Why Blades in the Dark feels less swingy than d20 – and why the bell curve (and variance) aren't the main reason

franz101 · 2026-03-20T12:44:03+0000

Retros_x said:
good point. Swingy is for me that you are less certain on a certain outcome. I think thats why people say 3d6 are less swingy because you have a higher chance of the result being in the middle of the bellcurve. For a linear distribution everything is possible every outcome has the same chance. So more swingy in that sense.

Good point – swingy can mean different things to different people, and you're right that for many the term is tied to "how likely is the result to land in the middle vs. the extremes." I think that's exactly why the standard "bell curve = less swingy" explanation feels incomplete to me.

My numbers show that at 3–4 dice, the extreme outcomes (pure failure and critical success) are already quite rare (6–13 % range), similar to PF2e crit chances (~5 % each). So the edges aren't dramatically more common in d20 than in Blades pools – which means the distribution shape alone isn't the main driver of the swingy feeling.

For me, the bigger factors are:

Progression via fixed, non-shiftable templates (chunky jumps in Blades) vs. smooth 5 % shifts (d20/PF2e)
The label shift via Position & Effect, which turns what would be a miss in d20 into a partial success in Blades → one more "positive" label and one fewer "negative" label, compressing hard failures to ~5–10 %.

That's why competent characters feel more reliably competent in Blades – not primarily because extremes are rarer overall, but because the system actively minimizes catastrophic failure and re-labels middling rolls as partial wins.

Bill Zebub · 2026-03-20T13:03:41+0000

franz101 said:
Good point – swingy can mean different things to different people, and you're right that for many the term is tied to "how likely is the result to land in the middle vs. the extremes." I think that's exactly why the standard "bell curve = less swingy" explanation feels incomplete to me.

That's why I made the comment about whether "the amount you are over/under matter". If you are just aiming for a specific DC, then anything equal or above is success, anything below if failure. The shape of the curve (or lack thereof) surrounding that target is irrelevant.

The only thing that reduces swinginess is repetition: a combat system with lots of rolls is going to be less swingy than a system that resolves it in a few rolls.

franz101 · 2026-03-20T13:20:26+0000

Bill Zebub said:
That's why I made the comment about whether "the amount you are over/under matter". If you are just aiming for a specific DC, then anything equal or above is success, anything below if failure. The shape of the curve (or lack thereof) surrounding that target is irrelevant.

The only thing that reduces swinginess is repetition: a combat system with lots of rolls is going to be less swingy than a system that resolves it in a few rolls.

Thanks for the reply.

I think we're talking past each other a bit. My point wasn't about binary success/failure or repetition reducing swinginess.Both systems have four outcomes (Crit Miss / Miss / Success / Crit Success in PF2e, Failure / Partial / Success / Critical in Blades), not just two.The percentages of those four outcomes are actually quite similar in the mid-to-high competence range (edges ~5–13 % in both). So the distribution shape itself isn't the main driver.The real difference is:

PF2e has one smooth, continuously shiftable curve (linear 5 % steps)
Blades has fixed, non-shiftable templates (discrete jumps when you add dice)
plus the label shift via Position & Effect, which turns many would-be misses into partial successes → three positive labels and only one negative.

That's why competent characters feel more reliably competent in Blades – not because of bell curve or more rolls, but because the system actively compresses catastrophic failure.Does that clarify where I'm coming from?

Retros_x · 2026-03-20T13:41:44+0000

franz101 said:
That's why competent characters feel more reliably competent in Blades – not primarily because extremes are rarer overall, but because the system actively minimizes catastrophic failure and re-labels middling rolls as partial wins.

I would argue that feeling competent and results less swingy has nothing to do with the extreme results on the outside of the curve, but with the average results (or more precise the results within one standard deviation of the mean) coming in more reliable.

franz101 · 2026-03-20T14:03:20+0000

Retros_x said:
I would argue that feeling competent and results less swingy has nothing to do with the extreme results on the outside of the curve, but with the average results (or more precise the results within one standard deviation of the mean) coming in more reliable.

I agree that swingy can mean "uncertainty about the outcome", and yes – bell curves make middle results more likely, which reduces uncertainty around the mean. But my point was specifically that the extreme results (pure failure and critical success) are not significantly more common in d20 than in Blades at comparable competence levels (both ~5–13 % on the edges).
So the "extremes equally likely as middles" explanation doesn't fully hold when you look at the actual four-outcome distributions. The feeling of reliability in Blades comes more from the label shift (miss → partial) and the fixed-template progression than from bell-curve math.