NUTS: Revolution among train sets?

After a while I am writing an article again. But unlike before, I will not mention building styles, techniques or anything similar. As some of you might know already, I started drawing my own trains some months ago. The set is called NUTS, (Nameless/NUTS) Unrealistic Train Set and it’s home is at the devzone. In this article I would like to make an official introduction to the set and describe why and how is it expected to work. Why do I write this? Basically because I am not going to just describe how is my set going to look as that info can be found on the devzone already, but I would also want to talk about the aspects and behaviours of other sets and what do I see good or bad, or what to consider when making a plan in our games, or even what to think about when choosing newGRFs for your own single player game.


Note: I am going to add images to the article later, but they are not too important so I publish the article now :).


There definitely are many ways on how to view train sets, how to classify them and how to rate them. You could focus on the graphical side, which I will not consider for now in this article. You could focus on the realism part, which I consider absolutely stupid and this article should show why. Or you could focus on the way how the train set actually works and what gameplay aspects or even effects it has and ultimately what does it bring to the table when playing your game. This is what this article is going to be all about: I am going to compare a few train sets (original engines, US Train set, 2cc set, UKRS) and then I am going to show how NUTS is supposed to work in these areas.


Choices and frequency of new engines

If there are any key elements which make a newGRF fun to play, I think giving the player some choice what does he want to actually do is one of the most important ones. Choosing whether to use rather fast engine, or a slower one, gaining something, sacrificing something else, adapting your train to your network. There are two kinds of a choice that I see. First one is the spectrum of trains available you get when you play the game normally (like when you start in 1950 or earlier and upgrade engines on the fly). The other one is much more used in our conditions and that is the end-game choice. After you get all engines, it is nice when you can still decide which train to use instead of having one ultimate engine which is ultimately strong and ultimately fast, fitting for every purpose and being used always again and again. In general I just think that having the choice to do something makes a person think, thus allows for a higher skill-level development as we could call it in general in computer gaming ;).

Another important element is the number of how many engines do you actually get to play with and how often do you get a new engine which is better than the previous engines.



Now the list of a few sets is going to follow, examining those aspects:



SRNW Stations

early game: EE1 – slower, stronger, shorter vs. Pacific/Streamliner/Merchant navy – faster, weaker, longer

mid game: Metrovick/Deltic – slower, stronger, shorter vs. Streamliner – faster, weaker, longer

late game: AL10 – strong, slower vs. GEC91 – overall balanced vs. Wardale604 – super fast, super weak

This is basically how you play UKRS, being able to pick any of those routes while all of them can be reasonable under some circumstances gives you unmatched amount of choice.

In the end of the game, any of these mentioned engines is usable, and there even comes the combination of passenger trains. You can have slower (160/225kmh) network with pax train in it, or you can go for faster pax trains and use them with Eurostars at 289kmh or even make a separate maglev network.


US train set:

SRNW Stations

early game: GG1 vs. nothing – GG1 is fastest, strongest, and short

mid game: GG1 vs. 8 – a pathetic challenge since 8 is only 17kmh faster and has half power

late game: SD110FC – strong, very slow vs. EL13k – strong, fast vs. Acela HHP-8 – overall balanced

As you can see, there is not much decision going on in the early ages. Luckily, it makes up for it in the end of the game as both of the faster engines are well usable. The SD110FC is awesome but just way too slow. Unfortunately there are no real passenger trains which would fit with the cargo trains on one network so you would have to use cargo trains with pax wagons. (Which is possible but imo ugly) So you automatically have to make separate networks in case you want to make both mines and cities happy.


2cc Set:

early game: Royal Hudson vs. nothing

mid game: Metrovick vs. nothing

late game: Lok2000 vs. nothing

Unfortunately this is basically everything you use while playing 2cc set which has like hundreds of engines. No choices are here and we get a very limited number of usable engines.


Original Engines:

Even more different are original engines. There we can’t really set too much of choices in the whole game. However we can set a very nice chain:

Kirby -> Chaney Jubilee -> Ginzu A4 -> SH8P -> UU37 -> Floss47 -> SH30 -> SH40 -> TIM -> AsiaStar

While we can decide between using SH40 – slower, better acceleration vs. AsiaStar – faster, worse acceleration, those choies are not really the main good thing about this set. The amazing thing is that you use so many engines throughout the game and change them often. If we started in 1926 when kirby comes out, we get the last engine (AsiaStar) in 1993. That is roughly 70 years during which we used 10 engines! This is what keeps the game very active and fun to play. Unfortunately when you look at the newGRF sets, none offers a variety of this many engines. (And I do not even mention modern rail types as monorail/maglev coming afterwards)


NUTS is hoping to use all of those points in the best way I see fit. I hope to provide variety during the game by making three cargo train classes based on speed/power ratio – you get speed or power or a bit of each, but never both. And two classes for passenger trains – local and ICE – speed or capacity.

As wasn’t mentioned above, another thing to consider in choices are also rail types. Most newGRF train sets just do not use additional rail types or make them limited like passenger only. The only set which uses monorail/maglev completely are original engines. As we all know, maglev has the highest throughput capability of all original engines. Therefore if you want throughput, you automatically choose it. There of course are many reasons why not to use maglev such as acceleration or slow loading times but throughput-wise, it is unmatched.ย What i want to do with NUTS is making faster trains not have higher capacities. To be more precise, rail wagons should have 30t capacity, monorail gets 25 and maglev ends up with only 20t per half-tile wagon while the speeds are kept in about the same place as the original train set has. This means all of the three rail types are considerable throughput-wise and you can focus on choosing based on systematic network requirements instead of raw power so to say.

Not just choices are the important things but as shown with original engines, it is really enjoyable if the game is active and keeps you up with new models of engines. With the way how I classed five types of trains, all of them get a newer version in a decade. Therefore you get a new engine every two years, and every ten years an engine which is surely better than the one you are currently using. So basically you could say NUTS can be expected to be five original engines aside of each other, thus providing a choice and at the same time all of the engines should be usable with some network.


If you wanted to get the root of “why” do some sets provide so limited choices and usable engine variety, just think about the word realism. By using real-life figures of trains and their number-information such as introduction dates, power, maximum speed, etc., you are signing your agreement with being limited in the bounds of realism. Therefore by using fictional engines, we have an absolute freedom over how the train set actually works in all those aspects mentioned above.



Other aspects of a train set

Wagon length

Things get quite interesting when you decide what wagon length do you actually want to use. Shorter wagons are definitely cute and they also have a great usage with early game stamers. You have a long steamer so you need smaller wagons in order to have as many wagons as you are going to need with a short engine later on. Later you just replace engine to a short one and wagons to longer version afterwards without having any trouble. However at the same time they bring a few disadvantages:

  • it is basically impossible to know how many wagons do you need for your train so you basically just buy wagons until the train is long enough instead of being able to tell how many wagons there actually are
  • with certain train lengths and short wagon combinations you can get various CL or length issues

There are basically two simple ways how to solve this, you use short wagons or not. If you do, then you can – but don’t have to – use longer steamers which just does look good. Although there are sets like US Train Set or Norwegian set which do use long steamers and at the same time use only long wagons, so you always get into problems if you use the steamers.

NUTS solves this by unifying all wagons into 0.5 tile length. However there are a few things to note:

  • There are visually also short wagons, but all of those are articulated wagons, so you just buy 0.5 tile wagons which consist of two 0.25tile segments hauled together. Simplifies vehicle purchase and looks still nice by having short wagons.
  • Because this would ultimately eliminate all long steamers from this set, all passenger trains (ICE or local) are double headed. Therefore 1 tile long. And steamers are going to be that long as well, so during all of the game you have an engine and a constant number of wagons. Unless you would want to change your 1 tile pax steamer to a 0.5 tile cargo locomotive – which is quite unlikely – and currently even unsure whether cargo locomotives are actually going to be able to transport passengers.
Train capacity
When a train is able to carry some cargo, it must not be able to carry more than a wagon. Therefore makes you choose either more power from the engines, or more capacity from the wagons. One could possibly say that two engines are more expensive so they should be automatically better, but firstly this just shows how stupid money in OpenTTD is. However even the starting point of such an argument might not be true as NUTS might use a bit weird costs as in cheap engines and costly wagons. After all, you transport cargo in wagons, and the wagons make you money. Engines only eat the money back on (significant) running costs so it would not mean that you could automatically buy ten engines and three wagons and be happy.
Wagon capacities
When you look at newGRF sets, you can see that basically every single of them has different cargo capacities. ย From low capacities like UKRS2 so far has to mindlessly high capacities such as 2cc or polish PKP set have. Let’s collect a few points on what do wagon capacities actually change:
  • Low capacity = train flood. It might not be good if we have too low capacity either. A 2295 primary industry could become such a force to be reckoned with that it might force you to build a separate line for it. This could easily lead to solutions like point to point traffic. Why would you make some junctions when you know that the lines are going to get full anyway? This is exactly what happens with ECS but in inverted way where it is the industries which have higher productions.
  • High capacity = more difference between full and empty trains. Hard to say whether this is good or bad because that greatly depends on the structure of the network. On one hand it would be just dumb to have low capacity trains and therefore not being able to distinguish acceleration of full and empty trains. On the other hand having too high capacity trains means that your empty trains are likely to accelerate instantly while full trains struggle to get to top speed. This could even mean that you need for example four lines for drop station entrance but the exit can live with just two lines going out. Simply because the empty outgoing trains get much more packed easily. Both of the extreme ends are definitely odd to me, so it is probably all about finding some golden ideal spot.
In the end I think that the 30 units of cargo per 0.5 tile wagon (be it tonnes, thousands of liters or crates/pieces of -things-) is a good standard which is used by the original engines and UKRS for the most part.
NARS and Tropic Refurbishment set with 33 and 34 units of cargo per 0.5 wagon are somewhere in the middle, which is quite nice point as well.
With 40 units of cargo per 0.5 tile wagon we have for example US train set or DB set. There the differences between empty and full trains get really significant, I think that even slightly too much.
And then there are sets like 2cc Train set or PKP polish set which use 60 units of cargo per wagon. When you try to have a reasonably accelerating full cargo train, you just end up having instantly accelerating empty train which looks really weird, not to talk about the eternity it takes to get an actually high train count which gets fun to play with.
NUTS is going to use 30-35 (not decided yet) units of cargo per 0.5 tile wagon for rail trains. As I already mentioned, monorail is going to have 25-30 and maglev 20-25. This is not only because of the throughput balance between rail types but also because faster trains tend to accelerate slower (unless they have ridiculous power). Therefore helping them with lighter wagons is also a nice option, but of course this reason rather is just a side-effect.
An another interesting thing is that many sets make goods wagons carry more crates of goods. Since goods are lighter than most of the other cargo, it just equalizes the weights of raw cargo:goods trains. I think it is a bit better to keep the capacities for goods trains with the same numbers as capacities of raw trains. Firstly because all industries already produce less than 1:1 ratio of material:product and secondly because I think goods trains should be a bit faster accelerating than raw cargo trains.
Extra utilities

Special engine

There is going to be an extra engine which is not going to influence how the train actually behaves but is going to have the ability to carry cargo and look like any wagon. This would make you able to make for example a train consisting of normal engine, four wagons and this special engine. Therefore you visually get a normal 3 tile long train with one engine and five wagons. The point is that later on in the game you can turn this wagon into a normal engine and suddenly have a single engine TL3 autoreplaced to double engine TL3.


Wagon refitability

Another thing to consider is the choice wagons provide for refit networks. This can’t really be considered a native feature of a newGRF – rather a feature I started abusing some time ago and developed into the refit networks we use nowadays. Still, it might be nice to offer the option to have for example a wagon refittable to any cargo. In the end, why not? Providing more possibilities and letting us consider even something we couldn’t have thought about until now. (unless the wagon we needed already existed in some other newGRF) At the same time it might (not sure) get easier to purchase vehicles – you buy a vehicle with some wagons, and refit to the appropriate cargo afterwards. On the other hand it might just get a bit more complicated for some people who like to purchase wagons by separate cargoes and the refit window tends to be rather small.

NUTS is definitely going to have a parameter setting (if not default [and only] mode that way) where we have one wagon for everything. You would just purchase an “universal wagon” and refit it to whatever you want. And it would change the sprites accordingly.



Whether is NUTS going to be better or worse than the other train sets, depends not only the way you look at it, the way you use it, or the things you expect from it or even the way how your subjective opinion commands. It is also going to have to look somehow so we don’t play with colourful bricks on rails. Now I am getting to the topic of what the next article is going to be, but about that next time. ๐Ÿ™‚ Thank you for reading and if you would like to join me, just poke me on the IRC or join our devzone.

3 comments so far

  1. LoPo October 18, 2011 20:24

    nice article V ๐Ÿ™‚ but i’m not sure if i’m ready for the Nyan train’s ๐Ÿ˜‰

  2. V453000 October 20, 2011 09:46

    That is what this article is not about. Not only because the graphical look is going to be just the “shell” in which the “nut” [the way how it works] is going to be covered ๐Ÿ˜› At the same time, the super-weirdo trains are not going to be in the main train table, therefore it does not belong here really. It is just something extra which does not force anyone to use it ๐Ÿ™‚

    This article defines the basic systematic way how the main engine pool is going to be made. Those engines should be at least “semi-sane” ๐Ÿ˜‰ Nyan cat and slug (currently only 2 extra engines) are going to be completely away from the main system. To make them “usable” they have the attractive/weird look but they will be slower/weaker than the main scheme trains,so you could classify them as useless something extra being available if you want to be really NUTS ๐Ÿ˜‰

  3. K?a??ep Mo??a??e#7051 January 8, 2020 19:54

    I’ve just have an issue, I am not able to combine trains, for example I want 2 wagons of mail and 5 of passengers and I cannot do it with NUTS (yet).

Leave a comment

Please be polite and on topic. Your email-address will never be published.