Confusion about shield spell and lucky feat timing.

[Barny]

In my last online session, we have an argument about lucky feat use timing.

GM describes the situation: "The enemy is making a deadly attack on PC A, and it seems that it's going to hit with 25 total hit roll"

PC A:" I cast Shield spell as a reaction, now my AC +5 to 26."

GM: "The enemy made a perfect strike, and deal a critical hit to A for 20 dmg."

PC A:" Wait! I use the lucky feat to replace the enemy's roll!"

GM:" Too late! I already announced the result. you need to use it before I announced the result"

PC A:" Unfair! I don't even have a chance to use this feature."

GM:" You do but you choose to use shield spell, not lucky feat against a deadly strike"

Is the GM maknig the right call with the rules or being unfair to PC?

If you are the PC, how do you best use the shield spell and lucky feat to avoid a critical hit?

[Bobthewizard]

I think the DM is right, but shouldn't have told you the total.

For shield, the DM just says if the attack hits, they don't have to tell you the total or if it is a critical. I agree with them that once you use shield you can't go back and change the die roll. You would need to use lucky first.

To use the lucky feat for defense. You roll a die and then choose to use that roll or whatever the DM rolled, without knowing what the DM rolled. When the DM rolls, you use it before they roll to give a sort of blind disadvantage. But that decision must be made before the DM tells you the results.

Here's how I think the order goes.

1. DM says they attack you, but not the result
2. You decide to use lucky or not, picking to use your die or theirs without knowing the result of the DM's die.
3. DM tells you if it hits
4. You decide to use shield
5. DM tells you if it still hits

Lucky is not useless. It's just better for your attacks or saves than enemy attacks or saves.

[Yakk]

There is no wording on the defensive use of lucky that it must be before the result is determined, nor that you have to do it blind.

You can invent such wording if you want. I'm sure many DMs have.

[Man_Over_Game]

There is no wording on the defensive use of lucky that it must be before the result is determined, nor that you have to do it blind.

This is true. Only the personal use of Lucky says you can use it after knowing the roll, not the defensive version. For what it's worth, Crawford says whether the player knows the enemy's attack roll is up to the player.

Although I feel it's really dumb for the DM to have ruled in that way, as the player was already going to use two resources to negate that attack, and there wasn't a realistic window for Lucky to be used in that situation. Crit against Lucky or Shield can be pretty dramatic, but both makes it feel very Player vs. DM.

[jmartkdr]

If the attack was a critical hit, saying "It seems like it will hit" or announcing the total (which is irrelevant) are both ... odd... things for the dm to say.

[Yakk]

"The enemy is making a deadly attack on PC A, and it seems that it's going to hit with 25 total hit roll"
Yes, that looks like a DM fishing and using specific wording to burn player reaction defensive abilities.

If your DM doesn't like you using shield and lucky defensively, just make a new character or rebuild it not to use it. Getting into rules arguments about it isn't a winning meta-strategy.

[Contrast]

The rules are silent on how its meant to work exactly but there are a couple of abilities that make a lot more sense if it is assumed that players can see the number rolled on the dice without knowing the modifier being applied.

Grave Clerics anti-crit ability would seem non-functional if the DM doesn't let you know when an attack is a crit for one thing.

[Barny]

This is true. Only the personal use of Lucky says you can use it after knowing the roll, not the defensive version. For what it's worth, Crawford says whether the player knows the enemy's attack roll is up to the player.

Although I feel it's really dumb for the DM to have ruled in that way, as the player was already going to use two resources to negate that attack, and there wasn't a realistic window for Lucky to be used in that situation. Crit against Lucky or Shield can be pretty dramatic, but both makes it feel very Player vs. DM.

Can you give me the link where JC said it's up to the player to know the enemy's attack roll?

[Barny]

If the attack was a critical hit, saying "It seems like it will hit" or announcing the total (which is irrelevant) are both ... odd... things for the dm to say.

I think it's fine for DM to use the wording "It seems that it will hit" on Natural 20 because he's describing an occurring event without a final result.
If the DM says "the attack is a guaranteed hit" or "a perfect strike", then he already announces the part of the result. Then how can he change the wording after PC use lucky at the right moment to change the final result?

JC said DM can hide the roll and tell the total attack roll.
https://www.sageadvice.eu/2018/11/20/does-the-lucky-feat-allows-the-character-to-see-the-number-on-the-enemys-die-attack-roll/

[Lunali]

If the attack was a critical hit, saying "It seems like it will hit" or announcing the total (which is irrelevant) are both ... odd... things for the dm to say.


I think it's fine for DM to use the wording "It seems that it will hit" on Natural 20 because he's describing an occurring event without a final result.
If the DM says "the attack is a guaranteed hit" or "a perfect strike", then he already announces the part of the result. Then how can he change the wording after PC use lucky at the right moment to change the final result?

I agree with half of both of you, if the DM says "It seems that it will hit" on a nat 20, that seems fine since he's giving you the chance to react without telling you whether it will work. If they say "It hits AC 25" on a nat 20, they're giving you inaccurate (because it will hit higher AC than that) information to try to get you to cast shield when it won't work.

[DeadMech]

As written the DM comes off as needlessly antagonistic to me.

[Keravath]

From Lucky:

"Whenever you make an attack roll, an ability check, or a saving throw, you can spend one luck point to roll an additional d20. You can choose to spend one of your luck points after you roll the die, but before the outcome is determined. You choose which of the d2Os is used for the attack roll, ability check, or saving throw.
You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours."

In order to choose which d20 roll is used, you have to know what those die rolls are. This is true whether the player rolls the die or the DM. The DM doesn't just get to say the "attack hits" which is all that is required to trigger the shield spell. When the Lucky feat is in play, the DM has to tell the player what the die roll was (not whether it hits or not) so that the player can decide whether to use the feat.

In this case, the events should have been ... creature attacks, DM reveals that the creature rolled a 20, player decides to use Lucky feat or not ... then based on the two die rolls, player decides which one to use, DM reveals whether the attack hits or not (since the player does not know the + to hit) THEN the player can decide to use shield if whatever the revised die roll still hits.

Avoiding crits is one of the main applications of the Lucky feat.

[kazaryu]

From Lucky:

"Whenever you make an attack roll, an ability check, or a saving throw, you can spend one luck point to roll an additional d20. You can choose to spend one of your luck points after you roll the die, but before the outcome is determined. You choose which of the d2Os is used for the attack roll, ability check, or saving throw.
You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours."

In order to choose which d20 roll is used, you have to know what those die rolls are. This is true whether the player rolls the die or the DM. The DM doesn't just get to say the "attack hits" which is all that is required to trigger the shield spell. When the Lucky feat is in play, the DM has to tell the player what the die roll was (not whether it hits or not) so that the player can decide whether to use the feat.

In this case, the events should have been ... creature attacks, DM reveals that the creature rolled a 20, player decides to use Lucky feat or not ... then based on the two die rolls, player decides which one to use, DM reveals whether the attack hits or not (since the player does not know the + to hit) THEN the player can decide to use shield if whatever the revised die roll still hits.

Avoiding crits is one of the main applications of the Lucky feat.

the part you bolded specifically applies to a D20 rolled by the player, not the DM. the relevant part of the feat is:


'You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours'

which is super ambiguous as to the intent.


for my part i believe the intent is that you decide to use a lucky dice when the DM declares an attack against you (but before you know whether the attack succeeded or failed) however, you still know the results of both dice so that you can make an informed decision on which to use...i.e. the lower one probably. unless you want them to hit? which could happen in niche scenarios.

However, this is, as i indicated, my own opinion. the wording itself is too vague imo.



As for OP: based on what i just stated, i think DM was technically correct. although i also think overall the situation wasn't handled super well, possibly for any number of reasons.

[Chronos]

The sequence is supposed to be:
(for an NPC attacking a player)
DM rolls a d20.
Everyone at the table sees the d20 result, but do not yet know whether it'll be a hit, because they don't know the monster's attack bonus.
Players can at this time use "before the result is known" abilities.
Any such abilities are applied.
Given any such abilities, the DM declares that the attack hit or not (if he doesn't know the PC's AC, he needs to ask).
Based on that, the target might decide to cast Shield.

(for a PC attacking an NPC)
Player rolls a d20.
Everyone at the table sees the d20 result, but do not yet know whether it'll be a hit, because they don't know the monster's AC.
Players (or NPCs) can at this time use "before the result is known" abilities.
Any such abilities are applied.
Given any such abilities, the player states the total.
The DM compares that total to the monster's AC and declares that the attack hit or not.

Now, in real life, groups don't always follow that sequence strictly. Like, here, the DM announced the total of the roll (including attack bonus), which he shouldn't have, because now there is no "before the result is known": The players know their ACs, and so, given a total, they know whether it's a hit. But the fact that the DM broke the sequence means that he has to make allowances for the effects of the broken sequence.

[stoutstien]

If the attack was a critical hit, saying "It seems like it will hit" or announcing the total (which is irrelevant) are both ... odd... things for the dm to say.

Agreed. If the roll is critical then they should probably just announce it so the player can make an actual sound tactical decision about how they want to react to that rather than gambling. I mean the DM knows when a player rolls a critical regardless on how they want to decide how the NPC would react so fair is fair.

[Bobthewizard]

There is no wording on the defensive use of lucky that it must be before the result is determined, nor that you have to do it blind.

You can invent such wording if you want. I'm sure many DMs have.


The sequence is supposed to be:
(for an NPC attacking a player)
DM rolls a d20.
Everyone at the table sees the d20 result, but do not yet know whether it'll be a hit, because they don't know the monster's attack bonus.
Players can at this time use "before the result is known" abilities.
Any such abilities are applied.
Given any such abilities, the DM declares that the attack hit or not (if he doesn't know the PC's AC, he needs to ask).
Based on that, the target might decide to cast Shield.


While I prefer to play with transparent rolls the way you both describe, and use Chronos' order when I am the DM, DM's are allowed to keep their rolls hidden. If they do, then I think my order in the first response above applies. Based on the OP's description, it seems their DM prefers to keep their rolls private, which they are allowed to do, and so that is how I responded to their post.

Lucky states "You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker’s roll or yours." If your DM keeps their rolls private normally, there is nothing in the feat that says you get to see it. The DM is perfectly allowed to say "when an attack roll is made" means you have to decide when the attack is made but before knowing if it hits. Once they announce the results, they can rule that it is no longer "when an attack roll is made" but has now moved on to "when you are hit."

Shield says, "Reaction trigger: You are hit by an attack or targeted by the magic missile spell." Again, nothing in the spell says you get to see the roll.

So while I prefer to play with open rolls by the DM, I don't think they are inventing wording or doing anything wrong by their ruling.

[Yakk]

The DM is allowed to do anything.

By their actions, they are making defensive reaction abilities less useful.

I mean, who says "hits with a 25" then LATER says crit. Basically, that is an example of "baiting shield".

[Bobthewizard]

The DM is allowed to do anything.

By their actions, they are making defensive reaction abilities less useful.

I mean, who says "hits with a 25" then LATER says crit. Basically, that is an example of "baiting shield".

I do agree that saying the total is 25 without saying the roll is a 20 is bad. They should go one way or the other.

[Composer99]

I roll behind the screen (unless I want to roll in the open for dramatic effect), so if one of my players tells me they want to use lucky defensively, personally I would tell them whether it would be worth it or not. If they don't think of it, I might even prompt them. Because my role isn't to play silly buggers with their features (unless it makes sense for an antagonist to do so, which wouldn't apply to Lucky by my reckoning).

Barny, the DM you describe made a hash of that call. It happens, but it seems like they were approaching the situation in too adversarial a fashion

[Nagog]

The DM should announce that it is a critical hit at the start. At that point, the player knows the Shield spell won't save them, as a natural 20 always hits, regardless of AC. Lucky, however, could save them, so they would know which tool to use for this situation. That's the issue with this one. With proper and clear communication, this is how a similar situation would go:

DM: Does a 17 hit your AC?

Player: It would, but I cast Shield as a reaction, so no.

The DM can choose to keep attacking this PC in hopes of hitting above the new AC, but the PC has chosen to expend a resource to buff themselves for the moment, an expenditure that should not be tossed out at the DM's whim.

Same principle applies to Luck. The DM rolls the dice, asks if the total hits the target's AC. At this point, the player can choose to expend a resource (their Luck die) to roll their own d20 and choose to use it or the original roll. The player does not know the creature's bonus to their attack roll, and in some cases may mistakenly choose the higher die roll as a result.

From the sounds of this example, the DM has taken an adversarial attitude towards the players and is wording the combat in such a way to give the enemies a subtle advantage over the players. Such an attitude and application is an abuse of the DM's power, and ignorant of the role of the DM to provide the players with a fun and enjoyable experience.

[Barny]

I think I find the answer from 5e official designer.

https://twitter.com/JeremyECrawford/status/791055742648197120
According to JC, it's perfectly fine for dm to hide the roll, and the dm did nothing wrong.


Lucky feat says: "You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours."
So PC can choose which rolls (dm's roll or PC's lucky roll) the attacker uses, but it didn't say PC must know dm's roll.

[Yakk]

I think I find the answer from 5e official designer.

https://twitter.com/JeremyECrawford/status/791055742648197120
According to JC, it's perfectly fine for dm to hide the roll, and the dm did nothing wrong.

No, that is far from saying what you wrote.

He said that players are free to state they'll use lucky before the DM rolls. He didn't say that DMs should ever refuse to show a roll before a player can use lucky, which is what you wanted to know.

[Keravath]

I think I find the answer from 5e official designer.

https://twitter.com/JeremyECrawford/status/791055742648197120
According to JC, it's perfectly fine for dm to hide the roll, and the dm did nothing wrong.


Lucky feat says: "You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours."
So PC can choose which rolls (dm's roll or PC's lucky roll) the attacker uses, but it didn't say PC must know dm's roll.

I'd also add that in the quote you cited, BOTH examples, whether the player rolls lucky at the same time as the DM or whether they roll it after the DM, in both cases the player sees the die roll of the DM and chooses which to use.

It is impossible to reasonably choose which die to use if you can't see both rolls. DM rolls ?, player rolls a 13. Which one do you want to use? In my interpretation of the feat as written, that is not how it is intended to work. The player knows the two die rolls and decides which one will be used.

Can a DM run it differently? Sure, it is their game. The feat doesn't explicitly state that you see the DMs die roll but the text strongly implies that in my opinion since it applies exactly the same mechanics and wording as used when the player rolls the die and wants to employ the Lucky feat.

"LUCKY
You have inexplicable luck that seems to kick in at just the right moment.
You have 3 luck points. Whenever you make an attack roll, an ability check, or a saving throw, you can spend one luck point to roll an additional d20. You can choose to spend one of your luck points after you roll the die, but before the outcome is determined. You choose which of the d2Os is used for the attack roll, ability check, or saving throw.
You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours. If more than one creature spends a luck point to influence the outcome of a roll, the points cancel each other out; no additional dice are rolled. You regain your expended luck points when you finish a long rest."

The feat gives you the option to spend the luck point after the roll is made, you can choose which die to use. There isn't much choice involved if you don't know both of the die rolls. In fact it is pretty pointless to utilize a luck point in that circumstance since the odds are (without any input on what the other die roll might be) that you will make things worse as often as better which is not the point of the feat.

Up to the DM how to run it but not revealing the die roll to the player considering the use of a luck point is a clear nerf to the feat.

[Torpin]

IMO This is poor DMing, the DM shouldve said he crits you, at which point you knew you could use lucky and shield wouldnt have mattered

[Tanarii]

It's fine for a DM to roll behind a screen, if they feel the need to fudge/cheat. Personally I won't ever play with them, but many players are okay with this.

Just so long as they tell the players what they "rolled" so the players can use any of the many PC features that require deciding after the roll but before the results are applied.

[Zhorn]

It's fine for a DM to roll behind a screen, if they feel the need to fudge/cheat. Personally I won't ever play with them, but many players are okay with this.
I roll behind a screen, but I also tend to let players peak over it for those nat 20 moments, or when those fist-full-of-dice moments ended up with a statistically improbably number of high or low values.

ME: "That Thayan Wizard's fireball burns you for" *roll* "... 11 damage"
PC: "But we all failed the save?" *peak* "huh... 11 damage"

[Keltest]

Personally, i dont generally announce the actual number rolled unless its a crit hit or miss (because those are different than regular hits and misses, and they therefore need to know). But as soon as lucky comes into play, they need to know the number rolled to pick which they want. So there is no "too late" until the table has moved beyond that turn, at which point, well, they had their chance.

[Barny]

I'd also add that in the quote you cited, BOTH examples, whether the player rolls lucky at the same time as the DM or whether they roll it after the DM, in both cases the player sees the die roll of the DM and chooses which to use.

It is impossible to reasonably choose which die to use if you can't see both rolls. DM rolls ?, player rolls a 13. Which one do you want to use? In my interpretation of the feat as written, that is not how it is intended to work. The player knows the two die rolls and decides which one will be used.

.....

Up to the DM how to run it but not revealing the die roll to the player considering the use of a luck point is a clear nerf to the feat.


Well, even in fatasy world, PC take a feat "lucky", which can provide more options to choose, but definitely not having god-knowing info including dm's rolls.

Imo, Lucky feat have 2 part of features:
1. You can use it to replace many d20s of yours, which is risk-free decision within your character's own knowledge and power.
2. You can use it to replace enemy's rolls, but it involes some unknown info beyond your in-game character's knowledge and power. So it's a decision with certain risk.

It's very similar to a powerful divination wizard class feature Portent, with which you have 2 certain d20s to replace monster's attack rolls, or possible crit rolls.
However, you need to decide to use this ability Portent before dm's rolling, so you can use a certain d20 to replace dm's roll, but would never know what dm's roll is.

[Aimeryan]

The enemy attack roll is not constrained to before knowing the outcome:


You have inexplicable luck that seems to kick in at just the right moment.

You have 3 luck points. Whenever you make an attack roll, an ability check, or a saving throw, you can spend one luck point to roll an additional d20. You can choose to spend one of your luck points after you roll the die, but before the outcome is determined. You choose which of the d20s is used for the attack roll, ability check, or saving throw.

You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours. If more than one creature spends a luck point to influence the outcome of a roll, the points cancel each other out; no additional dice are rolled.

You regain your expended luck points when you finish a long rest.


The paragraph with spending a luck point for when you roll the die requires you do so before the outcome is determined, while the paragraph with spending a luck point for when the enemy rolls the die makes no such requirement.

In other words, resolution can go as thus:

- Enemy rolls
- DM: "Looks like its going to hit."
- Player: "I use Shield!"
- DM: "Sorry, still hits, in fact turns out to be a crit."
- Player: "I use Lucky!"
- Player rolls
- Player: "I choose my die."
- DM: "Luckily, the apparent crit was actually deflected after all."

[greenstone]

I think the takeaway from the Lucky feat discussion is:

As your GM how they rule this before deciding to take the feat.


As far as shield is concerned, the trigger for the spell is being hit, which happens after the attack roll(s) is resolved, so it certainly happens after the Lucky feat.

[MaxWilson]

From Lucky:

"Whenever you make an attack roll, an ability check, or a saving throw, you can spend one luck point to roll an additional d20. You can choose to spend one of your luck points after you roll the die, but before the outcome is determined. You choose which of the d2Os is used for the attack roll, ability check, or saving throw.
You can also spend one luck point when an attack roll is made against you. Roll a d20, and then choose whether the attack uses the attacker's roll or yours."

In order to choose which d20 roll is used, you have to know what those die rolls are. This is true whether the player rolls the die or the DM.

Not necessarily. The player could just declare a policy like "reroll if it's 16-20" and the DM could execute it, then declare hit or miss, wait for a Shield declaration, then tell you if you were still hit or critted.

[Chronos]

While the Lucky feat used on an enemy attack roll doesn't say anything about when it's used (which lack of saying anything could be interpreted for or against the player), there are other abilities that a player might want to use in such a situation (such as a bard's Cutting Words or Combat Inspiration) that do specify "After the roll but before the results are known". So that step still needs to exist in the process.