I'm thinking about implementing the following houserules at my table to try to bring back the concept of "Critical Successes" by piggybacking off the Advantage/Disadvantage system.

Critical Success

Whenever a creature makes a d20 roll with advantage, if both dice in the advantage roll result in a natural 20, then something especially good will happen as a consequence, determined at the DM's discretion. For an attack roll, this might mean increased damage on the critical hit, or for a Saving Throw, it might mean avoiding all damage from a spell that normally only reduces to half.

Conversely, when a creature makes a d20 roll with disadvantage, if both dice in the disadvantage roll result in a natural 1, then something especially bad will happen as a consequence, determined at the DM's discretion. For an attack roll, this might mean dropping their weapon or causing a self-inflicted injury. For a Saving Throw, it might mean suffering an additional effect contextual to the spell/source.

Although unlikely, if a "Critical Success" occurs in a scenario where success is normally impossible (like a creature with +3 to Charisma making a check with Advantage against a DC of 24), a "Critical Success" should not result in the check succeeding; but it should confer some kind of nominal benefit that reduces the penalty of failure or offers a different path forwards. Conversely, "Critical Failures" should not cause a check to fail if the total still matches the DC (like a +9 check with Disadvantage against a DC of 10) but should confer some nominal penalty to the result.

There's a few things I'm trying to do with this rule:

  • I like the idea of "Critical Successes"/"Critical Failures" as a ludonarrative mechanism for especially unusual outcomes to gameplay situations,
  • But I don't like the high frequency of these outcomes when using traditional "Critical Success on 20, Critical Failure on 1" rules
  • I also like the idea of Advantage giving a small chance of an especially good outcome occurring, with a reflective chance for Disadvantage for something especially bad occurring.

Has anyone attempted to use a rule like this in their games? If so, have players enjoyed this adjustment to the d20 rules? I worry that as-written the rule might be too insignificant to really affect gameplay, is that a potential issue? Should I flesh out the potential outcomes that can occur when following this rule?

  • 2
    \$\begingroup\$ @BenjaminOlson As I tried to show with the examples, it would apply to any d20 check that has Advantage or Disadvantage. \$\endgroup\$
    – Xirema
    Apr 10, 2019 at 18:43
  • 2
    \$\begingroup\$ @BenjaminOlson See the examples. I specifically called out a Critical hit on an attack roll as one scenario where it might apply. \$\endgroup\$
    – Xirema
    Apr 10, 2019 at 18:48
  • 2
    \$\begingroup\$ Can you clarify whether this replaces or is in addition to the "normal" critical rule (i.e. nat 20 and 1s on attack rolls)? \$\endgroup\$
    – Someone_Evil
    Apr 10, 2019 at 20:15
  • 4
    \$\begingroup\$ Why are you rolling checks in cases where success is normally impossible? \$\endgroup\$
    – Mark Wells
    Apr 10, 2019 at 20:19
  • 2
    \$\begingroup\$ "Although unlikely, if a "Critical Success" occurs in a scenario where success is normally impossible " - If it is impossible to succeed, why are you rolling? Also, it's worth noting that outside of attack rolls a natural 1 is just a 1, I don't believe there is such a thing as a critical failure for anything other than attacks. Which also raises the question of why a character should have to roll if even their lowest roll succeeds. A player should only roll if success is possible, failure is possible, and difference between the two matters. \$\endgroup\$ Apr 11, 2019 at 5:01

2 Answers 2


It's probably fair, but also spectacularly unlikely to ever matter.

I've played dozens of sessions of 5e in the past year, and this has literally come up never.

Plus, if it ever does come up that someone rolls a double 20 or double 1, I'll probably just narrate awesomeness anyway, house-rule or no. You're probably just wasting your time preemptively coming up with things that might happen when this really unlikely thing occurs.

  • 3
    \$\begingroup\$ I disagree with the assertion that it's a waste of time. I do agree that it's ill-advised to expect this to come up frequently enough to warrant depending on it or expecting it. \$\endgroup\$
    – GcL
    Apr 10, 2019 at 19:38
  • \$\begingroup\$ I think this is really up to chance - In the past 40 sessions this has come up at least a half dozen times in my campaign. \$\endgroup\$
    – Joshu's Mu
    Apr 10, 2019 at 20:47
  • \$\begingroup\$ Mathematically, it's a 0.25% chance (so one in 400) that rolling two d20's will result in the exact same number, including 1's and 20's. Beyond that, the roller would also have the Advantage or Disadvantage, as appropriate to that roll. It could be estimated that that would halve the chance, but I don't think that's an appropriate approximation. Instead I'd look at it as 1 in 400 Advantage rolls will result in a double 20 and 1 in 400 Disadvantage rolls would result in snake eyes. It's hard to say if this is statistically insignificant for a given table. \$\endgroup\$
    – Ifusaso
    Apr 10, 2019 at 20:53
  • 4
    \$\begingroup\$ It largely depends on how often players and the GM roll for things, and how liberal the GM is with Dis/Advantage (in cases where it's not clear-cut). I've seen games where there are 400 rolls, and ones that there are 10. \$\endgroup\$
    – Ifusaso
    Apr 10, 2019 at 20:55
  • 1
    \$\begingroup\$ @Ifusaso - "Mathematically, it's a 0.25% chance (so one in 400) that rolling two d20's will result in the exact same number, including 1's and 20's." This is incorrect. The chance of double 1s is 0.25%, and the chance of double 20s is also (a separate) 0.25%. The chance of rolling doubles at all is 5% (1 in 20). Proof: Roll the two dice sequentially. Whatever number the first die rolls, there is a 1 in 20 chance that the second die will roll the same number, resulting in doubles. \$\endgroup\$ Apr 12, 2019 at 9:31

My group uses Critical Hits Revisited for improved critical effects. It sounds like it's very similar to what you're asking about (thought limited to combat, not ability checks/saves) because, once the player rolls a 20, they roll a second d20 to determine the nature of the critical. It would be straightforward to do the opposite with failures.

We really like the way it allows for a "severity curve" and isn't just a simple procedure of "you do double damage." For example, if a player crits with slashing damage, the result can be adding a bleed effect (in addition to whatever additional damge is one). But if a player/creature rolls particularly well, the target might accrue various wounds that have both narrative and mechanical ramifications.

In the case that a player rolls a pair of 20s with a slashing weapon, the result is an outright, hands (heads?) down beheading. This actually occurred with a one-shot my group ran a couple weeks ago. A brand new player ended up decapitating Rivalen Tanthul in an encounter.

One thing that makes it really enjoyable is that it makes a sort of mini-game out of the crit/failure. Once the first die is rolled, you know something good or bad is going to happen but you don't know how good or bad or the nature of the effect.

Then, you have the narrative implications. Things like lasting scars, limps, eyepatches for missing eyes, to say nothing of the sanity effects. We've found that a system like this really adds a lot to the game.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .