There have been more votes cast since I ended the experiment, and I can see that my noted trends are not slowing down. There are STILL people identifying the Beta terrain as being from 1.3.1, more than there are identifying 1.3.1 from being from 1.3.1.
You can't argue with a poll, people. If you do not agree with your finding, I suggest you make your own poll, gather your own data, and back your arguments with genuine evidence instead of saying that the game was rigged because you lost.
That's just immature.
Rollback Post to RevisionRollBack
The Internet is a big place, friend. I've been places you've n͍̺e̩v̦e̦̰͍͓̩ͅr̜̭̝̬̬͉̤̬ ͙ịm̖͇a͍͇̤͙̥g̤̘i͔͖̤̼̪̬n͖͔̳̬̯e̩̘ḓ͈͔̠̙͇̼̯.͎
There have been more votes cast since I ended the experiment, and I can see that my noted trends are not slowing down. There are STILL people identifying the Beta terrain as being from 1.3.1, more than there are identifying 1.3.1 from being from 1.3.1.
You can't argue with a poll, people. If you do not agree with your finding, I suggest you make your own poll, gather your own data, and back your arguments with genuine evidence instead of saying that the game was rigged because you lost.
That's just immature.
To be fair, the sample size IS kinda small. Most complaints towards the new terrain generator are in regard to its potential for variety, compared to the older one. Such a fact would only be obvious with a much larger sample size.
But I must say the current generator's getting better all the time, what with smaller features being added incrementally. My biggest problem was the dull grass (which I fixed myself with a custom grasscolor.png).
A single screenshot just doesn't do the terrain justice, especially given the low chance of actually spawning near something cool. The spawn point is never going to be as good as the rest of the world, and hand-picking what you want to photograph messes with the experiment. This makes a proper experiment very difficult.
I think it would really be better to show a generated image (using a world viewing program) of several chunks worth of terrain. Unfortunately you're not going to be able to do that without running into a jungle. This means you're going to have to cut out random parts of the terrain.
As much as I enjoyed this experiment, I have to say it does not change anything about the previous comparisions between the terrain generation of alpha, beta and the latest terrain generation.
First of all, the biomes. The fact you used a non-biome shading texture really changes everthing. Why? Well, one of the biggest complaints about the new generator is the fact it generates the height difference together with the biomes, and the biomes are not organized in any way. That means that all huge mountains are bluish-green, and do not have snow, or are green and filled with jungle trees. While the way the mountains themselves was also changed, it is generally not considered a problem, other than how cliffs no longer generate above the sea, which is not demonstrated in those pictures. Both the old and the new generator could make those kind of mountain ranges, the difference is that on the new one all mountains are like that, while the old one had variety. This can not be displayed by pictures.
The beta generator used a temperature and an humidity gradient, randomly placed from seeds. This made smooth transitions of color, and the biomes were generated separately from the altitude, allowing for huge mountains on all biomes, not exclusively bare ones and ones with jungle trees. It also had the effect of making the color changes of the grass smooth, and not rough like the new one.
Another missing part in this is the underwater bedding. The fact the beta generator did not have a pattern on the seafloor actually makes this hard to see, and that is why it was great! The old one felt more natural, so it is normal that people would not notice those details. You didn't think "Oh, this ocean is pretty because it have a smooth floor, and proper beaches". You didn't even consider the possibility of it having a pattern in the floor, because such pattern didn't exist yet.
The beaches are another problem. While the new generator do generate pretty beaches ocasionally, it once in a while drop you a patch of sand on grass, making an awkward presentation. The beta generator, instead, would have smoother beaches, with a few occasional flaws. This is the same as the underwater bedding, in that nobody really noticed it at time because they did not see how it could go bad. For the last pic, you also apparently choose a version on which the chunk errors were still present, which can be seen by the square formation on the cliffs.
By selecting only areas that could be confused, you ended up acidentally hiding those details.This is in a way similar to comparing microsoft paint and adobe photoshop by how well they can draw lines, rough brushes and text. The old generator had things that were lost in the new one, and the new one had good formations too, but also have awkward terrain. The best would really be a new generator, that was not rushed like the last one, and on which the developers actually paid attention to the community, not just reading the top tweets and petitions. The beta generator was not without it's flaws; a major part of the playerbase were supporting a terrain generation overhaul because of the problems of the swamps, and how the old generator didn't properly allow for new biomes, or features like volcanos (Which were promised by notch since around the halloween update, back on 2010). The community just said yes out of the hype generated by screenshots of swamps, and new features, and ignored the way other things were being lost in the process. That is to be expected, as we don't really value something until we see we can lose it.
The terrain generation of minecraft still have a long way to improve. There are many ideas frequently sugested by the players, and mojang itself should stop for a second to think about what minecraft really means as a game, and not be driven away by the popularity minecraft have, and what the players think (although if they are not able to take the right choices, I believe the ideas would actually be better in the hand of the player community, if it can properly get organized; just look at how many astonishing mods have been produced).
I really hope people really consider those points, and break off the bias that mojang always takes the right decisions, and what is new is always better. Thank you.
A single screenshot just doesn't do the terrain justice, especially given the low chance of actually spawning near something cool. The spawn point is never going to be as good as the rest of the world, and hand-picking what you want to photograph messes with the experiment. This makes a proper experiment very difficult.
I think it would really be better to show a generated image (using a world viewing program) of several chunks worth of terrain. Unfortunately you're not going to be able to do that without running into a jungle. This means you're going to have to cut out random parts of the terrain.
You are simply ignorant and refuse to read anything I write. For the very last time, I did not select ANYTHING. I made ONE SINGLE MAP PER VERSION WITH A RANDOM SEED FOR EACH AND TOOK A PICTURE OF WHERE I SPAWNED.
Sure, I could have "cherry picked" like you people keep saying, but I was honest and scientific about everything. I suppose you would have rathered that I used the "gargamel" seed for Beta, and a plains seed for 1.3.1? I wouldn't be surprised if that is what you would have done had you taken the time, effort, and energy to do an experiment rather than base your arguments strictly on ANECDOTAL EVIDENCE. But you didn't. If you truly had scientific proof that Beta and Alpha had better terrain, or even that Beta and 1.3.1 terrain were noticeably different, you would have done what I did and proven your argument logically.
But no. You didn't try to back your argument with scientific fact, largely because you knew that scientific testing would probably disprove your shaky arguments and anecdotal evidence. Instead, you find it easier to simply attack logic with anecdotes and opinions. The science suggests that you are wrong, so you nitpick and claim that the scientist lied.
I quit.
I have been pouring an enormous amount of time, effort, and energy into doing these experiments and polls for the good of this community. My only reward has been ignorant forumites who ignore EVERY LOGICAL ARGUMENT and say "you are stoopid and wrong cuz you are stoopid and wrongz".
I give up. IF I attempt to contribute meaningfully to this community again, it will be a long time from now. Everyone is just too wrapped up in their petty opinions (which they state as facts) to stop and consider when, how, and why science and logic conflicts with what they say. Instead they just attack it like spoiled preschoolers who can't have a turn with a toy.
Congratulations, terrain Nazis! Your ignorance has once again beat down science, logic, and reason once again, just like any true cult would. I hope you are proud of that.
Rollback Post to RevisionRollBack
The Internet is a big place, friend. I've been places you've n͍̺e̩v̦e̦̰͍͓̩ͅr̜̭̝̬̬͉̤̬ ͙ịm̖͇a͍͇̤͙̥g̤̘i͔͖̤̼̪̬n͖͔̳̬̯e̩̘ḓ͈͔̠̙͇̼̯.͎
And the very changes a lot of people are considering will help improve the terrain, leaving the current features (including the mountains spammed together, the infinite caves, and maybe the crappy beaches) able to exist. Most of us just want things to return. You can't deny the generator needs improvement. Cherry-picking or not, the current generator still cannot do things that the previous generator did, things that we loved. And that's that.
People who hold the feedback to a halt are the terrain nazis. With the changes I and many other people are considering, no one is going to be pissed off.
Again, I don't want anything taken away -- just more possibilities able to come to life.
You are simply ignorant and refuse to read anything I write. For the very last time, I did not select ANYTHING. I made ONE SINGLE MAP PER VERSION WITH A RANDOM SEED FOR EACH AND TOOK A PICTURE OF WHERE I SPAWNED.
I didn't mean to imply you did anything on purpose. The fact is that you mentioned some of the seeds spawned you in water, which means you had to have made a choice in which land you headed towards, which means you could have subconciously gravitated towards a certain piece of terrain without realizing it. This becomes a non-issue if you simply capture a larger amount of the world's terrain.
Sure, I could have "cherry picked" like you people keep saying, but I was honest and scientific about everything. I suppose you would have rathered that I used the "gargamel" seed for Beta, and a plains seed for 1.3.1? I wouldn't be surprised if that is what you would have done had you taken the time, effort, and energy to do an experiment rather than base your arguments strictly on ANECDOTAL EVIDENCE. But you didn't. If you truly had scientific proof that Beta and Alpha had better terrain, or even that Beta and 1.3.1 terrain were noticeably different, you would have done what I did and proven your argument logically.
What I would've done is used random seeds for each world (no retries, lots of samples,) then explored a wide area around the spawn on each. I would then render 3-D views of each world using one of the many available world mappers. I would then censor anything that gives away the version (gravel beaches, jungles, etc.) It might be necessary to also censor random areas on each world so that people can't guess what the censored regions are.
The uncensored versions would be available at the conclusion of the experiment (to prove nothing was censored unjustly.) Specific terrain generator version (not just "Beta", etc.) and world seeds would also be listed (in case skeptics wanted to verify the seed, or if people were just curious.) It's a little harder to prove worlds weren't skipped. I suppose a video of the process would be possible, if absolutely necessary (or maybe the seeds could just be fixed numbers in an ascending order so that you'd know if they were missing.)
And that, ladies and gentlemen, is how you do an experiment from behind (i.e. the anal way.)
But no. You didn't try to back your argument with scientific fact, largely because you knew that scientific testing would probably disprove your shaky arguments and anecdotal evidence. Instead, you find it easier to simply attack logic with anecdotes and opinions. The science suggests that you are wrong, so you nitpick and claim that the scientist lied.
Or maybe because I couldn't be assed to carry out the experiment myself, especially with my waning interest in the game (besides testing out the new features.) I'm hoping ChaosGuardian or someone else will carry out the modified experiment in another thread.
I have been pouring an enormous amount of time, effort, and energy into doing these experiments and polls for the good of this community. My only reward has been ignorant forumites who ignore EVERY LOGICAL ARGUMENT and say "you are stoopid and wrong cuz you are stoopid and wrongz".
I give up. IF I attempt to contribute meaningfully to this community again, it will be a long time from now. Everyone is just too wrapped up in their petty opinions (which they state as facts) to stop and consider when, how, and why science and logic conflicts with what they say. Instead they just attack it like spoiled preschoolers who can't have a turn with a toy.
I'm sorry to hear that and I appreciate the work you put into this. That doesn't stop me from criticizing your experimental methods.
Congratulations, terrain Nazis! Your ignorance has once again beat down science, logic, and reason once again, just like any true cult would. I hope you are proud of that.
While I do believe that you did this legitimately, without faking the results. I'm curious about this image.
It looks like an Extreme Hills biome (they also remind me of alpha, minus the green grass and trees), although I'm yet to find an Extreme Hills biome with so many trees scattered around. Also, noticed the underwater terrain is smooth and missing the blotchy pattern of sand, dirt and clay.
I haven't played much of 1.3.1, so maybe Mojang have improved on this. If you still happen to have the seed, would you mind posting it please?
While I do believe that you did this legitimately, without faking the results. I'm curious about this image.
It looks like an Extreme Hills biome (they also remind me of alpha, minus the green grass and trees), although I'm yet to find an Extreme Hills biome with so many trees scattered around. Also, noticed the underwater terrain is smooth and missing the blotchy pattern of sand, dirt and clay.
I haven't played much of 1.3.1, so maybe Mojang have improved on this. If you still happen to have the seed, would you mind posting it please?
Cheers
Looks like ForestHills maybe. Problem is that all height changes are spammed together. I would rather have random height variation in forests themselves.
Both of the 1.3 maps don't really look legitimate. The sea floor from image 1 is made of dirt only, and sea floors don't generate like that anymore. The biomes from both of the 1.3 images don't look legitimate either. The only biome in the game with that tree density is the swamp biome, which that obviously isn't, and the grass color looks almost the same as it is in the forest biome.
I have noticed that the terrain is slightly more varied than it was before though, so I could be wrong. If you want to prove me wrong, just post the seed and coordinates for each of the 1.3 maps.
While at first I thought the experiment was valid, Hayarotle's post convinced me otherwise and ended this thread WITH LOGICAL, REASONABLE ARGUMENTS.
Cadika is proudly stating that his experiment was scientific, and he was doing science... but forgets that science is based on repetition and multiple test samples.
I believe that deep inside his subcounsciouss mind he knows that because he ignored every post that asked for VIDEO samples instead of SINGLE FRAMED IMAGES.
It's easy to be mistaken trying to identify framed images that lack the same various features that make beta and 1.3 terrain so different, and the same features that are asked by hundreds of people to be re-inserted in the game.
This features were listed by Hayarotle on his excelent post, that has yet to be replied by Cadika. But I guess after that post he/she realised the mistakes of the experiment and decided to ragequit this thread to avoid admitting the experiment was biased.
There have been more votes cast since I ended the experiment, and I can see that my noted trends are not slowing down. There are STILL people identifying the Beta terrain as being from 1.3.1, more than there are identifying 1.3.1 from being from 1.3.1.
You can't argue with a poll, people. If you do not agree with your finding, I suggest you make your own poll, gather your own data, and back your arguments with genuine evidence instead of saying that the game was rigged because you lost.
That's just immature.
To be fair, the sample size IS kinda small. Most complaints towards the new terrain generator are in regard to its potential for variety, compared to the older one. Such a fact would only be obvious with a much larger sample size.
But I must say the current generator's getting better all the time, what with smaller features being added incrementally. My biggest problem was the dull grass (which I fixed myself with a custom grasscolor.png).
I think it would really be better to show a generated image (using a world viewing program) of several chunks worth of terrain. Unfortunately you're not going to be able to do that without running into a jungle. This means you're going to have to cut out random parts of the terrain.
Mostly moved on. May check back a few times a year.
Not really. Alpha and Beta trees look extremely different compared to 1.3's constant and boring mini pyramid shape.
First of all, the biomes. The fact you used a non-biome shading texture really changes everthing. Why? Well, one of the biggest complaints about the new generator is the fact it generates the height difference together with the biomes, and the biomes are not organized in any way. That means that all huge mountains are bluish-green, and do not have snow, or are green and filled with jungle trees. While the way the mountains themselves was also changed, it is generally not considered a problem, other than how cliffs no longer generate above the sea, which is not demonstrated in those pictures. Both the old and the new generator could make those kind of mountain ranges, the difference is that on the new one all mountains are like that, while the old one had variety. This can not be displayed by pictures.
The beta generator used a temperature and an humidity gradient, randomly placed from seeds. This made smooth transitions of color, and the biomes were generated separately from the altitude, allowing for huge mountains on all biomes, not exclusively bare ones and ones with jungle trees. It also had the effect of making the color changes of the grass smooth, and not rough like the new one.
Another missing part in this is the underwater bedding. The fact the beta generator did not have a pattern on the seafloor actually makes this hard to see, and that is why it was great! The old one felt more natural, so it is normal that people would not notice those details. You didn't think "Oh, this ocean is pretty because it have a smooth floor, and proper beaches". You didn't even consider the possibility of it having a pattern in the floor, because such pattern didn't exist yet.
The beaches are another problem. While the new generator do generate pretty beaches ocasionally, it once in a while drop you a patch of sand on grass, making an awkward presentation. The beta generator, instead, would have smoother beaches, with a few occasional flaws. This is the same as the underwater bedding, in that nobody really noticed it at time because they did not see how it could go bad. For the last pic, you also apparently choose a version on which the chunk errors were still present, which can be seen by the square formation on the cliffs.
By selecting only areas that could be confused, you ended up acidentally hiding those details.This is in a way similar to comparing microsoft paint and adobe photoshop by how well they can draw lines, rough brushes and text. The old generator had things that were lost in the new one, and the new one had good formations too, but also have awkward terrain. The best would really be a new generator, that was not rushed like the last one, and on which the developers actually paid attention to the community, not just reading the top tweets and petitions. The beta generator was not without it's flaws; a major part of the playerbase were supporting a terrain generation overhaul because of the problems of the swamps, and how the old generator didn't properly allow for new biomes, or features like volcanos (Which were promised by notch since around the halloween update, back on 2010). The community just said yes out of the hype generated by screenshots of swamps, and new features, and ignored the way other things were being lost in the process. That is to be expected, as we don't really value something until we see we can lose it.
The terrain generation of minecraft still have a long way to improve. There are many ideas frequently sugested by the players, and mojang itself should stop for a second to think about what minecraft really means as a game, and not be driven away by the popularity minecraft have, and what the players think (although if they are not able to take the right choices, I believe the ideas would actually be better in the hand of the player community, if it can properly get organized; just look at how many astonishing mods have been produced).
I really hope people really consider those points, and break off the bias that mojang always takes the right decisions, and what is new is always better. Thank you.
You are simply ignorant and refuse to read anything I write. For the very last time, I did not select ANYTHING. I made ONE SINGLE MAP PER VERSION WITH A RANDOM SEED FOR EACH AND TOOK A PICTURE OF WHERE I SPAWNED.
Sure, I could have "cherry picked" like you people keep saying, but I was honest and scientific about everything. I suppose you would have rathered that I used the "gargamel" seed for Beta, and a plains seed for 1.3.1? I wouldn't be surprised if that is what you would have done had you taken the time, effort, and energy to do an experiment rather than base your arguments strictly on ANECDOTAL EVIDENCE. But you didn't. If you truly had scientific proof that Beta and Alpha had better terrain, or even that Beta and 1.3.1 terrain were noticeably different, you would have done what I did and proven your argument logically.
But no. You didn't try to back your argument with scientific fact, largely because you knew that scientific testing would probably disprove your shaky arguments and anecdotal evidence. Instead, you find it easier to simply attack logic with anecdotes and opinions. The science suggests that you are wrong, so you nitpick and claim that the scientist lied.
I quit.
I have been pouring an enormous amount of time, effort, and energy into doing these experiments and polls for the good of this community. My only reward has been ignorant forumites who ignore EVERY LOGICAL ARGUMENT and say "you are stoopid and wrong cuz you are stoopid and wrongz".
I give up. IF I attempt to contribute meaningfully to this community again, it will be a long time from now. Everyone is just too wrapped up in their petty opinions (which they state as facts) to stop and consider when, how, and why science and logic conflicts with what they say. Instead they just attack it like spoiled preschoolers who can't have a turn with a toy.
Congratulations, terrain Nazis! Your ignorance has once again beat down science, logic, and reason once again, just like any true cult would. I hope you are proud of that.
And the very changes a lot of people are considering will help improve the terrain, leaving the current features (including the mountains spammed together, the infinite caves, and maybe the crappy beaches) able to exist. Most of us just want things to return. You can't deny the generator needs improvement. Cherry-picking or not, the current generator still cannot do things that the previous generator did, things that we loved. And that's that.
People who hold the feedback to a halt are the terrain nazis. With the changes I and many other people are considering, no one is going to be pissed off.
Again, I don't want anything taken away -- just more possibilities able to come to life.
What I would've done is used random seeds for each world (no retries, lots of samples,) then explored a wide area around the spawn on each. I would then render 3-D views of each world using one of the many available world mappers. I would then censor anything that gives away the version (gravel beaches, jungles, etc.) It might be necessary to also censor random areas on each world so that people can't guess what the censored regions are.
The uncensored versions would be available at the conclusion of the experiment (to prove nothing was censored unjustly.) Specific terrain generator version (not just "Beta", etc.) and world seeds would also be listed (in case skeptics wanted to verify the seed, or if people were just curious.) It's a little harder to prove worlds weren't skipped. I suppose a video of the process would be possible, if absolutely necessary (or maybe the seeds could just be fixed numbers in an ascending order so that you'd know if they were missing.)
And that, ladies and gentlemen, is how you do an experiment from behind (i.e. the anal way.)
Or maybe because I couldn't be assed to carry out the experiment myself, especially with my waning interest in the game (besides testing out the new features.) I'm hoping ChaosGuardian or someone else will carry out the modified experiment in another thread.
I'm sorry to hear that and I appreciate the work you put into this. That doesn't stop me from criticizing your experimental methods.
Okay...
Mostly moved on. May check back a few times a year.
http://img192.imageshack.us/img192/9819/20111023232636.png
It looks like an Extreme Hills biome (they also remind me of alpha, minus the green grass and trees), although I'm yet to find an Extreme Hills biome with so many trees scattered around. Also, noticed the underwater terrain is smooth and missing the blotchy pattern of sand, dirt and clay.
I haven't played much of 1.3.1, so maybe Mojang have improved on this. If you still happen to have the seed, would you mind posting it please?
Cheers
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumIt clearly contains tall grass...
Also what specific versions are we talking about? The "Beta" label is a bit vague.
Looks like ForestHills maybe. Problem is that all height changes are spammed together. I would rather have random height variation in forests themselves.
Mostly moved on. May check back a few times a year.
ANARKY, please give me that seed!
- The Cubic Chunks Mod is back! Be a part of it's rebirth and Development.
-- Robinton's Mods: [ Mirror ] for some of his Mods incl Cubic Chunks Mod, due to DropBox broken links.
- Dungeon Generator for the Open Cubic Chunks Mod
- QuickSAVE-QuickLOAD for the Open Cubic Chunks Mod
I have noticed that the terrain is slightly more varied than it was before though, so I could be wrong. If you want to prove me wrong, just post the seed and coordinates for each of the 1.3 maps.
Cadika is proudly stating that his experiment was scientific, and he was doing science... but forgets that science is based on repetition and multiple test samples.
I believe that deep inside his subcounsciouss mind he knows that because he ignored every post that asked for VIDEO samples instead of SINGLE FRAMED IMAGES.
It's easy to be mistaken trying to identify framed images that lack the same various features that make beta and 1.3 terrain so different, and the same features that are asked by hundreds of people to be re-inserted in the game.
This features were listed by Hayarotle on his excelent post, that has yet to be replied by Cadika. But I guess after that post he/she realised the mistakes of the experiment and decided to ragequit this thread to avoid admitting the experiment was biased.