channel button

However, I haven't found yet any utiity to convert the distributed DDS skins to TGA, in order to modify them with Corel. Apart from that, is it worth to seek some utility to convert the TGA back to DDS, or should I be content with TGA ?