My friends and I just started a new world on the latest version of 1.16, and we set the generation mode to Large Biomes. We started to notice that we were having a lot of trouble finding ore like iron and gold, and even more trouble finding diamonds. I got really frustrated and started digging out entire chunks, and we discovered that only about 1/2 of the chunks even had diamonds spawned in them! Is this a new thing for Large Biomes? Or does Large Biomes make ore more spread out in general?
For an example of this, use seed -597086656834133 with Large Biomes enabled. You'll spawn on an island. We dug out most of the underground of that island and didn't find diamonds in half of the chunks.
The only effect that Large Biomes should have on world generation is to make biomes larger, everything else is exactly the same, aside from biome-specific features of course, but this does not include the majority or ores (aside from emerald in extreme hills and gold above layer 32 in mesa biomes) or most other underground features (for example, caves will be identical aside from land/ocean).
Otherwise, there are specific seeds that break world generation, as mentioned here, but it is extremely unlikely that you've stumbled upon one by accident, and otherwise I consider an analysis with a tool like MCEdit to be the one way to verify if any changes have been made (unfortunately, it doesn't work on worlds since 1.13, though people have made their own tools, for example, in this bug report regarding reduced ore generation in 1.13 snapshots (which has been fixed, or claimed to be; the last comment claims otherwise, in particular, they found over 40% less diamond in 1.13.1 than 1.12.2, which is unlikely to happen by random chance, same for other ores being much less or much more common. However, if this were correct I'd think that many others would have noticed).
Note that every time a new updates comes out there are claims of greatly reduced ore spawn rates, such as in this thread for 1.7, but I and others found no evidence to support that, based on an analysis done with MCEdit (if anything, 1.7 has slightly more ores than 1.6 due to less dirt/gravel/caves, which increased much more significantly in 1.8, no idea about 1.13+ though).
Sometimes two small veins connect together to form one larger vein if they are close. Diamond veins range widely in size from 2 to 10 in my experience.
How many chunks did you dig out? You need a pretty big sample size to draw conclusions like this.
I've now sampled 100 chunks around spawn (with your seed and Large Biomes) and found 76 diamond veins, so 3/4.
I think the bit about ores being shifted between chunks must have changed, I've found multiple instances of 2 diamond veins in the same "shifted chunk" and not noticed any where there were more than one in the same chunk (except when one crossed the border.)
My sample area was from -80/1/112 to 80/15/272
--
Having nothing better to do with my life I repeated the test with 1024 chunks with the same result, diamond veins in 780 out of 1024 chunks or 76,2%
I've now sampled 100 chunks around spawn (with your seed and Large Biomes) and found 76 diamond veins, so 3/4.
I think the bit about ores being shifted between chunks must have changed, I've found multiple instances of 2 diamond veins in the same "shifted chunk" and not noticed any where there were more than one in the same chunk (except when one crossed the border.)
My sample area was from -80/1/112 to 80/15/272
--
Having nothing better to do with my life I repeated the test with 1024 chunks with the same result, diamond veins in 780 out of 1024 chunks or 76,2%
I'm curious, did you count the number of ore per chunk? There should be an average of about 3.7, which is about 2/3 of what would generate if every vein were intact (the average size of a vein is 5.8 but a lot is lost in bedrock, in particular, veins starting at y=0 won't ever generate anything at all because they only "grow" downwards from their start, which can also be one layer below the initial point so this also applies to some veins at y=1 (conversely, this is why diamond is so rare at y=15 despite an otherwise uniform distribution of initial starting points).
Otherwise, I see no real evidence that diamond has been made rarer, with the 1.13 snapshots being the only time an actual decline occurred (which was supposed to have been fixed), if not based on actual analysis since I don't know of any such tools that work in current versions but I'd think it would be all over the place if diamond were all of a sudden much rarer.
Also, based on previous observations it appears that Mojang did remove the "offset by 8" thing so decorations are now centered within a 3x3 chunk area instead of 2x2 (the main difference as far as world generation is concerned is that features can now extend up to 16 blocks from their center instead of just 8, though I don't know of any vanilla features, excluding structures, which use a different generation method which places them chunk-bu-chunk, that might take advantage of this, outside of mods; for example, the largest trees in TMCW require a reduced random offset so they don't spill into unloaded chunks, causing runaway world generation when they are loaded and in turn decorated. I've actually thought of changing this myself due to this; this change also greatly benefits modders, many of whom do not add the offset of 8).
I'm curious, did you count the number of ore per chunk? There should be an average of about 3.7, which is about 2/3 of what would generate if every vein were intact (the average size of a vein is 5.8 but a lot is lost in bedrock
I hadn't but I have now checked the 10X10 chunk area I did first.
I found:
0 2 4 1 3 0 1 3 5 8
0 7 1 8 1 0 0 5 3 2
2 5 4 5 5 6 6 6 6 4
1 4 0 0 0 0 2 0 4 4
7 4 5 0 2 2 6 0 8 6
0 4 4 4 4 1 6 4 7 0
6 5 4 4 4 5 1 8 0 4
5 2 0 3 5 0 5 3 1 1
3 4 4 3 0 5 4 0 4 0
8 8 1 9 3 5 5 0 0 4
ores per chunk (running in a serpentine pattern, so the horizontal repeats weren't actually next to each other.)
which I make as 328 ores in 78 chunks out of 100.
That's 3.28 ores per chunk and approximately 4.3 ores per vein using the 76 veins I counted last time.
The numbers are a tad inconsistent sounding since the 76 veins counted veins that I figured "belonged to" the chunks but the ores are strictly ores that are in the chunks in question, so there are a few veins that are split between chunks and a few ores intruding into the area from veins that "belonged to" chunks outside the area.
Am I correct in assuming that, with no shifting of the ores between chunks, a diamond vein can only extend a single block outside its chunk? (That's a blocks distance, not just a single ore block.) And that therefore any merging of veins can only occur along the chunk borders?
Am I correct in assuming that, with no shifting of the ores between chunks, a diamond vein can only extend a single block outside its chunk? (That's a blocks distance, not just a single ore block.) And that therefore any merging of veins can only occur along the chunk borders?
Yes, the only difference from before is where the "chunk borders" are, now they are actual chunks instead of "offset chunks".
Also, 3.28 ores per chunk isn't inconsistent with an average of 3.7 since there will be variance over a relatively small area, and it suggests that MC-126373 had actually been fixed (in the last comment somebody found an average of only 2.26 ore per chunk compared to 3.86 for the same area in 1.12.2. However, the former was supposedly in 1.13.1, after the bug has been marked as fixed, and while ore density does vary over relatively small areas 2.26 seems too low, so I'm not actually sure).
For comparison, I used MCEdit to analyze 10 different 10x10 chunk areas of a world created in 1.6.4 and found 307, 332, 287, 326, 376, 267, 337, 329, 317, and 289 diamond ore, an average of 3.167 per chunk, compared to the Wiki's old figure of 3.097 per chunk (not sure what version this was from but ore generation was unchanged from Beta 1.8 to 1.6.4, with a slight increase expected in 1.7 due to less dirt/gravel/caves and a greater increase in 1.8 due to an increase in average vein size).
The variation from the average was -15.7 to +18.7%, and a similar variation from an average of 3.7 gives a range of 3.12 to 4.39, or as low as 3.00 if a range of +/-18.7% is assumed. I'd expect that 1.7+ worlds have less variation due to less variation in cave density, although I found only 263 diamond in an area with very few caves and 341 in an area with a lot so this isn't the only factor.
Also, according to this report vein sizes 1-3 now generate ores, while before they would result in nothing, suggesting a changes to the code that generates veins of ore. The numbers given are otherwise consistent though; a size of 6 averages about 4.4 ore per vein in either case (1 million simulated veins using 1.6.4's/pre-1.13's ore generation code vs the results in the link for four 40x40 chunk areas at one vein per chunk), as does a size of 5 (about 3.6) and 4 (about 2.4) so whatever change Mojang made only affects the smallest sizes (3 is only about 0.46 in 1.6.4 but 2.65 in 1.13), and it would have no impact since the smallest vein size used in vanilla was 6 for lapis prior to 1.8 (ancient debris may be smaller).
Also, somebody (recently?) edited all the sizes given on the Wiki to range from 1 to the number in the code, which is only loosely related to the average size of a vein:
Coal ore can generate in the Overworld in the form of mineral veins. Coal ore attempts to generate 20 times per chunk in veins of size 1-17, from levels 0 to 127, in all biomes. https://minecraft.gamepedia.com/Coal_Ore#Natural_generation
Not even close - dirt actually averages about 112 blocks per vein, as anybody who has ever dug one out would know, and similarly coal averages about 22 with 20, 22, and 24 being the most common sizes; only sizes" 12-13 closely match the average with smaller sizes being smaller and larger sizes larger:
The reason for this discrepancy, and confusion, is because there is indeed a loop that iterates (size) number of times but each iteration places a roughly spherical blob of blocks each of which can be more than one block in size, with a total of 4 loops involved; for diamond, size 8, the average number of iterations in the innermost loop is 13.7 while for coal, size 17, it is 75.5, and dirt, size 33, is 496.
My friends and I just started a new world on the latest version of 1.16, and we set the generation mode to Large Biomes. We started to notice that we were having a lot of trouble finding ore like iron and gold, and even more trouble finding diamonds. I got really frustrated and started digging out entire chunks, and we discovered that only about 1/2 of the chunks even had diamonds spawned in them! Is this a new thing for Large Biomes? Or does Large Biomes make ore more spread out in general?
For an example of this, use seed -597086656834133 with Large Biomes enabled. You'll spawn on an island. We dug out most of the underground of that island and didn't find diamonds in half of the chunks.
The only effect that Large Biomes should have on world generation is to make biomes larger, everything else is exactly the same, aside from biome-specific features of course, but this does not include the majority or ores (aside from emerald in extreme hills and gold above layer 32 in mesa biomes) or most other underground features (for example, caves will be identical aside from land/ocean).
Otherwise, there are specific seeds that break world generation, as mentioned here, but it is extremely unlikely that you've stumbled upon one by accident, and otherwise I consider an analysis with a tool like MCEdit to be the one way to verify if any changes have been made (unfortunately, it doesn't work on worlds since 1.13, though people have made their own tools, for example, in this bug report regarding reduced ore generation in 1.13 snapshots (which has been fixed, or claimed to be; the last comment claims otherwise, in particular, they found over 40% less diamond in 1.13.1 than 1.12.2, which is unlikely to happen by random chance, same for other ores being much less or much more common. However, if this were correct I'd think that many others would have noticed).
Note that every time a new updates comes out there are claims of greatly reduced ore spawn rates, such as in this thread for 1.7, but I and others found no evidence to support that, based on an analysis done with MCEdit (if anything, 1.7 has slightly more ores than 1.6 due to less dirt/gravel/caves, which increased much more significantly in 1.8, no idea about 1.13+ though).
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?
But were there "too many" diamond veins in the other chunks?
The game does (or should anyway) try to generate one diamond vein per chunk but it doesn't necessarily end up in that chunk.
So called decoration, ores, trees, houses etc are shifted by half a chunk, to the south east I think.
So for your test you should measure between chunk centers rather than chunk edges.
Just testing.
From what we found there weren't extra diamond veins in adjacent chunks.
Sometimes two small veins connect together to form one larger vein if they are close. Diamond veins range widely in size from 2 to 10 in my experience.
How many chunks did you dig out? You need a pretty big sample size to draw conclusions like this.
I've now sampled 100 chunks around spawn (with your seed and Large Biomes) and found 76 diamond veins, so 3/4.
I think the bit about ores being shifted between chunks must have changed, I've found multiple instances of 2 diamond veins in the same "shifted chunk" and not noticed any where there were more than one in the same chunk (except when one crossed the border.)
My sample area was from -80/1/112 to 80/15/272
--
Having nothing better to do with my life I repeated the test with 1024 chunks with the same result, diamond veins in 780 out of 1024 chunks or 76,2%
Just testing.
75% is not a terrible rate though. Some people here have said diamond is too common, I think this is a nice balance.
I'm curious, did you count the number of ore per chunk? There should be an average of about 3.7, which is about 2/3 of what would generate if every vein were intact (the average size of a vein is 5.8 but a lot is lost in bedrock, in particular, veins starting at y=0 won't ever generate anything at all because they only "grow" downwards from their start, which can also be one layer below the initial point so this also applies to some veins at y=1 (conversely, this is why diamond is so rare at y=15 despite an otherwise uniform distribution of initial starting points).
Otherwise, I see no real evidence that diamond has been made rarer, with the 1.13 snapshots being the only time an actual decline occurred (which was supposed to have been fixed), if not based on actual analysis since I don't know of any such tools that work in current versions but I'd think it would be all over the place if diamond were all of a sudden much rarer.
Also, based on previous observations it appears that Mojang did remove the "offset by 8" thing so decorations are now centered within a 3x3 chunk area instead of 2x2 (the main difference as far as world generation is concerned is that features can now extend up to 16 blocks from their center instead of just 8, though I don't know of any vanilla features, excluding structures, which use a different generation method which places them chunk-bu-chunk, that might take advantage of this, outside of mods; for example, the largest trees in TMCW require a reduced random offset so they don't spill into unloaded chunks, causing runaway world generation when they are loaded and in turn decorated. I've actually thought of changing this myself due to this; this change also greatly benefits modders, many of whom do not add the offset of 8).
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?
I hadn't but I have now checked the 10X10 chunk area I did first.
I found:
0 2 4 1 3 0 1 3 5 8
0 7 1 8 1 0 0 5 3 2
2 5 4 5 5 6 6 6 6 4
1 4 0 0 0 0 2 0 4 4
7 4 5 0 2 2 6 0 8 6
0 4 4 4 4 1 6 4 7 0
6 5 4 4 4 5 1 8 0 4
5 2 0 3 5 0 5 3 1 1
3 4 4 3 0 5 4 0 4 0
8 8 1 9 3 5 5 0 0 4
ores per chunk (running in a serpentine pattern, so the horizontal repeats weren't actually next to each other.)
which I make as 328 ores in 78 chunks out of 100.
That's 3.28 ores per chunk and approximately 4.3 ores per vein using the 76 veins I counted last time.
The numbers are a tad inconsistent sounding since the 76 veins counted veins that I figured "belonged to" the chunks but the ores are strictly ores that are in the chunks in question, so there are a few veins that are split between chunks and a few ores intruding into the area from veins that "belonged to" chunks outside the area.
Am I correct in assuming that, with no shifting of the ores between chunks, a diamond vein can only extend a single block outside its chunk? (That's a blocks distance, not just a single ore block.) And that therefore any merging of veins can only occur along the chunk borders?
Just testing.
Yes, the only difference from before is where the "chunk borders" are, now they are actual chunks instead of "offset chunks".
Also, 3.28 ores per chunk isn't inconsistent with an average of 3.7 since there will be variance over a relatively small area, and it suggests that MC-126373 had actually been fixed (in the last comment somebody found an average of only 2.26 ore per chunk compared to 3.86 for the same area in 1.12.2. However, the former was supposedly in 1.13.1, after the bug has been marked as fixed, and while ore density does vary over relatively small areas 2.26 seems too low, so I'm not actually sure).
For comparison, I used MCEdit to analyze 10 different 10x10 chunk areas of a world created in 1.6.4 and found 307, 332, 287, 326, 376, 267, 337, 329, 317, and 289 diamond ore, an average of 3.167 per chunk, compared to the Wiki's old figure of 3.097 per chunk (not sure what version this was from but ore generation was unchanged from Beta 1.8 to 1.6.4, with a slight increase expected in 1.7 due to less dirt/gravel/caves and a greater increase in 1.8 due to an increase in average vein size).
The variation from the average was -15.7 to +18.7%, and a similar variation from an average of 3.7 gives a range of 3.12 to 4.39, or as low as 3.00 if a range of +/-18.7% is assumed. I'd expect that 1.7+ worlds have less variation due to less variation in cave density, although I found only 263 diamond in an area with very few caves and 341 in an area with a lot so this isn't the only factor.
Also, according to this report vein sizes 1-3 now generate ores, while before they would result in nothing, suggesting a changes to the code that generates veins of ore. The numbers given are otherwise consistent though; a size of 6 averages about 4.4 ore per vein in either case (1 million simulated veins using 1.6.4's/pre-1.13's ore generation code vs the results in the link for four 40x40 chunk areas at one vein per chunk), as does a size of 5 (about 3.6) and 4 (about 2.4) so whatever change Mojang made only affects the smallest sizes (3 is only about 0.46 in 1.6.4 but 2.65 in 1.13), and it would have no impact since the smallest vein size used in vanilla was 6 for lapis prior to 1.8 (ancient debris may be smaller).
Also, somebody (recently?) edited all the sizes given on the Wiki to range from 1 to the number in the code, which is only loosely related to the average size of a vein:
Not even close - dirt actually averages about 112 blocks per vein, as anybody who has ever dug one out would know, and similarly coal averages about 22 with 20, 22, and 24 being the most common sizes; only sizes" 12-13 closely match the average with smaller sizes being smaller and larger sizes larger:
Ores: 8; count: 5
Ores: 9; count: 6
Ores: 10; count: 52
Ores: 11; count: 109
Ores: 12; count: 757
Ores: 13; count: 664
Ores: 14; count: 4017
Ores: 15; count: 2927
Ores: 16; count: 21408
Ores: 17; count: 10040
Ores: 18; count: 65499
Ores: 19; count: 24672
Ores: 20; count: 175424
Ores: 21; count: 46353
Ores: 22; count: 133477
Ores: 23; count: 76266
Ores: 24; count: 146654
Ores: 25; count: 85366
Ores: 26; count: 87464
Ores: 27; count: 44552
Ores: 28; count: 40018
Ores: 29; count: 15718
Ores: 30; count: 10587
Ores: 31; count: 3975
Ores: 32; count: 2844
Ores: 33; count: 742
Ores: 34; count: 214
Ores: 35; count: 141
Ores: 36; count: 35
Ores: 37; count: 7
Ores: 38; count: 7
Average vein size: 22.7306
Dirt ("size" 33):
Ores: 50; count: 1
Ores: 51; count: 0
Ores: 52; count: 0
Ores: 53; count: 0
Ores: 54; count: 1
Ores: 55; count: 0
Ores: 56; count: 2
Ores: 57; count: 0
Ores: 58; count: 1
Ores: 59; count: 4
Ores: 60; count: 7
Ores: 61; count: 3
Ores: 62; count: 8
Ores: 63; count: 6
Ores: 64; count: 18
Ores: 65; count: 8
Ores: 66; count: 28
Ores: 67; count: 14
Ores: 68; count: 60
Ores: 69; count: 34
Ores: 70; count: 97
Ores: 71; count: 61
Ores: 72; count: 145
Ores: 73; count: 98
Ores: 74; count: 233
Ores: 75; count: 147
Ores: 76; count: 338
Ores: 77; count: 244
Ores: 78; count: 541
Ores: 79; count: 367
Ores: 80; count: 817
Ores: 81; count: 515
Ores: 82; count: 1159
Ores: 83; count: 798
Ores: 84; count: 1758
Ores: 85; count: 1120
Ores: 86; count: 2396
Ores: 87; count: 1542
Ores: 88; count: 3509
Ores: 89; count: 2180
Ores: 90; count: 4637
Ores: 91; count: 3017
Ores: 92; count: 6636
Ores: 93; count: 3930
Ores: 94; count: 8698
Ores: 95; count: 5078
Ores: 96; count: 11956
Ores: 97; count: 6684
Ores: 98; count: 15318
Ores: 99; count: 8662
Ores: 100; count: 19997
Ores: 101; count: 10626
Ores: 102; count: 25035
Ores: 103; count: 12993
Ores: 104; count: 31459
Ores: 105; count: 15822
Ores: 106; count: 37316
Ores: 107; count: 18440
Ores: 108; count: 45026
Ores: 109; count: 21395
Ores: 110; count: 49930
Ores: 111; count: 23543
Ores: 112; count: 55967
Ores: 113; count: 25212
Ores: 114; count: 57124
Ores: 115; count: 26267
Ores: 116; count: 57668
Ores: 117; count: 26224
Ores: 118; count: 53408
Ores: 119; count: 25095
Ores: 120; count: 48439
Ores: 121; count: 22592
Ores: 122; count: 39188
Ores: 123; count: 19074
Ores: 124; count: 31119
Ores: 125; count: 15769
Ores: 126; count: 21965
Ores: 127; count: 12082
Ores: 128; count: 15342
Ores: 129; count: 8751
Ores: 130; count: 9476
Ores: 131; count: 5817
Ores: 132; count: 5563
Ores: 133; count: 3671
Ores: 134; count: 3095
Ores: 135; count: 2034
Ores: 136; count: 1592
Ores: 137; count: 1025
Ores: 138; count: 771
Ores: 139; count: 456
Ores: 140; count: 298
Ores: 141; count: 198
Ores: 142; count: 110
Ores: 143; count: 61
Ores: 144; count: 52
Ores: 145; count: 18
Ores: 146; count: 9
Ores: 147; count: 3
Ores: 148; count: 3
Ores: 149; count: 2
Ores: 150; count: 1
Ores: 151; count: 1
The reason for this discrepancy, and confusion, is because there is indeed a loop that iterates (size) number of times but each iteration places a roughly spherical blob of blocks each of which can be more than one block in size, with a total of 4 loops involved; for diamond, size 8, the average number of iterations in the innermost loop is 13.7 while for coal, size 17, it is 75.5, and dirt, size 33, is 496.
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?