How I Made My Stats Better

For my ChoiceScript workshop (materials to be made available later!) I wrote up how I messed up my stats when I first started my game, and how I made them better. It’s an example of me running afoul of the Stats Disease pitfalls @Brian_Rushton documented in his excellent “What I Learned From Playing Every ChoiceScript Game” post.

Before I started my game, I thought I knew what it took to pick good stats. I’d played numerous Choice of Games games. I’d read articles and posts on selecting stats. I took the common approach of dividing my stats into skills, which would go up as the player gained in ability, and personality traits, opposed stats that would let the player express what their character was like.

You’re a professor at a magic school, so the skills reflect what’s critical in that job:

  • Lab Skills. Practical magic.
  • Library Research. Wresting information from magical tomes.
  • Logic. Reasoning, especially about magic.
  • Rhetoric. Use your words!
  • Political Insight. Use politics to your benefit.

I was pretty pleased with myself. I had skills that grouped nicely into magic-focused skills (Lab Skills, Library Research, and Logic) and interpersonal skills (Rhetoric, Political Insight, and Logic). Sure, I’d need to make sure not to make Logic overpowered, but I could handle that.

Then I chose the personality traits.

  • Structured/Intuitive. Do you plan or wing it?
  • Straightforward/Circumspect. Do you say what you mean plainly?
  • Steadfast/Flexible. How willing are you to make changes?
  • Tweedy/Modern. Do you favor historic or modern approaches.

That screech you hear is the train leaping off the tracks and plunging down the mountainside. I messed these up but good.

To start with, Tweedy/Modern is tough to involve in many choices. It’s not a broad enough category, and it’s not all that interesting.

Far worse, though, is Steadfast/Flexible. It’s awful. The other stats already capture a player’s flexibility. If they make varying choices, the personality stats will hew closer to 50. This stat could also influence players not to make interesting choices because they’re emphasizing steadfastness.

Then there’s what happens when you combine this personality stat with the others. Can you be flexibly structured? Is flexibly tweedy even a thing? Ugh, ugh, ugh.

If I’d been clever, I’d have thought about how the stats work in combination before I started writing a game. I’d have written ten choices involving the stats to see what I thought of them. Instead, I slogged through a good chunk of my first chapter before throwing these personality stats out and creating a new set:

  • Outgoing/Reserved. Do you show your feels?
  • Pathos/Logos. Feels or lasers?
  • Self/Others. Who do you focus on?
  • Driven/Relaxed. Type A or laid-back?

These aren’t perfect by any means. Pathos/Logos conceptually overlap with the Logic skill, though I’m trying to differentiate them through their descriptions on the stats page and how they show up in choices. Even with that bobble, these personality stats are a far sight better than what I started with.

In conclusion, I am going to invent a time machine and take @Brian_Rushton’s post to my past self. Barring that, I hope seeing how I floundered with stats will help others think through theirs before they’re tens of thousands of words into their story.

26 Likes

Thanks for sharing it, man. It’s terrific having authors laying bare the secrets to creating these games, especially the ugly and painful ones, in an effort to make the trailer easier to traverse for those who follow them.

10 Likes

I will be 100% sincere here. If I open a demo and see those stats is almost a guarantee of not buying that game.
Why? I always saw those type of complex stats, the game is a mess, and never know what stat is being used and end with continuity issues where the opposed stat drifts and half chapter I am selfless another half am selfish in the flavour text.

Those stats also seem very complex to understand in a demo. So except the writing is a masterpiece is a pass for me.

That’s a really good idea to write ten choices involving the stats and seeing how they hold up to scrutiny.

6 Likes

I feel like you’d be able to tell the difference between a selfless choice and a selfish choice lol. Those types of stats are usually pretty self-explanatory. One could make an argument that implementing them in a satisfying way could be difficult, but they, or related stats, are in a lot of games because they’re so easy to distinguish. Now I’ll admit, I’d like to see a game that toes the morally grey areas, where the selfless choice isn’t always the ‘right’ choice, but I have yet to see one personally, so it’s usually pretty easy to figure it out.

There is definitely an argument to be made about the crossover between stats though.

4 Likes

If you roleplay and not meta-game most of opposed stats switch wildly. I have beta dozens of cog and It is one of the issues I focus my reviews. There are several of selfish reasons to do good actions an example.
X is the king I want befriend him so I say that I don’t want the reward to gain his trust to betray him later.
For the opposed stat that is selfless. That can make my game flavour text changes completely in same chapter from a scene to the other, It’s very common.

4 Likes

I suppose that’s true. I’ll fully admit that I missed the “flavor text” bit at the end so I thought this was just about the stats themsleves. So, to me this sounds like people are having trouble implementing them in an impactful way and just relegate them to otherwise useless flavor text like internal thoughts or something. But that still seems like it’s pretty clear: if your character does something, good or bad, for a selfish reason, their selfish stat increases, and vice versa. However, I think this may be more confusing for games that use altruism as a stat. I get the impression that people often mistake atruism for ‘moral righteousness’ when in reality the definitions are quite different and in life they can be, and often are, different things. This makes differentiating things difficult when we don’t even really know what we’re measuring.

Understanding, devising and then implementing “opposite pair” stats is perhaps the most common mechanic that trips new author/developers up.

The idea of a mechanic with a gradient value just does not stick in people’s understanding very well.

5 Likes

It is because they forget the second part of good usage of personality stats ASK the player WHY he has chosen the choice

For instance, in my example about the king
A npc would later come and ask you

“Pc_name, it is weird you say no to all that money Why did you do that?”

Then a few lists of reasons to choose from evil to gray to pure real selfless. That is the easiest way to fix opposed asked players why.

2 Likes

So you’re saying, rather than the PC reflecting on their actions, you want them to explain their actions, and this would relfect the stats better? Is that correct or no?

It can be both using both is a way to make games better. The NPc can be your lover, a companion or a parent figure. Talking and discussing why and how. It should be the core element of any game with interactive elements. Reflecting can also do the same in other moments to get more attachment with your own character.

The witcher 3 is a great example of usage of both. Dragon age origins for instance, focus more in the party banter.

1 Like

But unless it impacts something, whether it’s a relationship stat or the story in some way, it’s still just flavor text right? The only person affected by it is you. I guess what I’m saying is, unless these kinds of stats actually DO something, regardless of their implications, do we even need them?

@ClaimedMinotaur and @poison_mara :

Perhaps it is time to split the posts into separated topics?