Thursday 21 August 2014

D&D 5: Roll-Under Skill Checks

So, D&D 5. I have a strange fascination with it. The idea of playing the latest thing which everyone's talking about does actually seem quite appealing. And there are a lot of things I really like about the new edition (backgrounds being one of the main ones). There are, of course, also a few things I'm either suspicious of (will need to be tested out in play) or distinctly not keen on. As ever, though, house rules can come to the rescue. This post is about one aspect of the new D&D that is distasteful to me and some ideas for making it more palatable.


I don't have anything much against them in principle, especially in a system (like D&D 5) where all characters can attempt to use all skills. That's cool. I also really like the simplicity of the proficiency bonus and the lack of "skill points" or whatever. What gets me, though, is the d20 system. I've ranted about this before on various occasions here and on google+.

My basic problem with it is that it introduces a very large random factor into things which, to me, don't seem that random. I'm totally cool with there being a large random element in combat (d20 + modifier vs AC) and with saving throws (d20 + modifier vs DC), but for skills it seems that the random element (d20) is way too significant. For example, the difference between the strongest person in the world (STR 20, +5 modifier) and an average person (STR 11, 0 modifier) is equivalent to only 25% of the random factor. I've always (since D&D 3) thought this seems weird, and WotC seem to disagree. Looks like it's time for a house rule then.

When I think about ability checks, what I really like is the semi-canonical, trad, "roll under your ability score on d20" system. Nice and simple. Rolling against a number on the character sheet. (Yeah, rolling low is better, how inconsistent... but whatever.) A couple of options:

Super Simple Roll-Under Skills
To make an ability check, simply roll d20 and compare it to the relevant ability score (as determined by the DM). Equal or under = success. A natural 1 always succeeds and a natural 20 always fails.

When making a check for a skill you're proficient with, add your proficiency bonus to your ability score.

Slightly More Complicated Roll-Under Skills
Sometimes you might want to give some kind of difficulty rating to an ability or skill check. You know... some chasms are wider than others, some walls slipperier than others, some locks more tricksy than others, etc. D&D 5 play material will no doubt be full of talk of "a DC 25 INT (History) check" and what-not. So it'd be nice to be able to use those difficulty ratings with a roll-under system. Here goes, DC to modifier:
  • Very easy (DC 5): +2
  • Easy (DC 10): +0
  • Medium (DC 15): -1
  • Hard (DC 20): -2
  • Very hard (DC 25): -4
  • Nearly impossible (DC 30): -8
That seems about in line with the kind of modifiers I'm used to in old-school D&D.

I know that this completely changes the probabilities of success and failure. My argument is this: who cares? (Well, I gather that some people do indeed care about such mathematical aspects of the game, but I and no one I play with fall into that category.)

Roll-Under Saving Throws?
This approach could, of course, be applied directly to saving throws too, for the full old-school "roll against the number on your sheet" approach. Personally, I'm happy with luck (i.e. the d20 roll) being more of a deciding factor than raw ability when it comes to saving throws, so I probably wouldn't go down this route. Something to consider though.

Opposed Checks
They don't come up that often, in my experience, but I suppose I should also come up with a less random solution for opposed checks. That's easy: ability score + d6 (+ proficiency bonus, of applicable). Highest score wins.


Any thoughts?


  1. Interesting. I use the terms "roll-under" and "roll-over" for various checks in my B/X inspired game.

    1. Yeah, with old-school games it's an essential distinction. Saves and ability checks (and thief skills) = roll-under, attacks = roll-over. WotC D&Ds are totally phobic of anything roll-under, but I feel it makes more sense sometimes.

  2. I actually talked about this same thing recently on Cooking with Charles:

    1. Interesting post! I'd wondered about a middle complexity system too, only using dis/advantage, without modifiers.

      That's also a very good point about monsters with superhuman abilities kind of breaking this. I'm not sure how often monster skill checks come up in practice...

    2. I talked about using roll-under for combat as well as skill checks, in order to still have one unified system. That's why I had to worry about monsters.

  3. Perhaps you should re-introduce THAC0 while you're at it. I mean, nothing like taking a straightforward, easy to understand system and making it more complex, right?

    If you don't like the level of randomness in skill rolls, lower the difficulty numbers.

    1. This comment has been removed by the author.

    2. Well, that would just make all checks easier, it wouldn't address the basic problem I have -- that purely using the ability *modifier* there's not enough (for my tastes) difference between an average person and a "best in the world".

      I hardly think "roll a d20 under a number written on your character sheet" is complicated.

  4. I like the simplicity of the [i]idea[/i] of a basic roll-under-ability skill system, but it always fees as though the chances of success are then too high. Opposite extreme to your issues. I mean, in your prime req, you likely have a 15+ even in old school editions. Should you start at 1st level with a 75% change of success? Well 85 really, if you still use the proficiency bonus...

    1. Yeah this is the trade off. It's a matter of taste, I guess, as to how skilful one expects characters to be, and how important a role skill checks play in the game. To me, they're not that central, so the idea of 1st level characters being very proficient doesn't bother me.

  5. I'm picking up what you're putting down.

  6. Don't use a D20, use , for example, 3 x D6 and you can get a much more gradual / nuanced range of results.......


Note: only a member of this blog may post a comment.