lower ruined portal structure spawn rate #3901

SmokeyDope · 2023-08-28T00:31:25+02:00

SmokeyDope commented

2023-08-28 00:31:25 +02:00

I finally got around to figuring out how to tweak the spawn rate of ruined portals. It is widely agreed by players that they are a little too common and break progression by saturating player with loot and gold.

The spawn rate of ruined portals is determined by the variable "fill_ratio". It was at '0.01'. To make sure this was for sure what controlled spawn rate I set it to 1 which made portals spam everywhere.

##Testing
My devised test was as followed:
Adjust the value, generate a new flat world in creative, pick a direction and start flying till I see a portal, then note the time it takes to find another portal after. Find a few portals to confirm the subjective statistical chance with a decent sample size. Not mathematical by any means. Exit world, adjust value, generate new world, rinse repeat.

If it took me more than 5 minutes of speed creative flying to find a portal I deemed it too rare. In these cases luck may have not been on my side. I gave 0.001 about 10 minutes before I gave up which is twice the time allocated to finding the others which indicates its just too low of a chance.

fill_ratio value: experimental results
1: spams portals
0.01: default, too common
0.006: portals much rarer but still findable within a few minutes of creative flight
0.001: Could not find portal, too rare

0.006 may still be a little too common but it is a big step in the right direction. I figured its best to not overdue it. I believe this change will allieiate many complaints of portals being too common.

Edit: Something I forgot to add is this. setting fill ratio to 1 caused the game to fail to exit correctly. Im not sure why but I had to force the program to close. May be worth looking into?

I finally got around to figuring out how to tweak the spawn rate of ruined portals. It is widely agreed by players that they are a little too common and break progression by saturating player with loot and gold. The spawn rate of ruined portals is determined by the variable "fill_ratio". It was at '0.01'. To make sure this was for sure what controlled spawn rate I set it to 1 which made portals spam everywhere. ##Testing My devised test was as followed: Adjust the value, generate a new flat world in creative, pick a direction and start flying till I see a portal, then note the time it takes to find another portal after. Find a few portals to confirm the subjective statistical chance with a decent sample size. Not mathematical by any means. Exit world, adjust value, generate new world, rinse repeat. If it took me more than 5 minutes of speed creative flying to find a portal I deemed it too rare. In these cases luck may have not been on my side. I gave 0.001 about 10 minutes before I gave up which is twice the time allocated to finding the others which indicates its just too low of a chance. fill_ratio value: experimental results 1: spams portals 0.01: default, too common 0.006: portals much rarer but still findable within a few minutes of creative flight 0.001: Could not find portal, too rare 0.006 may still be a little too common but it is a big step in the right direction. I figured its best to not overdue it. I believe this change will allieiate many complaints of portals being too common. Edit: Something I forgot to add is this. setting fill ratio to 1 caused the game to fail to exit correctly. Im not sure why but I had to force the program to close. May be worth looking into?

👍 1

kneekoo commented

2023-08-28 03:09:59 +02:00

Thanks for looking into this. But I wonder if this is the right path forward, by just tweaking the fill ratio. When I looked at a few seeds with Chunkbase, I noticed that each "super-chunk" (512x512 blocks) usually has one ruined portal. This makes them reasonably rare and the grid-based distribution (similar to slime chunks) also brings them closer to what we need.

Check out this earlier comment that I made about them.

Thanks for looking into this. But I wonder if this is the right path forward, by just tweaking the fill ratio. When I looked at a few seeds with [Chunkbase](https://www.chunkbase.com/apps/seed-map#y), I noticed that each "super-chunk" (512x512 blocks) usually has one ruined portal. This makes them reasonably rare and the grid-based distribution *(similar to slime chunks)* also brings them closer to what we need. Check out this [earlier comment](https://git.minetest.land/MineClone2/MineClone2/issues/3004#issuecomment-66174) that I made about them.

SmokeyDope commented

2023-08-28 03:29:25 +02:00

@kneekoo The general player consensus is that ruined nether portals are too common in the overworld. I've played enough mcl2 for this to be experimentally verified for me personally. Many players have reported this as an issue separately over time.

We can generate graphs, compare the math, and argue statistics till the cows come home. All the while nothing actually gets done to fix the percieved issue.

Or we can just try something, take a step forward and slightly tweak a value and see if it alleviates the issue. There may be multiple ways to fix this, I choose the simplest and easiest way that is easily reversible if it ends up not working out. As far as I am concerned I did the leg work to dial in the values and spent a good chunk of my time testing to make sure its rarer but not impossible to find.

The disagreement comes down to a difference in opinion. You apparently think that nether portals are spawning in the correct amount with supporting evidence of a generated graph of a sample area. I (and many others) played the game, ran into a dozen ruined nether portals within half an hour of playtime, and thought it was slightly too much after many hours of playtesting.

I like to think that experience and gut feeling intuition are important to consider.

Hopefully some other contributors will chime in on this with their thoughts. If it comes down to a community poll/vote at least we can put this old issue to rest.

see MineClone2/MineClone2#3726

Also go into the discord and search up 'portal' and read through some of it.

@kneekoo The general player consensus is that ruined nether portals are too common in the overworld. I've played enough mcl2 for this to be experimentally verified for me personally. Many players have reported this as an issue separately over time. We can generate graphs, compare the math, and argue statistics till the cows come home. All the while nothing actually gets done to fix the percieved issue. Or we can just try something, take a step forward and slightly tweak a value and see if it alleviates the issue. There may be multiple ways to fix this, I choose the simplest and easiest way that is easily reversible if it ends up not working out. As far as I am concerned I did the leg work to dial in the values and spent a good chunk of my time testing to make sure its rarer but not impossible to find. The disagreement comes down to a difference in opinion. You apparently think that nether portals are spawning in the correct amount with supporting evidence of a generated graph of a sample area. I (and many others) played the game, ran into a dozen ruined nether portals within half an hour of playtime, and thought it was slightly too much after many hours of playtesting. I like to think that experience and gut feeling intuition are important to consider. Hopefully some other contributors will chime in on this with their thoughts. If it comes down to a community poll/vote at least we can put this old issue to rest. see https://git.minetest.land/MineClone2/MineClone2/issues/3726 Also go into the discord and search up 'portal' and read through some of it.

kneekoo commented

2023-08-28 20:40:58 +02:00

I know they're too many (personal experience), and I'm fine with tweaking the value. But while it's still an experiment that we will have to wait and see if it's reasonable enough for everyone, I was also thinking about the algorithm AncientMariner used to have deterministic slime chunks.

I mentioned it just so it's not forgotten, and because it might not be hard to implement. But if no one wants to try that, I'm happy with lowering the fill ratio because we need something done about them.

I know they're too many (personal experience), and I'm fine with tweaking the value. But while it's still an experiment that we will have to wait and see if it's reasonable enough for everyone, I was also thinking about the algorithm AncientMariner used to have deterministic slime chunks. I mentioned it just so it's not forgotten, and because it might not be hard to implement. But if no one wants to try that, I'm happy with lowering the fill ratio because we need something done about them.

ancientmarinerdev commented

2023-08-28 22:23:36 +02:00

I know they're too many (personal experience), and I'm fine with tweaking the value. But while it's still an experiment that we will have to wait and see if it's reasonable enough for everyone, I was also thinking about the algorithm AncientMariner used to have deterministic slime chunks.

I mentioned it just so it's not forgotten, and because it might not be hard to implement. But if no one wants to try that, I'm happy with lowering the fill ratio because we need something done about them.

The algorithm I used for slime chunks wasn't great. You can get none near each other and then 2-4 really close. If I was to revisit that, I'd do it differently. Don't get me wrong, no regrets and we need slime chunks, but getting them better distributed could be an improvement. It's a problem we could probably looking into, and this work can be applied to where peaceful animals spawn in mc, but for now decreasing fill ratio would probably be a sensible tactical solution to make this less of a cake walk.

> I know they're too many (personal experience), and I'm fine with tweaking the value. But while it's still an experiment that we will have to wait and see if it's reasonable enough for everyone, I was also thinking about the algorithm AncientMariner used to have deterministic slime chunks. > > I mentioned it just so it's not forgotten, and because it might not be hard to implement. But if no one wants to try that, I'm happy with lowering the fill ratio because we need something done about them. The algorithm I used for slime chunks wasn't great. You can get none near each other and then 2-4 really close. If I was to revisit that, I'd do it differently. Don't get me wrong, no regrets and we need slime chunks, but getting them better distributed could be an improvement. It's a problem we could probably looking into, and this work can be applied to where peaceful animals spawn in mc, but for now decreasing fill ratio would probably be a sensible tactical solution to make this less of a cake walk.

👍 1

ancientmarinerdev commented

2023-09-01 00:45:15 +02:00

I have tested this, and it's definitely much rarer. It looks around 60% as much which I think feels a good amount based on my personal experience. They aren't needed to get to the nether, but help. I would rather people got excited by these rather than thinking "oh, not another one". Who knows, we might revisit this at some point if it turns out people feel it's too rare, and we need a value somewhere in the middle, but we will have to guage feedback, but I am personally happy with this (had about 4 on my starting island which would be quite findable).

👍 2

ancientmarinerdev approved these changes 2023-09-01 00:46:43 +02:00

ancientmarinerdev left a comment

Based on my previous comments, I will approve this. Thanks for fixing this and testing. It's clear you've put time into that, and I do trust that. The results look better from my testing.

ancientmarinerdev force-pushed lower_ruined_portal_generation_rate from 8e3b7974f6 to aa08a176fc

2023-09-01 00:47:11 +02:00

Compare

ancientmarinerdev added the

mapgen

label 2023-09-01 00:47:18 +02:00

ancientmarinerdev added this to the 0.85.0 - Fire and Stone milestone 2023-09-01 00:47:25 +02:00

ancientmarinerdev merged commit 273165ce3b into master

2023-09-01 00:47:54 +02:00

ancientmarinerdev deleted branch lower_ruined_portal_generation_rate

2023-09-01 00:47:54 +02:00

ancientmarinerdev referenced this issue from a commit