I'm playing a 5e campaign as a Fighter, planning on becoming an Eldritch Knight.

Thus far (we're still level 1), the DM has been resolving attack rolls against us by telling us the result of the attack roll, and asking us if it hits:

  • DM: "The goblin takes a swing at you!" (rolls behind DM screen) "Does a 19 hit you?"
  • Player: (looks at character sheet) "Yep. Ouch!"

This has worked fine so far, since there isn't really anything any of our characters can do about being attacked, and you can't derive very much useful information from knowing what a monster's modified attack roll was. However, in a couple levels, I'm going to learn to cast spells, specifically the shield spell, which changes things a bit.

I know that I don't have to declare whether or not I cast shield until I know whether the attack would hit me, but am I supposed to know the exact modified attack roll result when I decide? Or should the sequence of events be more like:

  • DM: "The goblin takes a swing at you!" (rolls behind DM screen) "What's your AC?"
  • Player: (looks at character sheet) "It's 16."
  • DM: "Okay, the attack hits."
  • Player: "I cast shield in an attempt to protect myself, raising my AC to 21. Does it still hit me?"
  • DM: "Nope, it glances off your magical barrier."

This is important because if I know the attack roll result, I know whether it's worth bothering to cast shield, whereas if it's hidden from me, I have to guess whether the attack beat my AC by 5 or more.

  • \$\begingroup\$ Related, possible duplicate: Do you know the value of an attack or damage roll on you before the hit is resolved? \$\endgroup\$
    – V2Blast
    Commented Oct 19, 2018 at 4:26
  • \$\begingroup\$ I would suggest that this should work consistently with players attacking an enemy spell-caster. If the DM keeps ACs hidden, then rolls are announced (or players can't tell if they have hit monsters); if the DM makes ACs public, then rolls can be hidden before shield is used. \$\endgroup\$
    – Dan W
    Commented Jan 5, 2022 at 11:05

5 Answers 5


According to this tweet by Mike Mearls, you're supposed to know with certainty whether the shield is enough to protect you or not, so that you can't waste the spell, unless you choose to:

Querent: Am I right..Wiz should know roll before using Shield,cant 'waste it' from not knowing roll?

Mike Mearls: correct

This is also how I have seen it be done in every game I have played. The DM announces the attack score, and the player compares it to their AC and announces if it hits or not.


Whether it hits

Usually, the DM asks your AC before the play, and they take a note behind the screen. When they roll an attack behind the screen, they can immediately determine whether it hits or not. In this style, the DM only announces the hit, and not the total roll.


your group has been using the other style: announcing the total attack roll.

It works with your group. Don't fix if it does not break.

You don't have to worry too much about "metagaming" the attack roll. What your player knows should be instinctively known to your character. If I'm the DM, if the roll can't be beat by Shield, I'll narrate the attack as a very fast attack, and you can't react fast enough to cast Shield, so you don't have to wonder whether to use shield or not. (Of course I tell them beforehand that was my intention: to speed up the encounter)


discuss your plan to take Eldritch Knight and shield with your DM. Ask how they want to manage the usage of shield. Mention the different style (announcing only the hit) and whether they prefer that one over the current style. Let them choose, and stick with that one.

Again, don't worry about metagaming. If your DM don't worry about it, most likely you are worrying over a problem that doesn't exist.

  • 4
    \$\begingroup\$ Unless this is very established at your table, narrating the attack as too fast to react to does not immediately make me think that the roll was too high. Mechanically, I am entitled to a reaction if I have not used one this turn, and without further elaboration this would only make me think you're bending the rules in a way to hurt my character. \$\endgroup\$ Commented Oct 19, 2018 at 17:53
  • \$\begingroup\$ @Winterborne I agree. I've added the relevant part. The only scenario I think this matters is when an abjuration wizard wants to summon their ward, however. \$\endgroup\$
    – Vylix
    Commented Oct 21, 2018 at 5:22

There quite a few answers already covering why a DM might chose to not announce attack roll results, and how this is incompatible with having the Shield spell not be "wasted" sometimes. I'd like to show two practical approaches to solving this dilemma:

Announcing an "accuracy margin" along with the hit

Eg. instead of just saying "yup, the attack hits" I'd say "the attack narrowly hits" or "narrowly misses". I use the following table:

  • Hit/miss by less than 5: "narrowly" hits or misses;
  • Hit/miss between 5 and 10: no special wording - it just "hits" or "misses";
  • Hit/miss by 10 or more: "soundly" hits or "wildly" misses.

This is the best of both worlds, I think. Besides allowing smarter uses of reaction abilities - you won't waste a Shield on an attack that soundly hit you - I found that this information actually helps player strategy and improves combat narration and realism. Players that notice that an strong but unskilled enemy does a lot of damage but often "barely" hits may use a full-defense action or other means to increase their AC; and they would be disinclined to do so if an enemy hits them "soundly" all the time, opting instead to just move away or disable such enemy.

Just have a Reflex roll

If a player, after being hit, declares an intention to use a reaction ability that may affect whatever the ability actually was a hit, I ask for a Reflex roll. If the roll succeeds, I tell the player if usage of such reaction ability will prevent the hit. If the roll fails, the player remains unsure. Regardless of the result, the player is not forced to use or not use the mentioned ability.

The roll has between Routine and High difficulty depending on the accuracy margins in question. For example, I'd say that knowing that an attack hit by a margin of 10 or more is a Routine task, while quickly discerning that an attack barely hit is a Hard task.

Those approaches have worked for me and my players before; and I specially like the first one. You may want to suggest that to your DM but - as usual - it'll be his call and if he's not very worried about your Shield spell, you probably shouldn't be either =).

  • 5
    \$\begingroup\$ Welcome to RPG.SE! Take the tour if you haven't already. Your terminology ("full-defense action", "Reflex roll") seems like it's based on a previous edition of D&D. This question is about D&D 5e. You may want to edit your answer correspondingly to make sure it fits the D&D 5e rules (and the fact that due to bounded accuracy, the margin between a hit and a miss will likely be much lower, since bonuses in general are smaller). \$\endgroup\$
    – V2Blast
    Commented Oct 20, 2018 at 21:48

If the OP is questioning whether the DM should reveal the roll, then the answer is, mostly no. Explained in the latter pages of the DMG, hidden rolls basically allow suspense and uncertainty. If you don’t trust your DM, then you need to have a talk with them. (See the "Dice Rolling" section on DMG pg. 235.)

The shield spell description states:

Until the start of your next turn, you have a +5 bonus to AC, including against the triggering attack, and you take no damage from magic missile.

As the spell states including the triggering attack, meaning when you confirm the attack as a hit, you then can choose to cast Shield as a reaction. RAW

You should know if the attack hits or not, any statistic ventures into the realm of metagaming. Now, a certain level of metagaming will happen, but is not a big issue if you leave that for OoC (Out of Character) times.

Take a moment to actually put yourself in the situation, a round is 6 seconds, as soon as you feel that weapon strike (that nanosecond the weapon would make physical contact, not just guessing it would) against your body, you mutter some words and wave your hand. Then POOF, you acted just in the nick of time to bring forth a barrier in hopes to block said strike.

If you know if the attack roll was too high enough to use the spell, then the spell is pointless. You are taking a chance to prevent yourself from injury or worse.

  • 5
    \$\begingroup\$ While it's true that you know whether you'd normally get hit or not because it's a reaction "which you take when you are hit by an attack or targeted by the magic missile spell"... That doesn't answer OP's question about whether the player knows the actual attack roll, so that they can choose not to cast shield if the attack would still hit even with the +5 to AC. \$\endgroup\$
    – V2Blast
    Commented Oct 19, 2018 at 4:46

According to Jeremy Crawford, the player doesn't see the dice roll:

The shield spell has two different triggers: being hit by an attack or being targeted by the magic missile spell. Neither trigger requires you to see a die roll. #DnD



You must log in to answer this question.

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