Note
In the original description below the data for Ender Pearl damage is accurate only for specific combinations of Protection levels, and the analysis is incorrect. For more extensive test analysis see comment.
According to the Minecraft Wiki, each "unit" of protection adds 4% to the damage reduction, meaning that it takes 20 units to reach the maximum 80%. (Note: Projectile Protection, Blast Protection, and Fire Protection are worth double, while Feather Falling is worth triple). However, it appears that the actual amount is 5% per unit, meaning that it takes 16 units to reach the maximum 80% protection.
How to Reproduce
Test different amounts of protection. Make sure you are testing a type of damage that armor does not normally protect you from. I tested this with ender pearls:
No protection: 5 damage
1-4 protection: 4 damage
5-8 protection: 3 damage
9-12 protection: 2 damage
13 protection and higher: 1 damage
I also tested this with higher falls, and a full set of Protection IV armor still reduces the damage by 80%. I can confirm that 16 units of protection is the maximum because you do not reduce the damage any further if you add Feather Falling.
Expected Result
Ender pearls would give the following amounts of damage:
No protection: 5 damage
1-5 Protection: 4 damage
6-10 Protection: 3 damage
11-15 Protection: 2 damage
16 protection and higher: 1 damage (Note: Feather Falling is required for any protection level above 16)
Actual Result
See the "How to Reproduce" section.
Linked issues
is duplicated by 1
Comments 9
By the way, can you tell if this has been the case for a while or if it has just newly happened in 1.16.201? Does this only apply to fall damage, or all types of damage?
The bug here is not that Protection gives 5% damage reduction per level; instead, (from what I can tell) the bug is that Protection 4 functions as if it were Protection 5 (the converse of MCPE-82952). Using different combinations of protection on multiple pieces of armor gives different results.
So with the ender pearl example:
Total prot | Prot breakdown | Damage |
|---|---|---|
5 | 2 + 2+ 1 | 4 |
5 | 3 + 2 | 4 |
5 | 4 + 1 | 3 |
9 | 3 + 2 + 2 + 2 | 3 |
9 | 4 + 3 + 2 | 3 |
9 | 4 + 4 + 1 | 2 |
I did not even try to test protection on fall damage because I could not get consistent fall damage from high falls. (Maybe a trace of MCPE-43356?)
I also tested 50-damage mele attacks against a health-boosted player wearing full leather armor (7 armor points) with different levels of protection. Results were as expected for 4% reduction/level applied after armor reduction, except when protection 4 was involved. Here is the most revealing data:
Total prot | Prot breakdwon | Damage |
|---|---|---|
4 | 3 + 1 | 30 |
4 | 4 + 0 | 28 |
5 | 3 + 2 | 28 |
5 | 4 + 1 | 27 |
8 | 2 + 2 + 2 + 2 | 24 |
8 | 4 + 2 + 2 | 23 |
8 | 4 + 4 | 21 |
Note that the damage of 23 with Prot 4 + 2 + 2 is what you would expect from Prot 9, and the damage of 21 from Prot 4 + 4 is what you would expect from Prot 10.
This bug also applies to blast protection, projectile protection, and fire protection. It most likely applies to feather falling also, but I have encountered randomness in fall damage and so have not tried to test it.
For blast protection and projectile protection, there is clearly an additional bonus for each piece of armor that has the enchantment. For fire protection there may be a very small bonus, but it is difficult to confirm because damage is slightly randomized.
Blast protection
Using TNT (51 damage) with full leather armor (28% damage reduction):
Total blast prot | breakdwon | Damage |
|---|---|---|
4 | 3 + 1 | 22 |
4 | 4 + 0 | 20 |
6 | 3 + 3 | 16 |
6 | 2 + 2 + 2 | 14 |
6 | 4 + 2 | 13 |
h4. Projectile protection
Using arrows set to 50 damage with a behavior pack, and full leather armor (28% damage reduction):
Total proj prot | breakdwon | Damage |
|---|---|---|
4 | 3 + 1 | 21 |
4 | 4 + 0 | 20 |
5 | 3 + 2 | 17 |
5 | 3 + 1 + 1 | 16 |
5 | 2 + 1 + 1 + 1 | 15 |
5 | 4 + 1 | 15 |
h4. Fire protection
Rapidly teleporting player into lava then water, with lava set to 50 damage with a behavior pack, and full leather armor (30% damage reduction--not sure why this is different for lava, but I consistently take 35 damage from 50-damage lava vs. 36 damage from a 50-damage arrow or 51-damage TNT). There is also some randomness in the final damage as shown below.
Total fire prot | breakdwon | Damage |
|---|---|---|
4 | 3 + 1 | 24 |
4 | 4 + 0 | 23-24 |
5 | 3 + 2 | 21-22 |
5 | 3 + 1 + 1 | 21 |
5 | 2 + 1 + 1 + 1 | 21 |
5 | 4 + 1 | 20 |
h4. Notes
The bonuses become very significant higher levels. For example projectile protection 5 should give 5 * 8% = 40% damage reduction after armor, and the final damage is expected to be floor(36 * (1 - 0.4)) = 21. The actual damage of 15 for projectile protection 4 + projectile protection 1 (see chart above) is 21 / 36 = 58% damage reduction instead of 40%. That means projectile protection 4 + 1 is (1 - 58/40) = 45% stronger than expected!
In terms of the "Enchantment Protection Factors" described on the wiki at https://minecraft.fandom.com/wiki/Armor#Enchantments, for a single piece of enchanted armor the EPFs seem to be as follows.
Expected:
| Level 1 | Level 2 | Level 3 | Level 4 |
|---|---|---|---|---|
Protection | 1 | 2 | 3 | 4 |
Speciality Protections | 2 | 4 | 6 | 8 |
Observed:
| Level 1 | Level 2 | Level 3 | Level 4 |
|---|---|---|---|---|
Protection | 1 | 2 | 3 | 5 |
Blast protection | 3 | 5 | 7 | 11 |
Projectile protection | 3 | 5 | 7 | 11 |
Fire protection | 2 | 4 | 6 | 9 |
This means that I have wasted time and effort in creating multiple pieces of armor with different enchantments for different situations. This issue makes a lot of armor enchantments like Feather Falling and Fire Protection useless in the late game. Can anyone else confirm this?