I was thinking about AC, attacks, and how to make it slightly easier on a DM, and more interactive for players. The idea I came up with was to reverse AC rolls.

To do this, a player's AC now becomes the Armor/Magic Item mods plus DEX modifier, and a monster's attack would be the listed attack bonus plus 10. Essentially this moves the +10 from player to monster. Finally, a player rolls a d20 and adds their new AC to this. From here, a tie or going over is a successful dodge, going under is a successful attack.

I've seen two problems with this, namely that you could see this as being able to have things that affect ability checks affect this, but this would follow the rules set out in the Ability Checks article on WotC website. The article states (talking about attack rolls, saving throws and ability checks) "If something in the game, like the guidance spell, affects one of them, the other two aren't affected unless the rules specifically say so." The other problem I saw was Advantage/Disadvantage on the monsters Attack rolls. This is also easily solved by applying the reverse to the player. For instance if a monster has previously been given advantage against a PC, then in this new system the PC would have disadvantage vs. the monster's attack.

I haven't tried it, but have thought about it for a while. My questions are these: Has anyone tried this? If they have, does it work? If you haven't tried it, can anyone see problems with it?

  • 2
    \$\begingroup\$ This was a pretty common modification in older versions of the game, for what that's worth :) \$\endgroup\$
    – Erik
    May 5, 2015 at 20:02
  • \$\begingroup\$ Also try to set a ruling on how to handle player vs player and monster vs monster, to prevent confusion when that kind of thing comes up. \$\endgroup\$
    – Erik
    May 6, 2015 at 7:56

3 Answers 3


Yes, this was tried in D&D 3.5e as a variant rule.

An officially published version of this rule existed in Unearthed Arcana, a 3rd edition D&D manual of variant rules. It can be used essentially unchanged with the 5th edition.

Unearthed Arcana p.133:

Players Roll All The Dice [...] Attacking and Defending

With this variant, PCs make their attacks just like they do in the standard rules. Their opponents, however, do not. Each time an enemy attacks a PC, the character's player rolls a defense check. If that defense check equals or exceeds the attack score of the enemy, the attack misses.

To determine the creature's attack score, add 11 to the creature's standard attack modifier (the number it would use, as either a bonus or a penalty to its attack roll, if it were attacking in any ordinary situation using the standard rules). For instance, an ogre has standard attack modifier of +8 with its greatclub. That means that its attack score is 19.

To make a defense check, roll 1d20 and add any modifiers that normally apply to your Armor Class (armor, size, deflection, and the like). This is effectively the same as rolling a d20, adding your total AC, and then subtracting 10.

  • Attack Score 11 + enemy's attack bonus
  • Defense Check 1d20 + character's AC modifiers

If a player rolls a natural 1 on a defense check, his character's opponent has scored a threat (just as if it had rolled a natural 20 on its attack roll).

Note that in D&D's core mechanic, a d20 roll that exactly matches its target DC is considered to succeed -- ties go to the person rolling the dice. So letting the defender roll actually gives them an advantage if you convert the numbers naïvely. This is why it's necessary to add 11 rather than 10 to the attack score to preserve the usual odds.

From the rules for Inspiration in the Player's Handbook, p.126:

If you have inspiration, you can expend it when you make an attack roll, saving throw, or ability check.

So you can't use inspiration to give yourself advantage on a defense roll. This matches the usual case, in which you can't use inspiration to give an opponent disadvantage on an attack roll.

  • \$\begingroup\$ What if both Attacker and Defender rolled? Would the odds be too a skewed? \$\endgroup\$
    – L.P.
    Sep 17, 2019 at 20:16
  • \$\begingroup\$ @L.P. That DOES change the odds, making things much "swingier" and giving the game a more capricious character. In recent years game designers have moved away from rolling both attack and defense simply because the game is faster with fewer dice to read, but earlier systems (GURPS, old World of Darkness) often did so. \$\endgroup\$
    – Sebkha
    Sep 18, 2019 at 23:52

This works fine, and has been done by groups since the dawn of RPGs. There are few to no traps awaiting you if you want to reverse the monters' to-hit math so that players make all the rolls.

Your concern about what might affect the roll changing needn't be an issue so long as you keep the situation clear in your head: just defer to the original way attack rolls are made when anyone wants to affect a roll. Think of it not as the players suddenly getting control over the roll, and instead as having your players do your rolling for you... and by the way you're flipping the math around to make that work properly. If you keep in mind how the game already works, and think of this change as just a different way of rolling the same dice, you won't run into any weird situations with players affecting things they couldn't normally.

Your observations about advantage/disadvantage and keeping these things operating as they originally did means that you're already thinking along the right lines to make this work. Go for it! Not having to roll dice frees you up for other DMing duties, to a surprising degree.

  • 1
    \$\begingroup\$ You can also do that with saving throws, though you need to tweak it a bit. I've never rolled for armor, but I've always thought it was rather illogical that some attacks require the attacker to roll while others require the defender to. \$\endgroup\$
    – Dungarth
    May 5, 2015 at 20:38
  • 3
    \$\begingroup\$ The only trap (and only real difference) is that the results go from behind the DM screen to infront of it. If you want to 'cheat' for any reason it becomes obvious to the players. This may or may not be an issue depending on DM style. \$\endgroup\$
    – Scott
    May 6, 2015 at 0:45
  • 2
    \$\begingroup\$ And watch out for criticals. I suppose you could cast them as fumbled defence rolls, instead? \$\endgroup\$ May 6, 2015 at 1:08

I feel like the math is off here somewhere, but I can't explain it, but maybe I can represent it using numbers.

Consider the following situation:

Player Character with an AC of 16, vs a creature with ATK bonus of +5

In order for the character to 'take damage', the attack roll on a d20 would have to be 11 or higher. 1-10 would be a miss, 11-20 would be a hit resulting in damage. The 11 causing a 'tie' but ties successfully hit.

Result: 50% chance of player taking damage.

Now use the variant rules:

Player character of Defense bonus of +6 ('AC of 16'-10), vs a creature with an 'attack check' of 16 ('atk bonus of +5'+11)

In order for character to 'take damage', the defense roll on a d20 would have to be 9 or lower. 1-9 would mean failing to defend, 10-20 would successfully defend. the 10 causing a 'tie', but ties successfully defend.

Result: 45% chance of player taking damage.

There is a 5% difference. When I check the variant rule set via algebra, I get the same result:

Formula for 50% chance of taking damage:

\begin{array}{c|c|l} \text{Attacker} & \text{Defender} & \text{Comment} \\ \hline d20+5 & +16 &\text{minus d20 from both sides} \\ d20-d20+5 & +16-d20 \\ +5 & +16-d20 &\text{Multiply by (-1) so we're adding the d20,}\\&&\text{not subtracting.} \\ -1(5) & -1(16-d20) \\ -5 & -16+d20 &\text{add 20 to both sides (forcing positive bonus}\\&&\text{and difficulty check)} \\ -5+20 & -16+20+d20 \\ 15 & 4+d20 & \text{adding 2 more so difference between}\\ &&\text{original AC and defense bonus is -10} \\ 17 & 6+d20 \end{array}

As shown here, if we change AC by subtracting 10 to create a defense roll bonus, we would have to add '12' to the atk bonus to get an attack difficulty check

  • 1
    \$\begingroup\$ I made an edit to make your working out readable. Please check that I didn't mess anything up, and feel free to ask if you have any questions about the MathJax (that's the code used to make that table) \$\endgroup\$
    – Someone_Evil
    Sep 17, 2019 at 20:10
  • 3
    \$\begingroup\$ Also, Welcome to RPG.SE! Take the tour if you haven't already and see the help center or ask us here in the comments (use @ to ping someone) if you need more guidance. Good Luck and Happy Gaming! \$\endgroup\$
    – Someone_Evil
    Sep 17, 2019 at 20:10
  • \$\begingroup\$ It actually took me a min to notice that someone else had edited it, was trying to figure out how my response changed so drastically xD but Thank You very much for cleaning it up. All the data is correct, so your edit was solid. I'm just curious about the math because these variant rules have been in use it seems since 3.5, but no one has been able to comment on the math being off 5% when I worked it out myself. \$\endgroup\$ Sep 17, 2019 at 20:48

You must log in to answer this question.

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