Why can't I love fixed target numbers?

05 Nov, 2026

This post is the first in a probable series on, for lack of a better term, "game feel." (Please someone give me a better term.) As I work on my own game systems, I think a lot less about the math behind individual mechanics and more about how they feel during play. These posts are heavily subjective and not really making any kind of an argument: I'm just exploring my own tastes through writing.

Over or Under?

Scenario A. A PC, Daisuke, is attempting a risky action and the referee asks the player to test their Strength to see how it goes. Daisuke has 14 Strength, and to make a test, the player must roll a d20 and get equal to or under. This gives Daisuke a 70% chance of success.

Scenario B. Daisuke's still attempting a risky action, and that action still requires them to test their Strength. Here, Daisuke has 5 Strength, and to make a test, the player rolls a d20, adds the appropriate attribute, and tries to get 12 or higher. This gives Daisuke a 70% chance of success.

For many, the longer they play RPGs and the more different RPGs they play, the specifics of the resolution mechanics get less and less important. Rolling under vs. rolling over, a d20 or a d12 or 2d6 or d100, whatever, what really matters is what's going on around that resolution, what you're resolving and how that resolution is applied during play. Through that lens, there's not really a whole lot of difference between A and B. Some players might prefer the lack of arithmetic in A.¹ Others might feel it's more intuitive to want to roll high, so they'd like B.² In both cases, referees who prefer not to have to set a target number for each roll are chilling. Most of the time, once the group plays a session or two and gets comfortable with whatever aspects didn't immediately come naturally, the distinction stops mattering too much.

So why am I still less comfortable with scenario B?³

What Does a Target Number Represent?

With a roll-under system, the target number is usually a stat or skill's value on a character sheet. It clearly represents the character's capability in that area. If someone has 14 Strength, they always have a 70% chance at succeeding at a task that tests that stat. That's their capability.

If we're rolling over with a fixed target number instead, that's a little less clear cut. That target number is the same for everyone, so it can't represent a character's capability: that's what the number added to the roll does. That means it must represent something external to the character. Naturally, to my mind, that means it represents the difficulty of a task--and it feels a little odd to say that difficulty is always the same.

But that doesn't make sense, right? Because the difficulty is always the same when you're rolling under, too.⁴ The difference is that it's not quantified. In fact, the difficulty always being the same is sort of obfuscated when you're rolling under. It's not when you're rolling over vs. a fixed target.

This is, of course, a purely aesthetic thing. It's a very subjective thing. I'm sure many wouldn't even really see it that way at all. But I think it's part of why I don't really like it as much as I like rolling under.

There we go. Question answered, right? I don't like fixed target numbers to roll over because it draws my attention to the artifice of every risky action being equally likely to fail. And since I prefer not to set a target number as a referee if I don't have to, that means roll-under is my happy place. Right? Surely there's nothing that might complicate thi--

Perturbed by the Apocalypse

I don't feel the same way about resolution mechanics with multiple ranges of outcomes. You know, your classic PbtA 10+, 7-9, 6- deal. Those are, by any definition, fixed target numbers. But all of a sudden, now that there's not just one target, I'm not really processing that as representing some kind of objective, external "risk" or "challenge." In fact, my brain processes it very similarly to how it processes rolling under.

The ranges somehow obscure the artifice again. What the hell?

I don't really have as clear of an answer to this one as I did to the above (and that wasn't even all that clear!), but I do have some guesses.

1. A referee usually doesn't set multiple difficulties.

If I'm setting a DC for a roll in D&D, I'm setting one DC, not two. Other DMs might not even really set a DC at all and run completely on vibes. (Some do this and don't admit it, others actively advocate for it.) In either case, there's no preexisting association in my brain there to see multiple target numbers, or a range of them, as attempting to represent anything tangible.

2. PbtA games are often explicit about rolls representing something else.

Many⁵ PbtA games are explicit about that dice rolls aren't about a PC's skill or ability. Instead, they answer different questions: "How does this uncertainty resolve in the unfolding story?" for some, or "Who gets to say what happens next?" for others. In that framing, the target number never represents "challenge" or "risk."

Of course, that's far from unique to PbtA games. Mythic Bastionland uses its Saves not to ask whether a Knight can achieve their goal, but whether they can avoid a negative consequence, for example. That's how I prefer to use dice rolls in my own work, too, so using that more PbtA-style framework feels just as natural for me.

3. Moves are scaffolding.

Most PbtA games that use the classic 2d6+stat with outcome ranges also scaffold that with discrete moves. Of course, there are PbtA-inspired games like Offworlders or FIST that don't use moves, but for those that do, the text of the move itself and its potential outcomes also pulls my mind away from thinking about what exactly it means that a PC rolled an 8.

On "Challenge"

Maybe the deeper reason I don't like fixed "external" target numbers is that I'm increasingly interested in expressing the risk or challenge of an action through the fiction and not the numbers. I'd rather just not call for a roll if something is too easy or too hard, or let an especially skilled character have greater impact, or an especially bad situation risk greater consequences. I'm not even really into advantage/disadvantage these days.

There's nothing stopping me from doing that with a fixed roll-over target number. But because I see that target number as some "external difficulty," there's just that little bit of dissonance there. And if I'm making my own game, why not steer clear of things that cause me to feel that dissonance? Even if it's just a "me thing,"⁶ well, I'm the one writing it, so I'd better like what I make, right?

So I think that's the conclusion, for now: I prefer not to set a target number on a roll-by-roll basis, and I prefer when a dice roll doesn't limit itself to a simple "can the PC pull this off?" question. Put those together and I'm going to gravitate towards rolling under, or a different framework altogether.

This won't be the last of these. For one thing, while I've ruled out one approach to task resolution, I haven't fully settled on one. Maybe next time I'll talk about changing die sizes vs. changing static numbers. Or something else altogether.