WTF is up with Mob Spawning?

[CyphurTheFox (OP)]

I have a few mob farms on VC, they work quite well, but I can never seem to get the projected rates out of them, despite double, triple, and quadruple checking my mob proofing techniques. I've done a couple of tests of this phenomenon:

Test 1: Grabbing the 9 chunks surrounding my guardian farm, as well as the farm itself into a MCEdit schematic, generating a new singleplayer world, then pasting the schematic on top of the location of an existing sea temple on that world (the World Downloader mod fails to save the monument bounding box) Comparing farm speed between these two, with me being the only person on the server, shows that I pick up XP in the single player world 4x faster than the (identical) copy on the server. This suggests that something server side is slowing down hostile mob spawning.

Test 2: I traveled into fresh (nether) chunks, then built a little box around myself and proceeded to afk for some time. I then used my World Downloader mod to grab the surrounding chunks and entities, while also configuring the download to have the gamerule "doMobSpawning" set to false. I then loaded up the singleplayer world. using the F3 debug screen, entity count read ~16. I also used a command to give all entities glowing, which revealed that the nearby mobs were all fairly close by, none looked like they were the up to 128 blocks away that mobs should spawn. I then did /gamerule doMobSpawning true and within a second, the entity count jumped to 50. The glowing test shows that mobs are spawning reasonable distances away again. This suggests that either mob cap is abnormally low, or the despawn chance of mobs far from the player is artificially increased.


Thus my question: What is up with Mob Spawning, why is it so slow?

[luisc99]

I'm not actually sure what's up with mob spawning. It's been lower than vanilla since before I even joined the server. I remember noticing the same thing back in 2012 when building my first spawner. Weird thing is, I've upped the spawn rates from the default value, iirc to a value even higher than vanilla, and its still low. I'll see if I can up it some more next time I get enough free time to test it all out.

[CyphurTheFox (OP)]

I'm not actually sure what's up with mob spawning. It's been lower than vanilla since before I even joined the server. I remember noticing the same thing back in 2012 when building my first spawner. Weird thing is, I've upped the spawn rates from the default value, iirc to a value even higher than vanilla, and its still low. I'll see if I can up it some more next time I get enough free time to test it all out.

Awesome, Thanks, I also had some ideas for tests I could toss around and I'll run some of them when I have the chance.

[CyphurTheFox (OP)]

I'm not actually sure what's up with mob spawning. It's been lower than vanilla since before I even joined the server. I remember noticing the same thing back in 2012 when building my first spawner. Weird thing is, I've upped the spawn rates from the default value, iirc to a value even higher than vanilla, and its still low. I'll see if I can up it some more next time I get enough free time to test it all out.

Awesome, Thanks, I also had some ideas for tests I could toss around and I'll run some of them when I have the chance.

So when I said "tests" I really meant "a test", which I performed just the other day. The idea was simple, a repeat of my nether test, but more people involved, so that's what I did. I borrowed a few accounts and placed them in the nether. Firstly, mine in the middle of fresh chunks, then recorded entity count, then add an acc on a high platform to up mobcap, then record again, then get a third on, loading chunks but not spawning mobs, then check again.

 So for those unfamiliar with the mobcap calculation, it boils down to a vastly simplified version of this: Players on server * constant = mobcap. ^{[See Footnote]} given this table:

1 Player: ~19
2Players: ~31
3Players: ~47

I think the line of best fit for this data would put the constant at around 15. For reference, the constant in a typical MC game sits closer to 70.

[Footnote]
The Actual calculation is based on a certain radius of chunks around the player, and is affected by player proximity.
For a better explanation of this, see https://minecraft.gamepedia.com/Spawn#Spawn_cycle.




Addendum(added after writing all of the above):

Actually, I think I just found the real source of the problem, and its really simple. I know from prior testing that the server render dist is 9. I used some simple measuring to get that because I needed it to do some bedrock removal with dragon eggs. Anyway, I was reading the gamepedia article on spawning , I noticed this little gem under "notes"

If the player's view distance or the server view distance in multiplayer is at 9 or below, mob spawning is severely reduced (or they despawn too quickly), and may result in the player encountering no mobs at all. Set the view distance to 10 or higher to ensure mobs spawn correctly.

So uh, yeah. Fix the problem (at least in theory) by increasing the server render dist to 10. Go Figure.