Wednesday, July 28, 2021

Reputations in the Solitaires Method

This post describes my approach to numerical reputation scores as they appear in the Solitaires Method. These opinions are based on my personal experiences of designing and implementing relationship mechanics in text RPGs. My approach is also influenced by the broader discussion of numerical reputation happening in the video game space for the last 30 years.

Renown is an optional mechanic described in the Dungeon Master’s Guide. It gives the party a set of faction-specific numerical scores starting at 0 and rising as high as 50. Faction quests (and downtime actions) affect these scores. With higher numbers, the party can win titles and support from the factions they help.

The Solitaires Method varies this system in a few ways. It allows the score to drop into negative numbers that represent lower-than-neutral reputations. It also abandons the 50-point maximum in favor of a scale that better matches a single-player, multi-book campaign.

If you are writing in the Solitaires Method, or you are interested in using reputation scores in your own text-based RPG or solo gamebook, these systems can help you extend choice-based consequences into the future. Importantly, they are one of the tools you can use to empower players to change the story as they go.

Why Renown?

Although the renown system tries to enforce consequences, it also tries to be realistic. It is not a piety or karma system because it doesn’t tally every good or bad action. Renown is a system by which the player’s deeds—when witnessed and passed on—alter faction reactions in some scenes.

The Solitaires Method also prefers reputation (renown) over karma for the reliability and realism of it. Reputation does not assume that every “good” action will accrue adjustments. Instead, these actions must be witnessed in a way that will spread among the relevant factions. Players should be able to rely on their ability to get away with various misdeeds if those go unseen. (See also, Simulationism.)

Another reason for reputation is that its adjustments can be subjective. This serves as a safety net for when—inevitably—the protagonist gets an adjustment that wasn’t earned or deserved. Perhaps the player’s route to that point in the story only marginally touched upon the relevant event or moral question from whence came the adjustment. Perhaps—in the player’s mind—the protagonist only rescued the kidnapped orphans to get the reward. When an unearned adjustment occurs, it is the NPCs’ misperception of the character’s actions that triggered it, not the writer’s inability to equally weight the consequence of every story route or failure to see into the protagonist’s heart.

From Simple to Complex

Early video games used simple reputations scores that went up and down. However, even simple video games now tend to track reputation along dual axes. Rather than a single number, computer games can track good and bad deeds separately. A character with good deeds and bad deeds is treated separately from a character with all good deeds or all bad deeds. Unfortunately, this abstraction doesn’t go much farther than a jump from binary to trinary results, letting factions have a good, bad, or “mixed” reaction to the protagonist.

In the Solitaires Method, renown scores remain numerical along a single axis. When I want a mixed reaction, the method leans upon another of tool, event codes. A character that has worked hard to serve a faction might have a high renown score but also possess negative event codes. For example, having “killed the mayor’s horse” or “burned the town,” the protagonist may have difficulty getting aid from the mayor, regardless of the protagonist’s renown. If a high renown score in this case does win an offer of aid, it would come only grudgingly or perhaps at a cost.

It’s also important not to overuse this second tool. Event codes should only stand in for renown adjustments when (1) the affected NPCs do not represent the faction as a whole, (2) the adjustment is so deep as to warrant references in future dialog, or (3) there is some other reason to want a mixed reaction. Otherwise, numerical adjustments to the score are preferred for consistency and to ease the chore of tracking persistent event codes.

Primary Measurements

Renown ratings let players feel the consequences of their actions, in the short term and the long term. They are tools that make the world more responsive to the player. They don’t fit every story. A solitary adventure with no NPCs or supporting characters can’t use—and doesn’t need—such tools. Renown ratings are only useful in stories where other people matter. However, it is my opinion that stories should include people that matter. Without spectators or people to help, stories tend to feel less meaningful.

In the short term, the immediate display of a renown adjustment punctuates a player’s action; the result is felt on a personal level and a mechanical level. Even though the number doesn’t have a concrete meaning in the moment, it portends a future consequence.

“Kenny will remember that.”

If you want to avoid the metaknowledge of an immediately-signaled renown change, you can apply a temporary event code that will later lead to a renown adjustment, probably in the moment the adjustment is felt most poignantly.

In the long term, renown makes that future consequence real. You don’t have to write apologetically if criminals “randomly” come after the protagonist; you know the protagonist went after them first.

Secondary Measurements

In the Eight Petals Argent series, the protagonist can gain and lose renown points relevant to four different factions. These include two sets of “opposite” groups. Many of the reputation-affecting decisions are binary in that the protagonist will gain or lose reputation with one faction or with its opposite, depending on the option selected.

The four factions in the Eight Petals Argent series are the Waterdeep Watch, the [Xanathar’s] Guild, the Slipstone Gang, and the Crimson Perytons. The first two of these represent a nebulous, high-level conflict between law and chaos in Waterdeep. The second two are more personal; smaller-scale factions that represent the tensions inside the protagonist’s friend group.

It’s useful to cast these factions as opposites to support the story’s theme and to help the protagonist “advance” along certain conceptual lines. By tying these developments to the factions, I can reward players for playing within their (very broad) archetypes.

My first pair of factions recognizes that D&D characters are particularly defined by their desire to operate within the law or outside it. By reporting to the Watch, the protagonist is preparing to walk in the sun, to collaborate with authority figures, and to manage conflicts with straightforward dialogue or combat. A low score with the Watch occurs when the protagonist commits crimes, disturbs the peace, or otherwise acts in a shady manor. Naturally, these criminal actions are admired by the Guild and will increase renown with that faction. A protagonist liked by the Guild has opportunities to buy poisons, learn details known only to criminals, and approach problems from the shadows.

The second pair of factions is the character’s gang of friends, the Slipstone Gang, and their rivals, the Crimson Perytons. The protagonist gains renown points with the former by making wise, friend-helping choices, and loses it by opposing the gang’s interests or acting against their “good” morals. Renown with the Crimson Perytons comes from making peace with individual members—though the Perytons leader is always antagonistic—and it is lost by killing members of that gang. A low score with the Perytons is easy to get if the protagonist kills a lot of that gang’s members, making it secondarily a measure of the character’s willingness to fight and kill opponents.

Although these four factions represent developmental directions for the protagonist, their renown adjustments are not independent of their own stories. They are living factions first, thematic tools second. This means the renown scores primarily represent the work the player has done with those factions. They only “secondarily” measure these other factors, the character archetypes and behaviors.

Downtime Adjustments

My use of the renown system allows the player to improving ratings with downtime. This is conceptually consistent with the fifth edition Dungeons & Dragons rules, but I use an altered mechanic for this activity, one that better reflects the scale of a single-player gamebook.

The use of downtime further degrades reputation ratings as a secondary measuring tool for player proclivities since any player can modify the ratings regardless of choices made. This is both good and bad; a player committed to a faction can wipe out an accidental loss by spending downtime to affect that renown score. The protagonist gets to stay on the track its player imagines, but the world is less accurate in its responsiveness to past deeds. This is another reason that persistent event codes are necessary to track exceptionally-influential deeds.

Tracking across books

In a series that uses renown, tracking the range is very important, as is tracking the reasons for the adjustments. Here is an example of the renown adjustment possibilities in book 1 of my series. (“DT” means downtime.)

Before the end of the book (before downtime options), I can use any negative renown rating with the Slipstone gang as shorthand for the one event that can cause it, killing a youth hired to serve as a rival gang’s scout.

As the books progress, this sort of fine detail gets lost in the numbers. However, each time I need to set a renown range as the trigger for certain events or reactions, I’m not looking at raw ranges. I’m looking at all the events and imagining different combinations of them that should trigger whatever I’m planning. I then translate those combinations using their combined numerical ratings.

Other Options

If you want a karma system, Dungeons & Dragons has in-built alignment mechanics you could track along the good-evil and law-chaos axes as seen in various CRPGs like Baldur’s Gate. However, the protagonist’s alignment doesn’t affect the game world; alignment-referential mechanics are almost entirely absent from the game’s current edition. The player might appreciate a mechanic that changes the protagonist’s alignment, but the effect isn’t visible outside that character’s brain. Rather than enjoying a character’s arc through alignment change, I think you will find players unreceptive to it.

The Dungeon Master’s Guide also includes an optional “loyalty” mechanic. If your story doesn’t have persistent factions, but it does have persistent, independent NPCs, the loyalty rules can be leveraged with similar result of renown.


If you’re writing gamebooks for D&D, feel free to find my Facebook group and engage in the discussion with us. Comments here aren’t likely to be noticed.

Monday, July 19, 2021

Loss of Continuity in Solo Gamebooks

Solo gamebooks divide text into segments that are read in a non-linear order. At the end of each part, the player receives options or instructions leading to the next part to be read. “Loss of continuity” is a problem that occurs here—in the space between understanding the instructions and finding the corresponding text to continue reading.

Loss of continuity has two forms. The first form occurs when players get lost trying to navigate to another segment of text. When they start reading at the next segment, the story doesn’t look familiar and confusion sets in.

The second version appears in option sets. When players should decide something based on prior events or observations, some players may have forgotten the necessary clues. Unlike a novel, the reader needs to remember what’s going on in order to progress meaningfully.

Loss of continuity may seem like a small problem, a momentary confusion cured by a bit of backtracking. This is not always the case. The worst losses of continuity occur not with player confusion, but with player certainty. Reaching the wrong text part, a player who misses the signposts and continues confidently may skip over to a different part of the story, never knowing that a problem occurred or finding out too late to do anything about it.

Losses of continuity are ubiquitous to the solo gamebook experience. No one has been able to prevent them entirely, but there metastructural tools that solo writers can use to minimize the occurrences.


 “Turn to page 124.”

Anyone who’s played physical gamebooks knows to mark the old text part before turning to the next; keeping a finger in the prior page lets you go back if you have a problem. It’s not just for cheating—you’ll sometimes need to recheck the instructions to make sure you’re in the right place. For example, you get to page 124 and start reading, but the location is ambiguous and the events don’t seem to flow from the prior text part. If you didn’t mark the prior passage, you’ll have to retrace your steps through the book to get back to that decision point and check that you read the instructions correctly.

In electronic books (PDFs), the problem is similar. You get to the new text part and start reading, only to wonder if you’re in the right place. Although hyperlinks solve some issues, they present several new quandaries.

It takes very little brainpower to click a hyperlink. When flipping physical pages, the mind actively engages with the title or page number as part of the search process. Conversely, clicking a hyperlink doesn’t require the reader to process or retain the identity of the target segment. As clicking hyperlinks becomes routine, players stop thinking about what they’re clicking on, trusting the hyperlink to do the work.

Loss of continuity can occur when the player arrives at the targeted segment of text. Not having committed the identity of that segment to short-term memory, the player may be confused if multiple segments of text are visible on the screen. This is common when each page of the document hosts multiple text parts. The player is left to scan each segment in hopes of finding one that matches the story’s continuity.

The gamebook’s routing can fail dramatically when players get onto the wrong text part, thinking they’re still on track, or if the PDF creator misplaced the hyperlink’s target. It’s very difficult to navigate a PDF back to the prior text part to check you’ve clicked on the right hyperlink or to confirm which text you’re supposed to read next.

We’ve had a long time to solve this but haven’t been able to do so despite the advance of technology. Instead, we use these tools to mitigate the problem.

1. Landing Text

In the Solitaires Method, the preferred anatomy of a text part includes “landing text.” After the title, the first text the player reads should repeat or affirm the action or decision that led to that text part. Even if the case of a hyphen prompt, where there was only one option to proceed from the prior text part, the new text part should repeat the last thing the player read. This repetition may seem cumbersome, but there’s research that shows players welcome strong landings.

Landings solve most problems with finding the correct text part on the page and they remind players of what’s transpired.

The drawbacks include having to write narrative text in a way that can kill the “chapter cliffhanger” effect, that pause between text divisions that draws out a dramatic moment. Having to repeat the last thing dampens the effect of this narrative tool. Additionally, because players will come to expect landings, you can’t fail to put them in every text part lest the players think something is missing.

2. Settings Instructions

Most players use Adobe’s reader for PDFs. Using a particular page view setting makes each hyperlink’s targeted text (presumably the title of the linked text part) appear at the top of the screen. This usually prevents hunting for the right text from among multiple parts on that page, though there is still room for confusion if the page has multiple columns of text.

You can instruct players to use those settings if your adventure begins with a set of gameplay instructions. However, players might use other PDF-reading programs, might not save those settings between sessions of play, or might find the instructions too complicated. Even more likely, those “correct” settings don’t display well because of your players’ screen sizes.

3. Proper Testing

Broken hyperlinks reliably create loss of continuity the player can’t avoid. And they’re all too common. Testing in the proper way and at the proper time is essential. When you think your PDF gamebook is ready to publish, run a final “test case,” methodically checking every hyperlink to make sure it leads to the right text part. This can take hours, but it’s worth the effort.

General testing isn’t enough to find broken hyperlinks. Unless the adventure is quite linear, there will be routes through your story that none of your testers use. You can’t rely on them finding everything, thus the need to test each link yourself.

This step must come at the end because edited prose or late-added art can change the position of your text on the page or the number of pages in your product. Depending on your publication software, this can change your link targets. Therefore, your hyperlink test case must be one of the final steps of production.


The second type of lost continuity occurs when players forget important details. There’s no way to predict which ideas will stick in a player’s mind, or for how long. Moreover, players may put the adventure down to take a break, coming back to the noted text part later. You don’t know how much real-world time passes between the player reading a particular detail and the player having to make a choice based on that knowledge.

This type of lost continuity can cause problems in option sets. It exists because of the gamebook’s segmented nature. In a novel, a player can reacquire forgotten details by scanning back up the page. However, in gamebooks, the player can only scan as far back as the top of the current text part. If that begins at a point after the critical information, the player is unable to reread it and must select from the options without full knowledge of their import.

The core problem is the separation of option sets from their supporting descriptive text. This is common in the metastructure of “recurrent nodes,” where the player can return to a short text part and choose another option from a set, repeating this process at least once. This small, recurrent text part often omits the necessary exposition for decision-making there, usually on the theory that the player shouldn’t reread a lot of text each time.

Here are some examples of separation problems:

1. Two-Part Decision Hubs

While it’s good to limit the length of the descriptive text in a hub, that text should still describe the forecasted consequence for every option. For example, an option set with three directions should remind the player what’s in those directions (if the story has revealed that information already). If the player already knows that the path toward the river leads to the character’s ultimate goal (the capital city), that should be in the narrative text of the hub or part of the option itself.

2. Omitting Hubs

The problem is harder to deal with if you avoid recurrent decision hubs. If you present an option set where multiple options will be chosen in a “multiplying” track, it’s easy to skip the foundational exposition further down-track.

Example: Choose one of three options (doors). After opening a “wrong” door, you find an option set that includes only the two doors you haven’t tried. But do you remember the clues about the correct door here, where you’re facing only two options?

Aside from requiring a lot of text and text parts, a multiplying track has the potential to skip the description of each option each time an option set appears. Be wary of this problem and construct winnowing tracks carefully or create a “test case” to apply prior to final editing to find such errors.

3. Hyphen Prompts

When one text part directs the player to another, with no alternative, that lonely prompt is called a “hyphen prompt.” The next text part might receive a reader coming from multiple routes (binding branches into a single route), or the separation might exist because the narrative text was simply too long for the page. As above, this can separate the description or stakes from the coming set of options.

Again, this requires checking all your option sets and making sure the relevant guiding details also appear in that text part.


Loss of continuity is often down to a failure of comprehension. The reader doesn’t understand the meaning or inference of your text. This is the writer’s fault; you can always do better with your descriptions. However, the constraints of locational mapping make this problem particularly difficult to improve.

In locational mapping, like a dungeon room, the narrative text often describes the room from a neutral, central perspective. This is because the character may have entered that space from multiple directions. Even if there’s only one “entry” and one “exit,” true locational mapping is recursive, letting the character move into and out of the space repeatedly. You can’t refer to the player’s prior direction of travel unless you write a different version of the room for each entry point.

This is less about reading comprehension; it takes a different sort of deduction to remember you went south from the last room, thus the north-leading option would be backtracking. The loss of continuity occurs not in the context of your narrative, but rather the decisions and movements previously made.

To reduce this problem, refer to the player’s past choices whenever you can safely do so. For example, a dead-end room can describe the door option as “back the way you came.”

Whenever possible, use map images of each location. This can reduce the chance for directional confusion, particularly among diligent players who are mapping as they go.

Finally, when a room has too much potential to cause confusion (or backtracking can create harsh penalties), try writing multiple versions of that room, including all text parts for interacting with its contents, so that you can explicitly mark which exit option is “backtracking.”