The highest AC in my games has been 22. I never give out +3 armor or shield, maybe +1 at most, so I'm not worried about high AC. Plus, I would give the ancient dragon a magic penetrating attack if the PCs were trying to make PCs this way. This goes back to other threads about BBEG and hitting and such.
The +5 reaction until your next turn is good. The bad guys that use it are frustrating to the PCs. I also have given it out as part of a magic item. Maybe I give out +1 chain shirt that allows casting shield 1/day. It comes in handy for the PCs and is only once per day.
I find making the +5 bump by 1 for each higher level slot feels cheap and not worth it. A 6th level spell slot for +10 to AC is powerful, but does not feel like a 6th level spell. Maybe +5 to all your allies in 30ft is better, but even that feels like a 4th level spell.
Your problem is this. Not the shield spell. The magic items are what is breaking bounded accuracy. The game isn't made for characters to be running around with 27 AC. Unless you're playing a high magic game and everyone has super high AC. Then the DM can adjust by increasing the attack bonus to monsters.
So, as mentioned above, this is an example of extreme cases.
The thing is, 27 AC doesn't break things at that scale of game. You are hard to hit, but not impossible. You are just wearing very rare armor and using attunement slots for less rare AC boosts.
27 AC plus shield brings things to the brink. Doing further AC optimization plus shield breaks things.
What more, more modest AC boosting ... plus shield breaks things at earlier points in the game. A level 1 character in plate+shield has 20 AC, high; add in the shield spell, and it breaks bounded accuracy at level 1.
But at level 1 this is less of a problem, because spell slots are rare.
By mid levels, you can pull off full tank AC, optimize even a little bit, and add shield, and it breaks bounded accuracy. And by that point spell slots (low level ones at least) are a dime a dozen.
And the DM adjust this by increasing the attack bonus of monsters was a problem I mentioned above and directly addressed, did I not?
I mean, did I not address that? Was it real life? Or just a fantasy? Caught in a landslide, No escape from reality.
When the DM responds to AC optimization by scaling ATK bonuses, this is a problem. You get, as I talked about above, a race between PCs to scale AC. Gaining AC for the tank simply makes everyone else easier to hit, instead of making the tank harder to hit.
I don't want that to happen. Because I consider that to be dumb, leads to bad in-play experience, etc etc.
Or is this just a hypothetical problem which could come up? Trying to solve a hypothetical problem that may or may not even be a real issue by completely reworking the Shield spell doesn't make much sense. Don't try to fix something if it isn't actually broken. Now if you have this problem in your game right now, I guess your solution would work. It's too late to not give them multiple +3 armors.
At almost every modest or higher level of optimization, a tank-like character using the shield spell breaks bounded accuracy.
While the shield spell is expensive, this isn't much of a problem. You get a miss. But if you get a tank-like character with lots of spell slots, you end up with basically unhittable character even when focus fired.
An example does not define the problem. An example illustrates the problem. There are many, many ways to get AC that is high for the level, and when you do such a build plus have plenty of spell slots plus learn the shield spell, you break bounded accuracy.
So I want the shield spell to be not as good in that case. I want it to be about as good for a traditional cloth character, especially at low levels. I am willing for it to be significantly better if a cloth character uses a higher level slot for it, and want it to be better than the 1st level slot. I am willing for it to be slightly better if a tank character uses a higher level slot for it.
Hence the design; limit the swing count of shield by the slot level, and make it reduce attack damage that gets through it.
This makes using 1st level slots for tanks less useful, as they can get focused; so they nullify or blunt a blow or 2, but not every attack.
If the tank-type burns a higher level slot to become nearly unhittable for a round, I am ok with that; higher level slots are much harder to get and generally rarer.
For a cloth caster who gets the occasional blow, 1st level slots are almost as good. If they get focused, they'll want to use a higher level slot. The higher level slot gives better ROI than the original shield spell, because on a cloth caster shield doesn't make you super hard to hit, just harder to hit. Half damage on the blows that do get through can be more valuable than the 5 points of AC.
If a wizard isn't using their 1st level slots for Shield, what are they going to use them for? You need some spells that still have uses late game so that casters still have uses for their low level slots.
I have played casters without shield. There are many uses for 1st level slots. Not all of them are as amazingly good as absorb elements and shield, but they don't need to be. They are 1st level slots.
Shall I provide a list?
Charm Person
Detect Magic
Feather Fall
Comprehend Languages
Find Familiar
Floating Disk
Expeditious Retreat
Disguise Self
Protection from Evil and Good
Unseen Servant
Alarm
Mage Armor
Identify
Longstrider
Snare
Catapult
All of the above could be useful to cast as a 20 level wizard using 1st level slots. Sometimes it just shaves 10 minutes of a ritual cast, sometimes it replaces gear, sometimes it is an emergency spell you might want to drop.
Second, if you examine my changes, this doesn't make the shield spell a bunch worse for traditional cloth casters. Because your miss chance isn't approaching 100%, the shield plus half damage on attacks that hit soaks as much or more damage than the shield spell does quite often.
The highest AC in my games has been 22. I never give out +3 armor or shield, maybe +1 at most, so I'm not worried about high AC. Plus, I would give the ancient dragon a magic penetrating attack if the PCs were trying to make PCs this way. This goes back to other threads about BBEG and hitting and such.
The +5 reaction until your next turn is good. The bad guys that use it are frustrating to the PCs. I also have given it out as part of a magic item. Maybe I give out +1 chain shirt that allows casting shield 1/day. It comes in handy for the PCs and is only once per day.
I find making the +5 bump by 1 for each higher level slot feels cheap and not worth it. A 6th level spell slot for +10 to AC is powerful, but does not feel like a 6th level spell. Maybe +5 to all your allies in 30ft is better, but even that feels like a 4th level spell.
I did not propose +5 AC plus 1 for each higher level slot.
I proposed:
+5 AC against triggering attack.
It works against at most 1 more attack per spell slot level, still granting +5 AC.
If the attack hits, you get resistance.
A 6th level slot gives you a shield that does +5 AC and resistance to hits for 7 attacks total, or until the start of your next turn, whichever comes first.
I want players to be able to say "I want to be hard to hit". I am not interested in nullifying investment in being hard to hit.
I am just aiming at the one piece that brings things from "hard to hit" to "basically impossible to hit" in my experience. And I don't want to make it not work, I want to make it cost more to use it to survive focus fire. I love the idea of using magic to bounce a flurry of attacks from a dragon, but it shouldn't be magic that is barely above cantrip grade.
But your problem is still players maxing out AC to break the game, not the shield spell itself.
Without magic items, the highest AC you can get is 20 (plate + a shield). The Shield spell boosts that to 25. And that is limited by the number of spell slots, so only available for a set number of rounds per day. And to get access to it and increase spell slots, you are not improving in your other class.
I have never seen (or heard anyone else talk about) the extreme AC optimization that you apparently have. But that problem still comes from your game/players. If you are seeing this often, your game has too many AC increasing items. I don't enjoy playing with the type players who want to optimize to that extreme (but to each their own, there is no badwrongfun). If your players are okay giving everything to one character to make them impossible to hit, then that just hurts the other characters.
I mean, artificers, forge clerics, defensive fighting style, bladesinger wizards (AC 23 naked), shield of faith, haste -- off the top of my head. That is not at all a full list.
Also relatively irrelevant. I want magic items in the game. When I do the math on making a tank without magic items, bounded accuracy works. But when I add in the shield spell on top of it, bounded accuracy doesn't work. A tank working hard on AC having 5 more AC than a non-tank is good; a tank having 10 more AC than someone not working hard on AC breaks the game.
The shield spell on any kind of spellcasting tank makes a 5ish point AC gap into a 10ish point AC gap.
I have never seen (or heard anyone else talk about) the extreme AC optimization that you apparently have.
I have seen it. I have seen people talk about it. I do understand if you are not experiencing this problem, then this thread is not relevant to your interests.
If you aren't seeing it, and aren't interested in doing the system math to see how it could happen, and you don't believe my math that shows how it can happen, this thread might not be of use to you.
I think the real problem is wizard spell slots. Because spells can be cast in higher slots, Wizards don't need 1st level spell slots to cast their 1st level spells. They also don't need a huge number of 1st level slots to make attacks as they have cantrips (which scale).
They really (and all casters too) need to lose low level spell slots as they level up. It seems counter intuitive but makes sense once you think about it.
I have proposed a simple nerf for this. At low levels, casters gain slots in the same manner. Once they have 10 spell slots, they don’t get MORE spell slots, their spell slots simply get more powerful (ie they simply get the 10 strongest spell slots they would get at their level in the PHB).
I don’t think limiting spellcasters to 10 spells per day is particularly onerous, and it really makes you reconsider the powerful level 1 and level 2 spells when you must use a level 3 slot to cast them.
I have proposed a simple nerf for this. At low levels, casters gain slots in the same manner. Once they have 10 spell slots, they don’t get MORE spell slots, their spell slots simply get more powerful (ie they simply get the 10 strongest spell slots they would get at their level in the PHB).
I don’t think limiting spellcasters to 10 spells per day is particularly onerous, and it really makes you reconsider the powerful level 1 and level 2 spells when you must use a level 3 slot to cast them.
Not sure 10 is the right number but it has potential. The downside is all the little stuff a mage does for party goes away at that point. Your mage gives up Utility to be only a buffer or blaster in important combats. If I've only got 10 spells a day definitely not going to waste one on endure elements, or ant haul, or remove curse on anyone but me.
If you play a game where you only have a combat or two a day it would work, If you play a game where you are constantly in danger and the clock is ticking it wouldn't work well
I think maybe a simpler solution would be something like 1st level spells top out in power at say 4 dice of damage. so something like burning hands would max at 4 dice of damage. But if you memorize it as a 5th level spell you can take it up to 8 dice, 6th up to 10, 7 up to 12, 8 up to 15 and 9 up to 20. Then your mage doesn't get anymore high level spells per day but all the older ones become more useful. You could even have spells that were U for upgradeable and stuff like Magic Missile might not be on that list.
Warning vancian magic is extremely hard to modify, sometimes little changes make things overpowered, or completely useless.
