Feats and randomness advice.

[SangoProduction]

So, I was thinking about offering the players two feats that they could take to alleviate the deviations in randomness in rolls. Can you guys tell me what you think? I made them feats as reducing the likely range of events does have consequences, but should be there if people really would like to invest in to reliability over potential.


Reliability
Benefit: You may roll 2d10 in place of any situation where you would roll 1d20.

Slow and Steady
Prerequisite: Reliability
Benefit: You may roll 3d6 or 2d10 in place of any situation where you would roll 1d20.


D20 average: 10.5
2d10 avg: 11
3d6 avg: 10.5

[PoeticallyPsyco]

The feats would be kind of meh, except for the fact that they completely eliminate the chance for an automatic failure or botch on a nat 1, which is probably a decent use for a feat for some builds. And if you've already taken "Reliability", the increased likelihood of getting mid-range instead of low-range would also be valuable to the same kind of build.

In conclusion... yeah, I think they're pretty well balanced. Stronger than the Luck feat line, but that's typically regarded as too weak.

[Jay R]

These are steps along the continuum between maximum variance (rolling a d20) and zero variance (taking 10).

In essence, these feats lets you get some of the benefit of taking ten even when threatened or distracted, but with some randomness retained. That doesn't seem out of line.

[Hiro Quester]

For some builds, critical fails can be devastating, while critical hits don't do that much extra. This seems a good idea for that.

e.g. I once played a kung-fu large cat whildshaping grappling Druid, with a one-level dip in Monk.
With iterative attacks and then claw claw bite (and possible rake), but not an especially high BAB, I was often trying beat high AC by sheer number of attacks, hoping one or two would get through (to start a grapple).

The DM had a chart of really awful critical fails. Most meant that if one attack was a 1, then I failed to beat AC on a confirm roll, then the whole attack sequence stopped (and I had some other horrible penalty kicked myself in the head and am stunned).

Rolling 2d10 to avoid critical fails would have been tempting to avoid that risk. But it does consume a feat.

A better solution might have been to just not have critical fails that so seriously add to the risk of those making many attacks per round.

[ExLibrisMortis]

Natural 1s usually aren't so bad. Hiro's example has homebrew critical failures included, and that's what makes the feat worth using. In vanilla 3.5, I don't think they're worth it, except perhaps on a very highly optimized casting-type who is mainly concerned about auto-failing saves, and not sufficiently protected by the Pride domain power.

[Hiro Quester]

I have been thinking about this more. It's not just avoiding critical failure 1s that is the issue here.

I can see it being a useful option for certain rolls, where you con't care about rolling high, but really don't want to roll low. Like skill checks in cases where there is no bonus to succeeding by a lot, but a penalty to failing.

e.g. My bard character now has +14 to a UMD roll. To use a device with a DC of 20 to avoid failure (or mishap!) he wants NOT to roll less than a 6. Failing means you can't try again for 24 hours. You can't take 10 on this skill either.

On a d20, he has a 25% chance (5/20) of rolling a 5 or less. On 3d6 he has about a 4.3% chance of failure (10/216) (based on this probability chart (https://divnull.com/blog/2009/3d6-probability-graphic/)). That is a huge reduction in risk.

Even on a 2d10 roll, the risk of 5 or less changes from 25% to 10% (10/100 possibilities that add up to 5 or less).

I could see many skill-dependent characters (esp rogues) wanting the option of a reduction in risk of failure on skill checks. I'm now worried that this is potentially too good for such characters.

Have you considered offering it as a magic item that grants this ability 1/day or something like that?

[Gruftzwerg]

Imho this kinda things is better solved with Action Points (https://www.d20srd.org/srd/variant/adventuring/actionPoints.htm).

Just give them the option to reroll the roll for action points.

And always when they complete an important encounter/task , give em one.

This way it still remains something special and they can have a few aces in their sleeves ^^

[Shpadoinkle]

On a d20, he has a 25% chance (5/20) of rolling a 5 or less. On 3d6 he has about a 4.3% chance of failure (10/216) (based on this probability chart (https://divnull.com/blog/2009/3d6-probability-graphic/)). That is a huge reduction in risk.

For a likewise relatievely huge investment of two feats.

Given that, I think these feats are fine. I'd add two restrictions to them though.

First, you have to be at least 3rd level (maybe higher) to take Reliability.

Second, you CANNOT take either of them as bonus feats, no matter the source of the bonus feat- you HAVE to spend the feats you get specifically from reaching a given level that grants you a new feat to take them. Maybe add a level minimum for Slow And Steady too.

[MaxiDuRaritry]

So, do these feats make you immune to nat 1s? What if you roll snake-eyes?