mojira.dev
MCPE-132746

Mobcap in the Nether for Monsters is way too low

Steps to reproduce

  1. In the nether, in the beta, fly around to various places.

  2. /execute @e[r = 45] ~~~ say hi

  3. /execute @e ~~~ say hi

Expected result

Based on changelog statements:

  • In step (2) 15-20 monsters say "hi".

  • In step (3) at simulation distances 6+ where spawn sphere is near full size, 70 monsters say "hi" (Java parity).

Actual result

  • The results below will be skewed if you are in/near a bastion or nether fortress because of the abundance of structure spawns/spawn spots.

  • Hoglins should be subtracted from Java results since they count toward the “hostile” cap in Java, but the “animal” cap in Bedrock.
    ||Range||Sim 4||Sim 6||Sim 8+||Java||
    |Step (2) / 45 block radius|8-12 monsters|4-6 monsters|4-6 monsters|???|
    |Step (3) / full sim distance|8-12 monsters|15-20 monsters|25-30 monsters|70 monsters|


Bug description
In the latest beta the following change was mentioned:
"Monster spawning in the Nether has been changed to more closely match Java Edition (MCPE-65762)
There will now be a maximum of 15 to 20 mobs around the player in the Nether, instead of 30 to 40"

This change does not make sense after testing, since there can now only be up to 8 mobs in the nether instead of 16 of the Monster population group (See

[media]

)

Related bug
Simulation distance 4
According to MCPE-65762, there are spawning way too many mobs in the Nether. The mentioned fix in the changelog was supposed to fix this bug, however the fix does not solve the issue in one way or the other.

Since the despawning sphere is so tiny (Mobs despawn 44+ blocks away) on simulation distance 4, new mobs tend to spawn more frequent and more closely to the player. The issue mentioned in the bugreport is about the mobs spawning too close to the player, making them overwhelmed by the amount of mobs. Reducing the Monster cave mobcap to 8 instead of 16 would still make this issue exist and doesn't change anything.

Ghasts
The bugreport also has a lot of duplicates to "Ghast spawning way too frequent". This change would not even fix that issue in the slightest, since Ghasts would spawn anyways, no matter how many mobs there are in the area. People would still be attacked in every direction by Ghasts.

Another issue with Ghast, which was also mentioned in a duplicate bugreport, that they would spawn in nether tunnels. There is currently a bug, like with Magma Cubes, where they can spawn inside 2-tall areas. It would be better fixing this bug than actually focussing on lowering the mobcap while Ghasts aren't much affected by it: (See Chri O' Briens comment about it https://bugs.mojang.com/browse/MCPE-65762?focusedCommentId=1016088&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-1016088)

Magma cubes
Also was mentioned that Magma Cubes spawn inside made nether tunnels everywhere. The change in 1.17.20.20 beta does not adress this issue, while it was mentioned multiple times in the bugreport that a lot of Magma Cubes would spawn there. The issue of Magma Cubes is that they can spawn in tiny rooms, even if their bounding box would align with a wall or ceiling. It is therefor better to focus on fixing this bug: https://bugs.mojang.com/browse/MCPE-46540, then reducing the mobcap in half.

Expected behavior
The Monster cavecap in the Nether is 16

Actual behavior
The Monster cavecap in the Nether is 8, making mobfarms even slower in the nether like they already were. A lot of people who make mobfarms would be affected by the lower cap, making their farms twice as slow, while the main issue here is not fixed

Proper fix
For the fix it is better to look at the mobspawning on simulation distance 4 and find a solution that would target only that simulation distance and not others. The simulation distance has a lot of problems regarding the small spawning range (24-44 blocks, outside that everything would despawn)

Another fix would be to get rid of simulation distance 4 entirely and focus on simulation distance 6. I hope most devices and servers would be able to run the small increase in loaded chunks, but this should stop any issues regarding spawning of mobs

Attachments

Comments

migrated
[media][media][media][media][media]
Desert Prince

This change is as bad if not worse than the sim distance change forced on realms. Give the OP the ability to control this from a pre-populated list of options on the game settings. This goes against the premise of taking away mechanics / undoing features that Mojang has always upheld. Mob caps do not need to be the same on Bedrock as Java. Bedrock has often been developed to make it MORE challenging than Java. I hope this change is undone. How it got recommended by anybody is mind-boggling. This was a non-issue that they were trying to fix.

The proposed fix will NOT work on realms. This was a huge issue when they tried to increase the sim distance and had to walk that change back. Realms infrastructure needs to be addressed before that can be done. It also broke a lot of farms.

GoldenHelmet

We should remember that MCPE-65762 was created, and received most of its votes, at a time when spawning was limited to a 44 block radius on all simulation distances. Since the spawn radius for sim6+ expanded in 1.16.100, that bug report now only applies to sim4. At the present time there is a huge difference in spawn density between sim4 and sim6+. You can easily observe this by spending a couple of minutes walking around killing monsters in a flat world at night.

For this reason, I expect that cutting the monster cap in half to make sim4 more bearable will negatively impact not only farms, but also general gameplay on sim6+. It goes in the opposite direction of the despawning changes in 1.16 and brings to mind some of the old comments on MCPE-21856 that describe the nether as a walk in the park prior to instant despawning.

If indeed a monster cap of 8 gives the right density for sim4, then it should be applied only to sim4.

Here is some mathematical analysis support my assertion. Consider the area checked for monster density during a spawn attempt:

  • Sim4, 25 blocks from the player: 21% of the checked area can spawn monsters.

  • Sim6, 25 blocks from the player: 79% of the checked area can spawn monsters.

  • Sim4, 43 blocks from the player: 16% of the checked area can spawn monsters.

  • Sim6, 43 blocks from the player: 68% of the checked area can spawn monsters.

The impact of these percentages is that over time the density of successful spawns "near" the player will be much higher on Sim4. That's because the cap check area that is spawnable ≈ the cap check area that can (a) contribute to the cap, and, therefore, (b) contribute to spreading out spawns.

Details on the calcuations above:

[media]

migrated

ItRichHeart not sure why getting rid of Sim distance 4 would benefit anything.

As for lowering the Mob cap Mojang @helper what is the rationale for that? Yes, it can be a pain sometimes but is it the number of mobs or the particular type that is the issue. I know for a fact... need to report this one...Skeletons standing on/in a block that impairs movement removes the random chance of them missing (yes I know coding perspective the chance isn't really gone, but their slowed movement makes shots that much more accurate). If we are going to implement actual parity fixes let us start with Agro distance of Zombie Piglins (increase from 20 effectively 10 blocks to 88 effectively 44 blocks), Magma Cubes attacking Golems of either type.

 

This is one of those unwelcome changes as it is rendering Nether Side Gold Farms useless for a start. Will make a video response to better clarify these points.

migrated

Slowly more old/low-end devices have been removed from the supported list. I think at this point we could standarized Simulation Distance 6 and maybe even 8. That could help solve this issue and a some more like mob spawning/despawning.

migrated

In the changelog MCPE-65762 was mentioned as the "inspiration" for this change, and I think that bug report is more of a personal issue. So if that bug report is what you want to address, maybe have each difficulty spawn different amounts of mobs so players with the want to play with less mobs can do so without effecting the entire community. With this implementation, people who feel that mobs spawn to frequently can play on easier versions with less mobs, while other players can play with a larger mobcap. This is also would be a small parity change as the rate of which mobs spawn on Java Edition changes per difficulty level. MCPE-65762 also seems to heavily lean on bugs such as MCPE-46540 where mobs spawn in smaller spaces than intended, so people would also feel less crowded with that bug fixed.

ravinmaddhatter

I did some math in Sim 10 the density is way way less after this change. It is not parity. it is actually a substantial reduction in cap. Honestly. if the nether is "too hard" that is what difficulty settings are for.

The max average mobs in bedrock is about 50 on bedrock after the change, but it is 70 in java....

GoldenHelmet
[media]

can be used to compare the effective density of "mobs around the player" on different simulation distances. In it, the player stands on the chain command block at 8, 8 and presses the stone button on the impulse command block to start a new test. A scoreboard system tracks time elapsed in seconds and continuously updates counts of monsters within 45 blocks and within 65 blocks. On Sim4 both counts are always the same because monsters despawn and cannot spawn past 44 blocks from the player. 65 blocks was chosen for the larger radius because that is roughly both the range at which mobs are rendered in Bedrock Edition, and the largest follow range for monsters (e.g. ghasts), so I thought that might be what was meant by "mobs around the player".

Demos running the test world for 1 minute on Sim4 and Sim6:

[media][media]
  • Within 45 blocks of the player, the number of monsters on Sim6 tends to be more than 50% lower than the number of monsters on Sim4.

  • Within 65 blocks of the player, the number of monsters on Sim6 tends to be ~20-40% higher than the number of monsters on Sim4.

  • If you compare the number of monsters to the monster-viable space being tested here, i.e. r45 on Sim4 and r65 on Sim6, then

    • For the 2D area the effective density on Sim6 is about 2/3 or 67% of the density of Sim4.

    • For the 3D volume the effect density on Sim6 is just less than 1/2 or 50% of the density of Sim4.

Some further technical notes on the test world:

  • It uses the overworld surface monster cap, which is the same (8) as the beta nether monster cap.

  • Every x, z coordinate is spawnable: There is a 51 x 51 spawnable platform 25 blocks above ground centered over the player, which fits entirely in the Sim4 spawn radius. The ground under the elevated platform is replaced with glass to block cave spawns.

  • The fact that nether spawns can occur on multiple Y-levels in a column is not likely to make much difference; if anything it might sometimes cause a few more monsters to aggregate at the edges of Sim4 (>32 blocks from the player on either side, where one side doesn't cap spawns on the other side), but the cap totals will still be the same.

  • The Sim4 spawn radius is marked by polished granite.

  • Zombie villagers are killed instantly so that they do not skew results. Zombie villagers do not count toward any cap (MCPE-62030).

  • Phantom spawning is disabled so that it does not skew results. Phantom's special spawn behavior makes them block spawns in different places from where they spawn (MCPE-106557).

ItsRichHeart

(Unassigned)

554980

Confirmed

Multiple

1.17.20.21 Beta, 1.17.20.20 Beta

1.17.30.24 Beta

Retrieved