The first time my local group sat down to play 1828 (in 2019, before 18xx.games existed), we hit a snag during the rules explanation: somebody wanted to know if there was a limit on the number of shares in the bank pool. This led to the rules re-reading, a fair bit of debate, and a tentative answer: the rules do not say either way, so, no? Wanting to get started, we posted the question online in numerous places and eventually received confirmation that our interpretation was correct. The episode added about twenty minutes of stall before we could get started playing, and I have since heard similar stories about this game from other groups.
So what happened here? Why would players new to 1828 assume there could be a rule governing shares in the bank pool? Why was the lack of a rule about this either way a problem?
The answer lies in what happens when player contexts intersect with written rules. By “player contexts” I mean the knowledge, experience, and expectations that players inevitably bring with them to the game table. In this post I will articulate the importance of considering player contexts when writing rules, and I will provide some tips for doing so. Much of the discussion below is set in an 18xx context, but the points I make apply to any set of written rules.
Definitional versus Contextualized Approaches
In this post I will discuss two schools of thought on rules writing: definitional and contextualized.
The goal of the definitional approach is to provide a formal definition of the game: what players can do, how to do it, and how to determine the winner. It eschews examples, lengthy explanations, and clarifications. J.C. Lawrence is an advocate for the definitional approach, and I think you can see a similar impetus in John Bohrer’s rules, although I have never read anything about him espousing a particular philosophy here, and Winsome games do include examples from time to time. The underlying theory of the definitional approach is that players will read the rules, interpret them literally, and follow them exactly. (I suspect, with no data whatsoever, that this approach is favored by writers and readers who spend a lot of time with computers.)
J.C. has referred to this approach as focusing on “precision, correctness, and completeness” (personal communication), and he often discusses how his goal with rules writing is to create a definition of the game. There is a lot to admire in his and similar efforts: the resulting text is short and concise, and often makes me wonder if my own efforts are needlessly wordy.
The goal of the contextualized approach, on the other hand, is to maximize accessibility (within reason) and make the game as easy to understand as possible (for your audience). The rules should enable the largest possible number of people to understand and enjoy the game.
This approach emphasizes the same core things as the definitional: stating what players can do, how to do it, and how to determine the winner. But it also recognizes that readers come to rules not as blank slates, but with personal and gaming backgrounds that influences how they read rules, interpret them, and draw conclusions about them. A rules writer taking this approach tries to anticipate questions and points of confusion, and includes clarifications and examples accordingly.
By accessibility within reason, I mean that you can assume some basic knowledge on the part of your expected audience. 18xx rulebooks necessarily assume the reader can do arithmetic; if they didn’t, there would be quite a few pages explaining how to pay for things. This seems like an obvious point, but I have seen many writers–of everything from game rules to academic papers–struggle with the question of how much knowledge can be assumed on the part of their readers. The key is to identify your expected audience and go from there. If you are writing rules for an 18xx game and your expected audience is toddlers, well, best of luck.
In the rest of this post I will discuss four aspects of player contexts: naturalized conventions, weak conventions, errata, and inclusivity. I will also give some tips on how to think about using examples in these terms.
Examples of Player Contexts
This section describes four examples of player contexts, and shows how not addressing them can create problems or misunderstandings.
The contextualized approach is particularly important when a game violates a genre’s (e.g 18xx, war games) or form’s (e.g. card games) naturalized conventions. By “conventions” I mean things that players can and cannot do in a game; they may or may not be stated in the rules. These conventions become “naturalized” when they are so ubiquitous that they seem to be a natural outgrowth of the genre or form, and players assume them to be part of all games within that group. These are usually negatives: in financial games you cannot steal money from the bank, or assets from your opponents; in war games you cannot move your opponents’ units. But they can also be positives: when dealt a hand of cards you can look at the cards and sort them.
When a game violates a naturalized convention, it can require the rules to specify what the players cannot do. Normally rules focus only on what players can do, because the list of things they cannot do is potentially infinite, but in this case accounting for player context requires it. For example, the rules for the card game Bohnanza specify that:Bohnanza Rules wrote:“Players may never change the order of the cards in their hands. Thus, sorting cards by variety or any other means is not allowed.”
Taken in a vacuum, this text violates the principle of only telling players what they can do. But if you have played nearly any card game in your life, you have likely sorted the cards in your hand to make it easier to understand as a whole. This freedom is a naturalized convention: when I’m teaching bridge, I never think to tell players that they can sort their hands. The card game Hanabi does something similar in that the rules specifically tell players that they cannot look at the cards dealt to them. Being able to look at one’s cards is a naturalized convention, so the rules must call-out this violation.
A strict adherence to the definitional approach of rules writing might lead an author to omit these sorts of explanations, because the goal is to only include what players can do. Now, this statement seems like something of a strawman: I don’t think anyone writing the rules for Hanabi would ever say “I use the definitional approach so I’m not going to say player’s can’t look at their hands. They should know this because I don’t say they can look at their hands.” That attitude seems insane. But the only reason it seems like a strawman / hypothetical insanity is that we automatically expect game rules to explain things that need explaining, which is just another way of accounting for our player contexts.
A weak convention is what it sounds like: a convention weaker than a naturalized convention. It is not strong enough that players automatically expect it to be the case in a new game (within a given genre or form), but it appears frequently enough to be familiar, and can still be a source of confusion. The 1828 anecdote I opened this post with is a good example.
In nearly every 18xx game, players can sell shares they own in a corporation to the “bank pool,” also known as the “open market.” Some games use the first term, others the second, but they are functionally the same thing; effectively these shares would now be owned by the bank. In many–if not most–titles, there is a limit on the amount of per-corporation shares that can be in the bank pool; this limit is a weak convention.
As I noted above, the 1828 rules do not specify whether there is a bank pool limit or not. This is consistent with the definitional approach to rules writing. But amongst veteran 18xx players this can be a problem: not having a limit violates a weak convention, so not clarifying the rule can create confusion. It also leads to a situation where, in order to determine that there is no limit, the players must read the entirety of the relevant sections, or even the whole rulebook, to be sure that no limit is mentioned. This is a bit like proving a negative: the reader must prove that there is no such rule by pointing to its absence (this also relates to “Errata,” below), which creates work for them that could be avoided by a simple sentence; in my group’s case it would have saved us 20 minutes, a lot of reading, and spamming various online communities with our question.
But on the other hand, if there was such a rule in the rulebook, newcomers to the genre would find it awkward or confusing. They have no reason to expect that there could be a limit on shares in the bank pool, so why state that there is none? The limit on shares is a convention that has not been naturalized amongst the gaming public, and even if 18xx were to become as ubiquitous as card games, it never will be naturalized because some games break from it anyway. This seems like a strong argument in favor of the definitional approach.
But the way to avoid confusion in this instance is to account for multiple possible player contexts: players who know 18xx, and those who do not. One way to do this is to include a page or two at the end of the rulebook that summarizes the rules for 18xx veterans. Such a summary is a good place to include clarifications and shorthand that are extremely useful to veterans but risk confusing neophytes. For example, the term “semi-restrictive track” communicates a wealth of information to genre veterans but is meaningless to newcomers. Another viable approach is to include clarifying footnotes written for veterans (they would need to be identified as such), but this is less useful than clumping all of that information together in one place.
Of course, the problem with these methods is that even the idea of an “18xx veteran” is not well-defined, and so there is a limited amount of knowledge that can be assumed.
It is a reality in board game publishing that errata happens. Mistakes get made, and games go to press with errors. This is especially true with complex games: the more rules and components, the more places mistakes can slip through. Some games even ship with key rules missing from the rulebook (I can think of a few examples and you probably can too), or even rules only appearing in examples. Many of the game boxes on my shelves include errata lists and rules clarifications that did not come with the game. A result of this reality is a doubt lingering in the back of every seasoned gamer’s mind: if something seems weird or incorrect, then maybe it is incorrect. This experience with errata is part of their player context.
This is another reason to provide clarifications and examples: it creates redundancy. If something gets left-out or omitted on accident, having examples and clarifying text can still communicate the essential information. This is not an ideal scenario, but it is better than nothing.
This also applies to my 1828 example: this session took place soon after the game was made available, but the designer was still stating that it was a prototype (or something similar), and had stated that the game was not in its final form. In this context, it was reasonable to assume that a rule about bank pool limits was accidentally omitted.
As an aside, this problem is hardly unique to board games. For several years I made a living as an indexer of academic books published by major university presses, such as MIT, Cambridge, and many of the Ivy League presses. Despite being the fourth professional looking at the proofs–after the author, editor, and proofreader–I regularly found errors small ("I agree that the Prussian army was 'stronk' but I'm not sure that's what you meant") and large ("This section titled 'Birds and Reptiles' only discusses macaws"). Executing highly detailed work in any field is challenging, and doing it perfectly is next to impossible.
Part of a player’s context is their identity: who they are, what their background is, and so on. Inclusivity is about making all players feel welcome. In terms of writing rule books, this often means being sure that you do not exclude anyone. In the past, it was a common convention to use “he” and “him” as gender-neutral pronouns in rulebooks, but of course these terms are not gender neutral and they create the impression that the author is assuming the reader identifies as male, thereby reinforcing stereotypes about who does and does not play games. By not making assumptions about your readers’ gender, you show that you respect them and their identities, and you help to create a more welcoming community.
This problem most commonly comes up with pronouns, with rulebooks consistently referring to players as “he” or “him.” It is not sufficient to simply change pronouns to “he or she” and “him or her,” as these constructions can still inadvertently exclude. One way to avoid this issue is to simply write around pronouns, for example by always referring to “the player” or to “player A, player B,” and so on. Another method is to write in the imperative: give the reader instructions as if they are performing the actions described. Here are some examples:
To sell shares, place the certificates in the bank pool.
This is clear and succinct. However, the imperative can get a bit tricky in complex games where multiple entities are doing things in rapid succession, or simultaneously:
To sell shares, place the certificates in the bank pool. The bank pays you the current market value for each share sold. Then move the corporation’s share price token down one row per share sold.
The problem here is that it does not matter who moves the share price token, so it is more accurate to use the third-person:
The selling player places the certificates into the bank pool. The bank pays the selling player the current market value for each share. Then the share price token is moved down one row per share sold.
Although this is wordier than the second-person example, the advantage is that it is clear that it doesn't really matter who gives the selling player their money, or who moves the share price token. (Dividing-up these tasks can save a significant amount of time in the long run.) Also note that “the selling player” avoids using a gendered pronoun.
Writing around pronouns does lead to wordier sentences, which can be a problem too. If you prefer to use pronouns, I recommend using the singular “they.” This is not a perfect solution either, but unfortunately English lacks a third-person neutral singular pronoun that is also not dehumanizing; nobody wants to be called “it.” The difficulty with “they” is that it can be confusing to non-native speakers. If I was reading a French sentence (a language I pretend to understand), and I saw “ils” used in place of “il,” when the rest of the sentence was conjugated around “il,” I would be totally lost. But “they” is the best option we have for now. Rules of grammar inevitably follow patterns of use (in the long term), so if “they” becomes widely used in this fashion, eventually it will become an accepted use and more people will know to expect it. I think this is already happening, as this usage of “they” seems to be becoming more common in everyday use.
When to Include Examples?
Examples can be a good way to account for player contexts. You can use them to further explain complex procedures or abstract ideas, but they are best used sparingly as they can be difficult to write and are time-consuming to read. (I almost never read them, probably out of some misplaced sense of pride.) You also have to make sure that any rules text in an example is also in the relevant section. Knowing when to use examples is a bit of an art, so I will just discuss three cases here: when you are having trouble explaining something, when a rule is unusual or violates a convention, and when you know from experience that players struggle with a rule.
When You Are Having Trouble Explaining Something
If you have spent any time writing rules for a complex game, you have doubtlessly found yourself writing and rewriting something that in your head is perfectly simple, but for some reason is impossible to describe with just words. This is a good time to take a step back, stop trying to explain abstract or general concepts, and instead just write-out a hypothetical sequence of events. (Sometimes when I write-out an example I accidentally discover a better way to explain things and then realize I don’t need an example after all.) I ran into this trying to describe the rules for issuing shares in 21Moon, because there are several limits on how this is done. Here is what I came up with:21Moon Rules wrote:A corporation may issue stock from its IPO to the bank pool, receiving the value of each issued share from the bank.The layered constraints on how many shares can end-up in the bank pool was a bit tricky to describe, and after writing it out a few times I felt that it was ripe for misunderstanding, so I added the two brief examples to clarify how the various rules interact.
The number of shares that may be issued is limited. The 50% limit on shares in the bank pool applies. Additionally, after issuing, the bank pool may not hold more shares in the corporation than the total number of that corporation’s shares owned by players and other corporations.
Example 1: During stock round 1, Beth starts Deep Space Explorers by purchasing the president’s certificate and one other share, for a total of 30% held by players. During the following corporate round, she can issue up to three shares of DSE.
Example 2: Paul is the president of Mining Alliance. He owns 30% of its stock, and 20% is in the bank pool. During the corporate round he may only issue one share of MA.
When A Rule is Unusual or Violates Conventions
This is similar to my discussions of naturalized and weak conventions above: when a game violates conventions, that can be a good time for an example, just to make things as clear as possible.
Instances when something about a game seems strange or counterintuitive can be good places for examples. One case from the world of train games comes from Kansas Pacific (Winsome edition):Kansas Pacific rules wrote:The railroad must now pay its shareholders dividends. Divide the railroad’s Income by the number of sold shares. Round up. This is the Dividend paid per share to each shareholder from the bank.The Winsome rules do not provide an example here, and the first time I encountered this rule I was actually fairly confused by it. An example is the best way to illustrate my confusion. According to this rule, if a railroad’s income is $12, and there are five shares, each share pays $3. To me this is quite counterintuitive, because it means that $15 was paid out, not $12. Money was created by a rounding error, which as I remember is a big plot point in Office Space. This makes no sense to me, and my brain wanted the rule to be something about only rounding-off odd shares, or paying dividends proportionally, or anything that did not magically create money. This is one of the limits of the definitional approach: not everyone is good at reading rules literally, because it is hard not to bring our expectations and past experiences with us.
As an example of this, I recently saw a thread on BoardGameGeek where the poster was asking questions about the R6H trains in 1849. At first glance the questions looked odd, because nothing in the text that I could see implied the interpretations that he was asking about. It turned-out that he was bringing knowledge from 1848 Australia to his interpretation of the 1849 rules. Reading literally and disregarding one’s prior experiences is difficult, and it is certainly something I struggle with.
When You Know from Experience that Players Struggle with It
This last case is a bit more specialized, and only really applies if you were involved with playtesting the game, or maybe played a prior edition and are now updating the rules. This case is self-explanatory: if you just know from teaching and playing the game that players tend to struggle with something, that thing probably warrants some extra time and attention in the rules.
A rulebook that takes player contexts into account is going to be easier for players to understand. Misunderstandings easily lead to frustration, so if you want your game to be played and enjoyed by as many people as possible accessibility should be a primary consideration. I hope that the discussion and suggestions above have been useful. If you are interested in further discussing these topics, please feel free to post below, or contact me directly via GM.
If you are writing 18xx rules specifically, I invite you to check-out my Rules Template and Style Guide. Here you will find sample rules and descriptions for many aspects of 18xx games which you are free to copy or make use of, even for commercial purposes. My hope is that it will save you some effort trying to define or explain things that have been defined and explained dozens of times already.
Thanks to Scott Petersen and John Harres for their comments on an early draft of this post.
Rules and usability.
11 May 2023
- [+] Dice rolls
28 Mar 2022
(This has been posted in a few places already, apologies for the spam if this is the third time you've seen this.)
I am excited to announce the existence of a project I have been working on for a while now. I call it "Jason's 18xx Rules Template and Style Guide," or RTSG for short. It is a document that aims to make the rules writing process easier. It does this in three ways:
1. It provides a bunch of generic rules text that anyone can freely make use of. Not sure how to phrase or describe something in your rulebook? Feel free to copy text from the RTSG and modify accordingly. Modification will almost certainly be necessary as the RTSG aims to be quite generic.
2. It includes a short style guide that discusses best practices regarding terminology and formatting.
3. It has a handful of writing tips aimed at people unfamiliar with technical writing.
The RTSG is based on my experiences working with All-Aboard Games: I have indexed the rules to nearly all of the games in print, and wrote the rules for several of them (and for a few upcoming titles).
Right now the RTSG exists as a Google Doc that can always be accessed from my website: https://www.jasonbegy.com/the-18xx-rtsg.html
The document itself provides a much more thorough introduction than this missive.
It is a living document that I will be expanding and improving upon as time allows. I had hoped to put a few more days' worth of work into the RTSG before launching, but I am starting a new job in a few days and I am not sure how much time I will have to work on it in the coming weeks and I wanted to get it out the door, as it were.
I'm happy to answer questions about it, but please read the introductory text first.
I hope you find it useful!
- [+] Dice rolls
One of the least popular aspects of board gaming is rulebooks. Few people like to read them (as evidenced by the popularity of instructional videos), even fewer like to write them (as evidenced by the number of poor rulebooks), and these two trends probably reinforce each other. They are, however, a necessity: after all, if nobody can figure out your rules, nobody is going to play your game!
This article is for anyone who writes rulebooks in any game genre, although I will be pulling examples from 18xx. I use the word “rulebook” intentionally to refer to the kind of long rules documents typical of 18xx and wargames, although much of this article is applicable to shorter documents.
Specifically, this article discusses how to use sections and section numbering effectively in your rulebooks by analyzing specific examples from three 18xx games, however, the emphasis here is more on providing tools to think with than on absolute rules or best practices. Some of the things I will be discussing come down to judgment calls or individual preference, but after reading this you will be better able to think through how to structure your own rulebooks.
The rulebooks I will be discussing are for 1761, 1822, and 1837, all available from All-Aboard Games. I indexed the rules for the first two (among many others), and have been working on a rewrite of the 1837 rules with Scott Petersen as well. I have selected these rulebooks for analysis because they were all written by different people, have different strengths and weaknesses, and use section numbering differently. Each rulebook will be evaluated and analyzed in terms of its readability, referability, and changeability, all of which are defined in the next section.
Although I will at times be critiquing these rulebooks, I want to state clearly that the intention here is not to attack or indict anyone. Rules writing is very difficult, and most rulebooks are written by game designers who didn’t get into designing games so they could write rules! The games discussed here are all excellent titles, and my critiques of their rulebooks should in no way discourage anyone from checking them out.
Full disclosure: I received a one-time small amount of store credit from All-Aboard Games for each of these indexes as a token thank you. I receive no compensation based on sales of these games, or any other All-Aboard Games.
Who Uses Rulebooks? Readability versus Referability
It is useful to think of rulebook users as falling into two categories: readers and referrers. Readers read the rulebook straight-through, and they do it in order to learn how the game is played. They may be reading it for the first time, or reading it again to refresh their memory. They might be reading the rules because they will be playing the game soon, or maybe just out of curiosity. But their goal is to attain a holistic understanding of how to play the game that is sufficient for play. The “readability” of a rulebook refers to how well it supports this usage.
Referrers have a question about the game that they are trying to answer. They likely already have a broad idea of how to play, but they need specific information. When a player picks-up the rulebook to answer a query in the middle of the game, they are using the rules as a referrer, and this is probably the most common use of 18xx rulebooks. The “referability” of a rulebook refers to how well it supports this usage.
(It should be clear that what determines the category is the current use of the rulebook, and we all naturally go back and forth between these categories, depending on the context.)
An effective rulebook, then, is one that caters to both of these uses simultaneously. Readers must be able to learn the game from the rulebook, and referrers must be able to answer questions about the game as easily as possible.
Nowadays, it practically goes without saying that rulebooks can live two lives: as physical objects shipped in the box, and as “living” rulebooks available for download.
Naturally, one of the benefits of the latter form is that it can easily be edited and improved. There are many reasons for doing this, from fixing typos or adding helpful clarifications all the way to changing the rules of the game.
“Changeability” refers to how well a rulebook supports this. Rules can be structured in a way that supports edits and changes, which is particularly useful with the type of complex games I am discussing here.
On Indexing Rulebooks
Having an index in your rulebook makes life much easier for the referrer (who is, do not forget, your player and your customer). Instead of flipping through pages and skimreading (which is highly error-prone; I myself have made more than one rules mistake this way), they can use the index to go right to the relevant section(s). Just by indexing your rulebook you can greatly increase its referability and save your players a lot of headaches.
As a professional indexer I often wish my rulebooks had indexes. For instance, I can never remember where Priority Deal goes after the opening auction. I always have to hunt for that piece of information, and it is never in the same place.
When I index 18xx rulebooks, I index not to the page, but to the section number. By this I mean that the index entries point to the section that the subject is in, not the page as in most book indexes. There are three reasons for this. First, sections are almost always shorter than a page, and so indexing to section number is more precise. Second, this system supports the changeability of the rulebook, because it is unlikely that edits will alter section numbering, whereas the pagination may very well change--correcting an index due to a pagination shift is an indexer’s nightmare. Even if a new section is added, as long as it has a new number it is easy to add that into an index. Lastly, there is almost always a table of contents pointing to page numbers. Indexing to section numbers helps the index work alongside the table of contents by giving the reader another set of access points.
Of course, this means that how a rulebook is structured has a big impact on how the index turns-out, and consequently, how useful the rules are to referrers.
The following sections analyze the use of sections and section numbering in the rulebooks for 1761, 1822, and 1837, and show how this affects their readability, referability and changeability. You do not need to be familiar with these games to follow the discussions, but you may find it helpful to download their respective rulebooks to follow along; this is not necessary, however.
Because the goal of this article is to provide some insight into how to think about using section numbers, the following analyses often include multiple perspectives, and they do not always agree with each other. As indexer Fred Liese likes to say, "there are no rules, only contexts." The same is true of writing rules.
1761 (Rulebook version 1.0)
In terms of its overall structure, the 1761 rulebook has a small number of sections. Looking through the rules, we can see that they only go down one level. For instance, operating rounds are covered in section 11, which is further subdivided into “routes” (section 11.1), laying track (section 11.2), and so on. Section headings are frequent enough that a reader won’t get lost, but not so frequent as to interrupt the reading process. The information is also provided in a logical order, making for a very readable rulebook.
In terms of referability, there are places where deeper section levels might help. Consider the second-to-last index entry, on page 15: “track, laying and upgrading, 11.2.” What this entry shows is that all of the rules covering laying and upgrading track tiles are in section 11.2. However, this section contains several paragraphs amounting to around half a page of text. If a referrer has a question about whether tile upgrades are permissive or semi-restrictive, for example, they will need to read a bit to answer their question. To be fair, this is not a lot of text, but in the middle of a game session referrers need answers as quickly as possible to avoid holding up the game.
That said, it is also important to consider consistency and parallelism throughout the text. Although section 11.2 would have greater referability if broken down into 11.2.1 (covering laying track) and 11.2.2 (covering upgrades), there really aren’t any other sections that could be broken-down further. Subdividing section 11.2 would thus create inconsistency in the rulebook, which risks confusing users, and is a good argument against doing so.
Another advantage of this broad structure is that it supports changeability. For example, if the designer or publisher decided they wanted to add some clarifications, or tweak some rules, the headings are general enough to support this with minimal difficulty. Changes kept under existing headings would not affect the index, and new sections could easily be added and then indexed. The only real effect would be a need to update the table of contents, which can be done automatically.
As an aside, this rulebook has a nice example of how useful an index can be. Section 11.5, “Calculate Earnings” has a lot of information about routes that is not covered in section 11.1, “Routes.” Given the order in which information is presented in this rulebook this makes sense, as trying to cover how routes are valued in the earlier section would be confusing. (Many rulebooks make the mistake of trying to put too much information upfront, which is not the case here.) But a referrer with a question on routes who goes to the table of contents might miss all of the information in 11.5. Naturally, the index entry for “routes” points to both sections, thereby enhancing the overall referability.
1822 (Living Rules version 2.0)
The rules for 1822, on the other hand, are quite different. Section headings are still in two levels, but every rule, which can be a paragraph or sentence, has a unique identifying number, which is often at the third level. For example, section 4 is titled “stock rounds,” which is followed by section 4.1, “introduction,” beneath which are three paragraphs numbered 4.1.1, 4.1.2, and 4.1.3. This pattern is repeated throughout the rulebook.
The advantage of this method is its accuracy: individual rules can be referenced in discussions and indexing is very precise because entries can point to the exact rules covering the topic. The referability of this rulebook is generally quite high. However, it does make the rules a bit harder to read through, because the proliferation of bolded section numbers is distracting to the eye and interrupts the flow of reading.
On the other hand, while this format allows for a very precise index, the index itself can be hard to read. For instance, under the entry “minor companies,” there is a subentry as follows: “initial share values, 3.3.4–3.3.5, 4.4.1, 4.10.1–4.10.2.” When grouped together like this, three-level section numbers are challenging to read. Sometimes this can be overcome by indexing differently, but that often entails wordier entries that will also take-up more space.
This example also demonstrates a little-acknowledged benefit of indexing: it reveals things about your rulebook that you might not have realized, and in turn suggests editing strategies. This entry shows that minor company initial share values are discussed in five separate sections, and so it might make more sense to use a single section to discuss the topic. Then in-text references to that section could be added where relevant, thereby potentially saving some space in the rulebook and keeping the information well-organized.
However, the downside to this method is that the user may find themselves constantly flipping around through the rules in pursuit of these references. Repeating information in all the places that it is relevant uses more space, but it makes things easier on your user. In indexing this is known as “double-posting,” and is standard practice. For example, in the 1822 index, “London” has its own main entry with five subentries, but London is also a subentry under several other main entries. This increases the odds that the referrer will find what they are looking for.
The choice between these two methods is largely a stylistic preference, however, and one of the advantages of having an index is that information that has been scattered throughout the rulebook is neatly tied together.
Another interesting thing about the 1822 rules is the small number of main sections: there are only six level-one headings covering the meat of the rules. As a result, the individual rule numbers tend to get quite long and hard to read, like “5.12.11.” Naturally this becomes easier with practice, but for users who are used to standard decimal numbers it looks odd and takes a moment to comprehend. Having a larger number of level-one headings would help mitigate this problem by spreading everything out.
From an information hierarchy perspective, however, this structure makes a lot of sense: most of the rules cover things that happen in stock rounds or operating rounds, and so most of the rules should be under these two sections. The table of contents shows this to be the case: the core of the rules are on pages 2 to 24. Of these, stock rounds are covered on 10 to 14, and operating rounds from 14 to 24. Long section numbers are a natural side-effect of this method because everything is grouped together.
One way to address the number readability problem would be to create more main sections, trading-off the accuracy of the information hierarchy for usability. For instance, “first turn housekeeping,” section 5.6, could be moved up a level to its own main section. Although it is technically part of the operating rounds, it won’t happen in every round, unlike most of the other rules under ORs. Further, because these rules are unusual (in 18xx terms), it is likely that players will be referring to them often, and making them into a top-level section would also make them more visible.
As noted, such a change disrupts the information hierarchy that has been established. Neither method is strictly better, but these are important things to consider when structuring your rules. If the rules were edited such that section 5.6 became its own first-level section, it would then make sense to move other, similar subsections to the top level, for consistency's sake. Section 5.18, "Minor Company Acquisitions," would be a good candidate for the same reason that section 5.6 is: these rules are likely to be referred to often because they will not come up in every operating round.
What if these rules need to be modified later? A dense numbering system like this, as opposed to 1761’s, makes changes harder. If a rule has to be added in the middle of a section, it can have a cascading effect on the following rules, thereby requiring updates to the section numbers in the text and corrections in the index, which introduces the possibility of errors creeping-in. Incidentally, the version 2.0 rules illustrate this perfectly by way of the 1822+ expansion. Rules covering this expansion are formatted in red text, and you can see that they are often (but not always!) added into existing sections. Sections such as 2.1.7 and 5.18.8 are exceptions, and required adjustment of the following section numbers.
1837 (Rulebook version 3.0 draft) - Link
The rules to 1837 have a bit of a storied history: the rules in the original, German-only version of the game were just a list of differences from 1835, much like how Winsome’s 18xx games come with rules differences from 1830. This original document was translated and expanded upon at various times by Stuart Dagger, Paul Work, and John Webley, resulting in version 2.0 of the rules.
When I received my copy of the game in 2018, I saw that the rulebook could use some further help, and I contacted Scott Peterson about it. This lead to a large collaborative restructuring and rewriting effort; the current version is linked above.
From the outset, I decided that I wanted to strike a balance in section numbering: three levels for greater referrability, but not to the point that every rule is individually numbered. 1837 is similar to 1822 in that there are a lot of rules (more so than in 1761), and many of them are unusual, so the rulebook benefits from having a third heading level.
For instance, the rules covering the city of Bozen, Italy and Bosnia-Herzegovina are all unusual, in 18xx terms, but are also too short to warrant being placed at the top level. As such, they were put into their own third-level sections (10.4.2, 10.3.3 and 10.3.4, respectively). This helps the readability because it signals to the reader that these sections will be covering lots of particular exceptions, and helps the referability by collecting these rules into easy-to-find locations, which in turn makes for precise indexing.
Keeping the section numbering broad was also important for changeability. The currently available version has been online for public feedback for over a year (as of the time of this writing), which has been invaluable. This has inevitably meant numerous changes, and the section numbering allows us to make improvements in response to feedback without causing problems elsewhere in the document.
The current version also uses a large number of top-level sections, 14 in all (although this ended-up being fewer than in version 2.0, which hardly used any subheadings. For instance, although forming the national railways happens in the ORs, putting these rules under “operating rounds” adds a lot of rules to that section which are not often used; this is analogous to the "first turn housekeeping" example in 1822. The same goes for the coal company exchanges. These sections were placed at the end of the rulebook because players will not need to refer to them in most ORs, and they are easier to find when needed.
Another way of supporting rulebook readability and referability is the use of "see" references which point the user to a related section elsewhere in the rulebook. This is useful for readers because if they have a question or do not understand something as they are reading, the "see" reference signals that the concepts are explained more fully elsewhere. This helps avoid overly frontloading rules. For referrers, "see" references increase the odds that they will find what they are looking for.
For example, section 7, "Corporate Entities," contains many "see" references to other sections that give fuller descriptions of the different types of companies, signalling to the reader that more rules will be forthcoming. But it also supports referability: suppose that the user has a question about minor companies. They look up "minor companies" in the index, and they see the first entry: "about, 7.2." Section 7.2 gives an overview of how minor companies work in this game, and it also has a "see" reference to section 10.7.2, which specifically covers how minor companies operate. If section 7.2 did not answer their question, they already have an idea of where to look next, without having to turn back to the index.
The current version of the rules is not yet indexed, as it is still being tweaked here and there. However, if I were to index it at this moment, it would look something like this:
track tiles, laying, 10.3
--in Bosnia-Herzegovina, 10.3.3
--in Italy, 10.3.4
--and mountain railways, 10.3.2
--terrain costs, 10.3.1
track tiles, upgrading, 10.4
In terms of usability, this hypothetical index is easier to read than in 1822, but more precise than 1761. Even when referring to third-level sections these entries are easy to read, because at the third level the rules sections cover more material than one or two rules, and so there is no need to use spans of difficult-to-read section numbers in the index. At the same time, the use of three levels of headings enables a more precise index than just two, greatly enhancing the referability of the rulebook.
The structure of a rulebook has a direct impact on its readability, referability and changeability, all of which affect your players' experience of your game. As such, it is important to think about all of them throughout the writing and editing process. However, writing rules is more art than science, and what is right for one rulebook may not be right for another. But by thinking through the issues and examples discussed above, you can hone-in on the best structure for your rulebooks. A well-structured rulebook goes a long way towards helping your players get into your game, and stay there.
- [+] Dice rolls