2d10 for Skill Checks

I have run many, many games now over the years that have used the standard d20 for skill checks, and over time have grown more and more disillusioned with it. Principally due to the fact that the d20 die roll produced way too large a variance for me, resulting in a PC's personal modifier having much less meaning in the grand scheme of results. A PC with a +6 to a skill versus one with a +0 did not make a perceptional difference that often... when the +0 could roll 16s to 20s while the +6 would roll 3s, 5s, 6s etc. The lack of a bell-curve meant modifiers had less import.

As a result of this, I decided for my two new Eberron games that I just started this past fall that I was going to replace the d20 with a 2d10 system for skill checks-- roll 2d10 and add your modifier. And due to the fact that this would now produce a bell-curve (bringing almost half the die rolls into the 9 to 13 range and very few at the extremes) a PC's ability modifier + proficiency would have a much greater impact. Those that were all-in on certain skills (with high ability score, proficiency and/or Expertise) would notice it even more.

I'm here to say that thus far I feel it has been working very well. It hasn't caused any issues (other than it took a little while for players to remember to roll 2d10 rather than a d20), advantage/disadvantage has been fine (in both cases roll 3d10 and take either the highest/lowest two of the three dice), and even those abilities/spells that grant bonuses to skill checks have not made any outrageous impact.

And in fact, I think it has made things easier for me to determine good DCs because the odds of reaching certain numbers has made them easier to say "Yeah, this check is difficult". With an 11 being the middle of the bell-curve for a common person with no modifier (and thus my foundation to build off of) my DC chart looks like this:

8 - Simple
11 - Easy
14 - Moderate
17 - Difficult
21 - Hard
24 - Extreme
27 - Nearly Impossible
30 - Superhuman

Knowing that most skill modifiers for most PCs will fall in the -1 to +4 range (basically anything that isn't a high ability + proficiency modifier and/or Expertise)... the bell curve will usually give me rolls in the 9-13 range and thus produce most modified results between like 10 to 17. Thus it's pretty cake to decide in the moment whether something is easy, moderate, or difficult... and the result of who should make or should not make these checks usually bear themselves out. And the times when I decide something is really hard for anyone to make, I just pop the DC to the ones above 20, knowing that its only going to be the ones who either get exceedingly lucky on their die roll, or who have really focused themselves on that particular skill (due to the higher bonus from high ability modifier and proficiency/Expertise.)

When I ask for a perception check, the elven rogue tends to be the one hitting the high numbers pretty consistently, whereas in the old days every single PC could be hitting DC 15s or 20s most of the time. I can now set DCs that play into the strengths of their characters, knowing that the odds are in their favor to be the ones hitting them, not just every PC who happens to have a 1 in 20 chance of getting a nat 20.

This system won't be for everyone obviously, but if anyone has been considering doing something like this and wanted confirmation that it *can* work... I'm here to say that yes, a 2d10 bell-curve for skill checks does not cause any undue problems over the d20 die roll system. It works excellent in my opinion, and has resulted in those PCs who are geared towards certain skills to be perceptionally more successful more of the time at the table than those who aren't. And it has made it more obvious to the players what the holes are in their skill roll-outs. Which I feel has been a boon thus far for our tables.

Glad it works for you! I've thought about alternatives myself. Thanks for sharing your experiences with it!

FYI it isn't a "bell curve" really, but a pyramid. Either way, it gets what you want: which is away from a linear distribution. We've used 4d6-4, really rolling 4d6 but treating 6's as 0's. The sum is 0-20 and it is more of a true bell curve. However, it is a bunch of dice so 2d10 is probably better for your purposes.

Optionally, a d8 and d12 will give you more of a bell curve (really a plateau). d8+d12 removes one occurrence of 10 and 12, and two from a result of 11.

Here is the chart of the three distributions for any who are interested:



EDIT: added Chart.

I use d8+d12, myself.

I like the flat probability from 9-13 and the d12 is the coolest due and doesn't get used enough.

I've found in practice it greatly benefits PCs who are good at stuff as it's really hard to roll low (and high).

I keep my DCs mostly in the 10-15 range with the maximum from 5-20. At low levels a 20 can be very hard to achieve.

Hmm, I was ready to discount this when I read the subject but it's actually quite interesting. Do you do it for attack rolls and saves as well?

I do like the idea of less variance, at least with respect to someone's ability level and their success rate with common tasks. Although that could also be resolved by simply not calling for skill checks as often.

I do wonder if it results in more homogeneous results. It also makes succeeding at higher DCs more difficult.

DEFCON 1 said:

I'm here to say that thus far I feel it has been working very well. It hasn't caused any issues (other than it took a little while for players to remember to roll 2d10 rather than a d20), advantage/disadvantage has been fine (in both cases roll 3d10 and take either the highest/lowest two of the three dice)

Click to expand...

Interesting. Why are you using that method and not rolling your 2d10 twice? And have you tried applying it to other d20 rolls?

Interesting idea. At first I had two issues, but those were resolved by the end. I personally use 1d8+1d12 for wandering monster charts, rather than the more common 2d10, because it provides a flatter "curve." I also felt that this would massively reduce the importance of advantage/disadvantage, since the rolls are going to average out more often. However, using 3d10k2 is an innovative way to keep advantage and disadvantage relevant, since the randomness of each die makes it important.

I've really been disliking the sheer randomness of 1d20 for a while now, to the point where I am enjoying D&D less because the huge variance completely overshadows being trained at something (especially in the early levels). I think I am going to give this a try.

DEFCON 1 said:

I have run many, many games now over the years that have used the standard d20 for skill checks, and over time have grown more and more disillusioned with it. Principally due to the fact that the d20 die roll produced way too large a variance for me, resulting in a PC's personal modifier having much less meaning in the grand scheme of results. A PC with a +6 to a skill versus one with a +0 did not make a perceptional difference that often... when the +0 could roll 16s to 20s while the +6 would roll 3s, 5s, 6s etc. The lack of a bell-curve meant modifiers had less import.

Click to expand...

Simple solution: have the players switch dice. That way the PC with the +6 could roll 16s to 20s while the +0 would roll 3s, 5s, 6s. Amirite?


On a more serious note, I have noodled over this before but concluded that the d20 just works. In the multiple dice method, any combo of dice is going to have results that strongly favor the middle of the range thereby lowering the chances of spectacular successes and failures. High and low rolls alike can create memorable moments at the table and reducing the chances of those would be a net loss, IMO (especially the failures ). As [MENTION=6987520]dnd4vr[/MENTION] demonstrates, rolling a 20 (or a 1) is 5x less likely to happen with 2d10 than it is with a d20.

In game, when there is a meaningful consequence for failing a check and so a roll is called for by the DM, the +0 PC could simply offer to Help (or Work Together with) the +6 PC. If that is appropriate in the given situation, the "problem" of the lesser skilled PC rolling higher goes away - instead they have teamed up to gain advantage and gain a better chance to succeed.

One might also argue that 2d10 for skill checks also somewhat diminishes the impact of the Rogue's Reliable Talent since it's less likely to roll less than a 10... or that it diminishes the value of Bardic Inspiration as the truly skilled PC won't need it as often... but maybe neither of those is really that significant...

That said, if 2d10 works for your table in a fun way, that's cool - and I'm glad you shared it!

I am reminded of Traveller with its roll 2d6.

The OP references Traveller: 2300 AD with 4d6-4.