I recently finished playing and reviewing every Choicescript game. Based on what I learned, I feel like it might be helpful if I share patterns I noticed in popular vs unpopular games. It helps that I have written what is possibly the least popular Choice of Games title of all time, In the Service of Mrs. Claus. So when I talk about what doesn’t seem to do well, I talk from experience!
First, a few caveats:
-This isn’t a guide to making a great game, it’s just examples of features common in good or bad games. Correlation is not causation, and the next great game might go against all of this advice.
-Games are like that boat in the Suez canal. You spend months or years writing them and they are huge beasts, difficult to turn or control. If you finish at all, it’s a real achievment. Some games just end up stinking and you can’t do much about it besides wrap it up.
-This list is certainly tainted by my personal opinions.
-There’s not much writing advice here. It is a huge factor, but not one I’m qualified to discuss.
-Every game that I find flaws with has good points, too, and my game is one of the worst-selling.
To help reinforce those caveats, for each rule I list I’ve provided counter examples of good games that ignore them to show that none of these rules are hard and fast.
With that in mind, here are the four things that I noticed the most!
1. Your first chapter is vital!
Your work of interactive fiction is both a story and a game. Your first chapter serves simultaneously as a tutorial for the game, a character creation menu, and the hook for your story. For most Choicescript games, the first chapter also serves as a book cover or blurb, as many links to Choicescript games just link directly to the first chapter.
This makes the first chapter absolutely critical. One thing the lowest tier of games sales-wise have in common (besides those which are extremely short) is bad first chapters.
What makes a first chapter good or bad? Here are some questions you can ask yourself:
Is my first chapter as interesting as the rest of the book?
Several of the lowest-selling games have purposely slow first chapters with later ‘big reveals’ that make the story much more exciting. While this can pay off with reader involvement, there is no payoff if no one’s reading in the first place.
When I discussed this with a friend of mine, they pointed out that famous books tend to start much slower than games narrative-wise, and that habit can sometimes carry over. Having dozens of pages of Bilbo’s birthday and Frodo walking is great for Tolkien, but doesn’t do much for gamebook sales.
Counterexamples: A few very well-written and well-regarded games have slow starts. Heart of the House is one of my top 2-3 Choicescript games, but I remember I looked at it a year earlier and didn’t buy it because the first chapter seemed boring. Asteroid Run: No Questions Asked is a fantastic, very fun sci-fi adventure that, I feel, starts very slow.
Is my first chapter a good tutorial for the rest of the game?
Your work of interactive fiction is, again, both a story and a game. For many people who buy your game, this will be their first Choicescript game, and even for experienced players, your game will have its own quirks that need to be taught.
If your game is going to be about resource management, you need to show what the resources are and how they’re used. If the game is going to be about moral dilemmas, you should show moral dilemmas in the first chapter that are similar to ones you’ll have later.
In my own poor-selling game’s original version, the first chapter featured no NPCs and a great deal of slow-paced flashbacks, both very different from the rest of the game.
In a twitter summary of a talk by Jon Ingold (80 Days, Heaven’s Vault, Pendragon), Twitter user Axel Hassen Taiari (@axel_hexed) reported the following comments:
“C) important that a game communicates early on what kind of game/experience it’s going to be. Not just thematically or diegetically (eg a Spider-Man game has to have Spider-Man) but rather give SOME sense of the possibility spaces you’re dealing with.”
One great way to communicate this is to explicitly show the player that choices change stats and say what stats are being increased (Creme de la Creme does this in some of the first chapter, as does VtM: Night Road and Choice of Robots. This is also central to Heroes Rise with the MeChip, as well as the Versus series).
Some games that have great tutorial chapters include Werewolves: Haven Rising, Vampire: The Masquerade – Night Road, Creme de la Creme, Choice of Magics, Psy High, Choice of Robots
Counterexamples: Ironheart is fun and popular and has a first chapter that is absolutely nothing like the rest of the game in genre, or in just about any other way. Silverworld and Stronghold: A Heroes Fate are two other good games where the main gameplay is very different from the first chapter.
Is my first chapter a good character creator?
Imagine rolling a Dungeons and Dragons character (where stats are between 3-18 with 9 or 10 being average) and getting someone with 5 8’s and one 10. That’s what a lot of character creation chapters are like in lower-selling CoG, even in otherwise award-winning games.
In fact, there are several games where you can pass all checks having all stats at 10% except for one in the 20%s. The game just never makes use of the visual space the rest of the stat bars provides.
Here is an example of the stats screen of a good-selling game and a bad-selling game after 1 chapter:
Outside of stats, more customization is better in character creation. I share lots of CoG games with my students. One girl (whose parents immigrated from India) who played Creme de la Creme was thrilled that she could create someone with her skin color and hair type. For myself, I enjoy Choice of the Dragon’s customization.
Counterexamples: The biggest counterexample would be Creatures Such as We, a wonderful and popular game that has no stats whatsoever.
2. Stat disease
As I played through the vast ‘middle’ chunk of Choicescript games which are not bestsellers or worstsellers, I found a pattern that was repeated over and over again in regards to stats, a pattern that differentiated the upper-middle from the lower-middle. It came to the point where I could open up a new game, check its stats page, and immediately know about where it would be placed in bestseller/rating lists, based on a problem that many games have.
I call this problem ‘stat disease’. What are the symptoms of stat disease?
-My own (low-selling) game, where there is a skill called Cheer, an opposed stat called Happy Helper, and an achievement meter called ‘Christmas Spirit’. If a choice is about the power of Christmas love, which of these three skills is involved?
-A game that has opposed stats ‘idealistic vs cynical’, ‘indifferent vs empathetic’, ‘honorable vs shameless’, ‘cautious vs maverick’, and ‘loner vs sharer’, as well as skills ‘charming’ and ‘cunning’ and meters ‘integrity’ and ‘pride’. If you have to make a choice to listen to your boss or ignore them, what’s being tested? Ignoring your boss could certainly be described as being a loner, and a maverick, and not idealistic, and indifferent, and not honorable.
-A game that has ‘charm’ as a skill and ‘guile’ as an opposed stat, and ‘confidence’ as a skill with ‘boldness’ as an opposed stat.
It should be easy to distinguish stats from each other. The very smoothest games are so clear that you never even need to check the stats screen; you can just say ‘my character has always been greedy, of course I can sympathize with the dragon’ and not worry about it.
Some games with great, easily distinguished stats are Tally Ho, Heroes Rise: The Prodigy (the opposed stats especially are very clearly differentiated), Choice of Magics (especially nice how differentiated the different kinds of stats are on the stats screen), and Champion of the Gods. Even in these games, there is some overlap, but most stats are distinguishable from each other.
Counterexample: Creme de la Creme has some stats that may be confusing at first glance. It’s not clear what ‘poise’ vs spirit is. The game overcomes that by explicitly describing what those skills are and aren’t in the first chapter, by name (with the question that marks your greatest skill and weakest skill). So confusing stat names can be overcome, but it requires conscious, guided effort.
Having too many stats getting spread out or generally being very weak
Little differentiation between good and bad stats
These two go together, and both go along with the ‘character creation’ part of point 1 above.
Opening up a stat screen and seeing 15 different stats starting out at 20% is very discouraging, especially when chapter 1 ends up with you only adding 5% to one or two stats. There are very few examples of popular games where you have tons of diverse stats with only small boosts in a few of them at the end of the first chapter.
One reviewer said:
Story captured my interest but stats mgmt hinder fun. There were too many stats to pay attention to. I found myself in scenario where a situation needs 2 or more different stats to align and my character meets only 1 or 2 requirements and none of the other choices can work either because of the same reasons. It ends up being annoying to keep track all these choices and scenarios to get what you want. It almost as if you cant be what you want without being penalized.
The other direction can be problematic, too. Having only 2 or 3 stats can get frustrating if everything depends on them but they are frequently changed without notice.
Stat checks and stat changes are indistinguishable from each other or happen unannounced
Opposing stats easily overriden on accident
Game frequently lowers your skills
The absolute worst thing is when you’re playing a game, focusing on a specific strategy, when suddenly a choice that looks harmless actually sets one of your opposed stats 40 points in the opposite direction of what you’ve been focusing on. You thought you were compassionate? No, you’re actually 75% sadistic!
Reviewers have pointed this out in many games. For instance:
this ruined the game for me, in this game you have personality stats like independence/respectable, and they fluctuates a lot depending on your minor choices, like greeting your friend one way instead of another. A big problem with this is at the end game, there will be major choices that will decide what ending you will get and only a few one will be available
to you depending on your personality stats, the rest will be grey. and since in this game what choices raise or lower stats is NOT CLEAR AT ALL! tries after tries i tried different combination of choices to raise my stat to a certain degree so i can select the ending that i considered to be best and it just never works,"
Choice of the Robots and Vampire the Masquerade: Night Road handle this well by indicating what’s being checked and when. I know a lot of authors (myself included) have eschewed this because it brings players out of the story, but it significantly contributes to the enjoyability of the game aspect.
Counterexamples: The Fog Knows Your Name is a fantastically popular game, but many reviews cite difficulties with the stats. For instance:
It is nice to see another story where stats matter but i think the way in which they are implemented leaves something to be desired. It’s never clear what choices raise your stats or when a stat check is behind a choice. Like for example my character in this was relatively physical, i had something like 63 physical so when confronted with an option to use a branch to defend myself i chose it. But obvious i failed that check because somehow picking up a stick wrenched my arm and made me more vunerable. But i had no way of knowing if 63 was a good level or if it was even checking my physical stat.
(Heavily criticized) Opposed stats are used for built-up skills or personal relationships
By skill I mean stats that represent your accumulated talent or knowledge. Making, for instance, agility an opposed stat to strength means that your character can be an incredible dancer or runner, but if they lift a weight once they become slow and clumsy. Making relationships opposed means that the best way to romance someone is to attack or abuse their ‘opposite’ person. It just doesn’t make sense.
Empyrean had to be completely revised to remove this feature after massive backlash. Treasure Seekers of Lady Luck, a personal favorite of mine, placed rock bottom in the last poll on underrated choicescript games, and it uses these opposed stats for relationships. People seem to really dislike it.
Overall Stat Disease
The following are a great deal of reviews exemplifying stat disease for various games:
God this game has intrigue and I want to learn more about it but the game’s mechanics itself won’t let me be interested. There are way too many stats to keep track of, such that I don’t think the creator properly factored in the crossover of having certain combinations. You will almost always hit walls of choices where none of your stats are suited for any options, and it doesn’t feel intentional or good to play. The game doesn’t clearly communicate which stats mean what and the crossover between stats makes it all the more confusing. I didn’t even get a chance to decide if I was honorable or cunning in the tutorial so I went through the game without the ability to choose either option. And in choices where 2/3 were the honor and cunning choice I’m obviously restricted.
Another review for a different game with stat disease:
Not enough chances to increase your skills.
Most choices seem to not affect the story as you may want them to, since in most cases it seems to fail a check of sorts. It mostly seems you’re just always making a bad choice.
Sometimes choices aren’t as straight forwards as they seem.
A save and reload function would’ve been nice to have; At least in between chapters.
A skill stat for stealth would’ve helped indicate if choices were being made properly when choosing options of that nature.
As I said, [game] lacks meaningful choices and reads more like a non-interactive novel at times. For that, we have entirely too many stats. Considering how the stats don‘t even impact the story all that much, we have too many of them and it gets confusing. Especially because micromanaging them is a chore.
There are a lot of instances where it‘s not at all clear which choice is going to change which stat. And even several instances where it is deeply unfair. For example, let‘s say we have answers A, B and C. Answer A raises a skill stat by 10% (very valuable). Answer B raises a personality stat by 10% (quite okay). And answer C reduces a skill stat by 10 % (really bad)! As a reader, it‘s almost impossible to see that coming.
Genuinely a bit of a pain. It’s hard to raise your stats high enough to do anything, and balancing things almost never works. Rather little variation in paths, despite several playthroughs with different choices;
I kept dying halfway through the game because I couldn’t figure out wtf the author’s choices meant, kept ruining my stats because I couldn’t figure out wtf the author wanted, and, had I known then that I could get a refund for a s*** game, would’ve demanded it. I expected a game that was more plot oriented (I mean, “choice of” means interactive novel, right??) and got a stat-oriented bunch of drivel with a shallow plot, shallow NPCs, and shallow relationships. I was completely unimpressed and found the game annoying as h***.
Games without stat disease
When talking about Mecha Ace, I never missed an opportunity to mention my favorite scene: I focused my skill on ranged weapon with a bit of piloting, so my mech is built for that. In one of the battle, I was given the opportunity to take a position in the formation which I took the frontline. My plan was to cause massive damage to opposing formation before they’re able to wreck ours, so at the start, I’d charge in with my squad. The game mentioned the pros and cons, including the time it’d take to assist my main force if they’re overwhelmed. Guess what? Our forces got overwhelmed. Being left alone, I pick the option to assist them, but the game mentioned that my mech is not the best at moving and my piloting is too low. Wham bam, we take a massive hit and there’s nothing I could do: too far to continue the deep strike, too far to return to the frontline. Power fantasy got fulfilled, but the consequence is brutally fair. At that time, I also realized: I have the long-range build. Why would I go that aggressive?"
From Creme de la Creme:
And unlike Choice of Romance, your lowest stat doesn’t get stuck being super-low level the whole time, because in the course of the story you get tutoring in your worst subject!
For Psy High:
If you, like me, love watching arbitrary sliders go up for acting like a Lannister, you will enjoy this game. If you had a good, or at least not-awful time in High School, chances are you will enjoy this game.
For Choice of Rebels:
With a sprawling, well-written story and intuitive mechanics (a first for me), there’s no reason not to try it out if you’re even mildly interested in the premise.
and the tension in plots are insane.
It’s long, detailed, complex, and gives you a huge amount of freedom to shape the story to your whims.
I fear the author might of written himself into a corner with how much choice was in this installment.
Origins of Stat Disease
It’s my belief that stat disease is caused by authors not knowing how to make interesting interactivity in their games (including me when I wrote my game). They throw in stat checks, but then the game is too easy since people can just pick a favorite stat. So the easiest option is just to make things difficult. Think about it: every symptom of stat disease is a way of making regular stat checks more difficult.
This is bad.
Engagement in your game does not come from making stat checks difficult! The primary use of stats is for the game to remember who the main character is.
Or, as Eiwynn put it, “For the bulk of text-based choice games, mechanics are not the key that unlocks the kingdom.”
So where does engagement come from, then? From having mutually worthwhile and interesting goals that must be pursued simulataneously but cannot all be achieved at once (especially relationships, but also things like saving something you care about that’s dishonorable vs being honorable. Choice of Rebels, Slammed!
The best games are when you can get any one goal you want but you have to sacrifice everything to get it (like Psy High where a dream crush can be dated but you find out they are a terrible person).
The game on the omnibus with the lowest overall rating is one which allows you to select your difficulty, with the default being ‘very difficult’. Difficult stats don’t, in general make for a fun game by themselves.
3. Give your game a satisfying plot arc (no matter the player’s choices)
The shape of your plot
Every game has a plot arc of some type. Here are some typical plot arcs:
Among poorly selling or lower-rated games, many have arc C, where the action picks up slowly and then kind of peters out. Like one review said:
There was a lot of blank space throughout the story. By that I mean several days were skipped to rush you to the main goal at the time. This could have been an excellent time to build upon the lore
The premise had a lot of potential but the execution of the story leaves much to be desired. The ending was completely underwhelming, no build up, no pay off, no climax, it just… happens, written in the same tone as the beginning and middle of the story.
Another review says:
Still working my way through but noticing the game breaks down in the middle of the story.
And the worst (my own game, In the Service of Mrs Claus, before I edited it):
The beginning is a chore to get through. It feels too much like an exposition dump.
It would’ve been better if it had just given you a character customisation first and skipped all the flashbacks that just bombard you and leave you confused on what you’re reading this time.
It’s as far as I’ve gotten cause like I said. The beginning chapter is too boring to get through right now.
Type B arcs are generally only found in the longer games (although Sixth Grade Detective is a shorter, but fun, example). These games are generally episodic, with no rising tension, and can become very dull after longer periods of time.
Type D arcs just stop suddenly, like the end of Monty Python and the Holy Grail.
I’ll shout out Choice of the Petal Throne, because I thought it was an awesome game (incredibly complex worldbuilding borrowed from a pre-existing shared world). It had a really interesting plot happening and right at what seemed like the midpoint of the game it just stopped like a truck driving into a mountain cliff. No real denoument or indication it was the climax.
A review of another game with ‘drive into a cliff’ ending:
I’m sure there are many out there who will enjoy this game. Unfortunately, I am not one of them. The story is definitely unique, and some ideas caught my interest, but it ended rather abruptly and with too many loose ends. It honestly felt like the story was going somewhere but the author cut it short/didn’t have time to develop it.
Only major negative I would say (which other reviewers have pointed out) is that the ending seemed to come out of nowhere and felt a bit rushed. I hope in future if the developer made another game that it would last a bit longer, I feel like there was more potential that could have been gotten out of this game.
Almost all bestselling games have an A arc, where the excitement picks up early on and steadily increases until a cool climax, then a significant (but not too long) denouement. Examples include Choice of Robots, Creme de la Creme, Choice of Magics, Tally Ho, and Heart of the House.
Wren Harper, who helped brainstorm this essay, pointed out that most good games actually have some self-similarity, with each chapter having its own climax and denouement, each higher than the one before:
The same contributor also noted that most static fiction builds a lot more slowly than this, so the best plot arc for a game is not the same as the best plot arc for popular (static) fiction.
Hitting up all important plot points
There are a couple of games where important plot points can be missed entirely if you take the wrong branches, but the game assumes that you knew them all along.
Emily Short famously said:
Your job is to make it as hard as possible for the player to finish your game without understanding your story.
From that same article, she says:
If you have a key story beat, don’t just narrate it and move on. Players skim in interactive stories, especially in choice-based stories where they know they’ll be able to keep moving forward regardless of how well they learned the establishing details.
If you need the player to remember something, give them a choice about that thing. If you can’t let them choose whether the thing happens, you can still let them choose how it happens, or how the protagonist feels about it, or what they’re able to salvage from it.
When games mess this up, it doesn’t show up in reviews because the reviewers literally have no clue what they missed.
A good example of hitting up important plot points is Choice of Robots. Events like the global war happen whether you’re a CEO supplying the military or a bum living with her mom.
Counterexample: Jolly Good, Cakes and Ale has one night where 3 or 4 incredibly important events happen at once. I completed the game remaining entirely at the opera, and that left 2 or 3 threads completely unresolved. But this is a very popular and fun game.