Minecraft worlds are random, it could be from one to one every block on the surface
It is impossible for there to be villages everywhere - the game makes one attempt per 512x512 block region, choosing a single point within that area, which only succeeds if the biome happens to be one that allows villages to generate. Even in Superflat you can only reduce the distance to 9 (one village per 144x144 block area).
The same goes for most other "random" world generation - it only seems to be random (even less so than a pseudorandom number generator, which is technically what the game uses, would appear to be; a true RNG would not be useful since then there is no way generate a world in a predictable manner; it would just be a jumbled mess; which is also why you can recreate a world with just its seed, which is combined with chunk coordinates to set a seed for each chunk, or each region in the case of villages); for example, the game makes exactly one attempt at generating a vein of diamond ore per chunk, so it is impossible for there to be more than 10 ore (the maximum size of a vein, which is constrained by the ore generator) per chunk, even with certain seeds that break world generation (the seed "102496288339226" causes chunks along diagonals to have the same ores).
Likewise, it is impossible to have a world with only one biome and no villages if the biome doesn't have them unless you use Buffet, since the probability of every one of 225 trillion points (the biome generator has an underlying resolution of 4x4 blocks) is too astronomically low for a PRNG that has only 2^64 possible states (the generation of biomes for each point also involves a few dozen RNG calls, further reducing the probability). As far as I know, based on the code, it is impossible for there to be a seed which breaks the biome generator so only one biome generates (the way the biome generator sets the seed makes it impossible for biomes to repeat in any manner; even the "repeating ores/caves" seeds only affect chunks in rows and there is still normal variation between rows).
I'm not sure what the frequency of villages is since you also need to know the frequency of the biomes villages generate in but I've found 12 in my first world, which was 109595 chunks at last count for an average of one every 9133 chunks; this is in 1.6.4 though (two out of 7 common biomes can have villages; excluding Ice Plains regions and oceans they all have the same probabilities, though the small forests in plains make them less common. Also, prior to 1.10 a significant fraction of villages fail to generate because the biomes around the entire village, not just at a single point, had to be valid so they may very well be more common now, especially when you consider that about 3/4 of a 1.6.4 world is ocean while 1.7+ is the opposite). You can get a good estimate by using a tool like AMIDST and counting the number of villages within a large area (large meaning on a scale like this map; you want it to be large enough to average out variations due to climate zones).
It is impossible for there to be villages everywhere - the game makes one attempt per 512x512 block region, choosing a single point within that area, which only succeeds if the biome happens to be one that allows villages to generate. Even in Superflat you can only reduce the distance to 9 (one village per 144x144 block area).
The same goes for most other "random" world generation - it only seems to be random (even less so than a pseudorandom number generator, which is technically what the game uses, would appear to be; a true RNG would not be useful since then there is no way generate a world in a predictable manner; it would just be a jumbled mess; which is also why you can recreate a world with just its seed, which is combined with chunk coordinates to set a seed for each chunk, or each region in the case of villages); for example, the game makes exactly one attempt at generating a vein of diamond ore per chunk, so it is impossible for there to be more than 10 ore (the maximum size of a vein, which is constrained by the ore generator) per chunk, even with certain seeds that break world generation (the seed "102496288339226" causes chunks along diagonals to have the same ores).
Likewise, it is impossible to have a world with only one biome and no villages if the biome doesn't have them unless you use Buffet, since the probability of every one of 225 trillion points (the biome generator has an underlying resolution of 4x4 blocks) is too astronomically low for a PRNG that has only 2^64 possible states (the generation of biomes for each point also involves a few dozen RNG calls, further reducing the probability). As far as I know, based on the code, it is impossible for there to be a seed which breaks the biome generator so only one biome generates (the way the biome generator sets the seed makes it impossible for biomes to repeat in any manner; even the "repeating ores/caves" seeds only affect chunks in rows and there is still normal variation between rows).
I'm not sure what the frequency of villages is since you also need to know the frequency of the biomes villages generate in but I've found 12 in my first world, which was 109595 chunks at last count for an average of one every 9133 chunks; this is in 1.6.4 though (two out of 7 common biomes can have villages; excluding Ice Plains regions and oceans they all have the same probabilities, though the small forests in plains make them less common. Also, prior to 1.10 a significant fraction of villages fail to generate because the biomes around the entire village, not just at a single point, had to be valid so they may very well be more common now, especially when you consider that about 3/4 of a 1.6.4 world is ocean while 1.7+ is the opposite). You can get a good estimate by using a tool like AMIDST and counting the number of villages within a large area (large meaning on a scale like this map; you want it to be large enough to average out variations due to climate zones).
Now, that was worded wrong from me... i meant like if you were to tweak the world gen (removed in 1.12 big rip ;( ) you can make it anything. And your points very valid. There is a limit if you dont go to the strange world of mods
I did some digging in Amidst for 1.15.2, counting villages across areas of world about 90 million square blocks in area. I found that on average, there were about 60 villages for each of these areas. So, taking 90,000,000 / 60 gives about one village per 1.5 million square blocks. Taking (60,000,000)^2 / 1,500,000 I get about 2.4 billion villages in the world.
In addition, using data from one of my own survival worlds in which I am trying to reach the world border, I have generated about 3GB of world data, and have kept track of how many structures I have encountered. I have found 63 villages, and extrapolating that data to the size of an entire Minecraft world generated (which is estimated at about 100,000,000 GB), I find that I would get about 2.1 billion villages. So I think it's probably safe to say that there are about 2-3 billion villages total.
Just out of curiosity, on average, how many villages do you think are in a regular Minecraft Java Edition world?
Assuming, for example, there is one village for every 1000x1000 block area, that would mean there would be around 3.6 billion villages!
Minecraft worlds are random, it could be from one to one every block on the surface
“Bees are the best addition to the game”
- TotallyNotThomas, 2021
It is impossible for there to be villages everywhere - the game makes one attempt per 512x512 block region, choosing a single point within that area, which only succeeds if the biome happens to be one that allows villages to generate. Even in Superflat you can only reduce the distance to 9 (one village per 144x144 block area).
The same goes for most other "random" world generation - it only seems to be random (even less so than a pseudorandom number generator, which is technically what the game uses, would appear to be; a true RNG would not be useful since then there is no way generate a world in a predictable manner; it would just be a jumbled mess; which is also why you can recreate a world with just its seed, which is combined with chunk coordinates to set a seed for each chunk, or each region in the case of villages); for example, the game makes exactly one attempt at generating a vein of diamond ore per chunk, so it is impossible for there to be more than 10 ore (the maximum size of a vein, which is constrained by the ore generator) per chunk, even with certain seeds that break world generation (the seed "102496288339226" causes chunks along diagonals to have the same ores).
Likewise, it is impossible to have a world with only one biome and no villages if the biome doesn't have them unless you use Buffet, since the probability of every one of 225 trillion points (the biome generator has an underlying resolution of 4x4 blocks) is too astronomically low for a PRNG that has only 2^64 possible states (the generation of biomes for each point also involves a few dozen RNG calls, further reducing the probability). As far as I know, based on the code, it is impossible for there to be a seed which breaks the biome generator so only one biome generates (the way the biome generator sets the seed makes it impossible for biomes to repeat in any manner; even the "repeating ores/caves" seeds only affect chunks in rows and there is still normal variation between rows).
I'm not sure what the frequency of villages is since you also need to know the frequency of the biomes villages generate in but I've found 12 in my first world, which was 109595 chunks at last count for an average of one every 9133 chunks; this is in 1.6.4 though (two out of 7 common biomes can have villages; excluding Ice Plains regions and oceans they all have the same probabilities, though the small forests in plains make them less common. Also, prior to 1.10 a significant fraction of villages fail to generate because the biomes around the entire village, not just at a single point, had to be valid so they may very well be more common now, especially when you consider that about 3/4 of a 1.6.4 world is ocean while 1.7+ is the opposite). You can get a good estimate by using a tool like AMIDST and counting the number of villages within a large area (large meaning on a scale like this map; you want it to be large enough to average out variations due to climate zones).
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?
Now, that was worded wrong from me... i meant like if you were to tweak the world gen (removed in 1.12 big rip ;( ) you can make it anything. And your points very valid. There is a limit if you dont go to the strange world of mods
“Bees are the best addition to the game”
- TotallyNotThomas, 2021
No matter how many there are, I will pillage them all. I need all the haybales. All of them.
People are so dramatic. "DoN't DiG sTrAiGhT dOwN" They say. I'll dig straight down if i want to dig straight down.
What's the worst that could happ-
Treedude13 Tried to swim in lava
I did some digging in Amidst for 1.15.2, counting villages across areas of world about 90 million square blocks in area. I found that on average, there were about 60 villages for each of these areas. So, taking 90,000,000 / 60 gives about one village per 1.5 million square blocks. Taking (60,000,000)^2 / 1,500,000 I get about 2.4 billion villages in the world.
In addition, using data from one of my own survival worlds in which I am trying to reach the world border, I have generated about 3GB of world data, and have kept track of how many structures I have encountered. I have found 63 villages, and extrapolating that data to the size of an entire Minecraft world generated (which is estimated at about 100,000,000 GB), I find that I would get about 2.1 billion villages. So I think it's probably safe to say that there are about 2-3 billion villages total.
So good luck pillaging them all.