Recently, I've noticed more method proposal threads popping up on this site than normal, especially from newer members. It makes me happy to see that so many people have become interested in method development, as it's a super cool art! However, I thought it prudent to make a thread to discuss the topic as a whole.
To begin: new method and subset ideas should not be given their own individual threads. The vast majority of them belong in the official New Method / Substep / Concept Idea thread (linked here). Once a method has been thoroughly developed, has been peer-reviewed in that thread, and has been confirmed to meet the three criteria below, then it can perhaps be moved to its own thread. It can be annoying to others to see an idea given its own thread and popping up in the "Latest posts" section of the site, only to find out that the method is a duplicate of one that's been proposed many times before. Now that I'm a mod, I may start moving posts in these new threads to the New Method thread.
But before posting even to the linked idea thread, there are three questions you should be asking yourself:
1) Is it original?
When designing a method, it's important to make sure you aren't accidentally repeating something that's been done before. The first post in the new method idea thread I linked has a small list of commonly-proposed ideas, but there exist many other niche methods and subsets. One thing that's important to keep in mind is that in general, the less a method/subset deviates from existing popular methods, the more likely it is to be a duplicate idea. CFOP but skipping a cross edge only to insert it later while solving EO, or combining the blockbuilding of Roux with the last layer of CFOP, or skipping a D-layer corner in Petrus and using the slot to solve something else later—these are all very close to existing methods and have been proposed many times, and are usually slower than just doing the popular version of the method.
Another way to decide if your method is likely original is by asking how obvious it seems. For example: starting a solve with making the 4 F2L pairs, then solving corners of the last layer, then solving the M and S slices? This seems like an obvious deviation from something like CFOP, and in fact has been proposed by many different people. I myself "discovered" this method years ago, thinking it new and exciting, when it fact it was just a duplicate of the PCMS method. If the method doesn't take much effort to think up, then unfortunately, you are probably not the first to think of it. This is one of the consequences of a hobby being around for decades: most of the good ideas are taken.
2) What makes your method unique?
For 3x3 alone, there are already hundreds of methods, subsets, alg sets, and other concepts out there. If you're making something new, it's important to consider the why. What does your idea bring to the table that others don't? Making something original isn't hard—I can think up a new 3x3 method right now: solve the corners of the D layer, then the edges of the U layer. Solve edge orientation, then finish the corners, then solve the last 8 edges with commutators. This method is almost certainly original, but it isn't good. It's hard for beginners, and it's slow for intermediate and advanced solvers. It doesn't have particularly few algorithms, and it isn't especially good for BLD or OH solving. If you want your idea to go somewhere, there has to be something special about it, something that sets it apart from what is already commonly used. Think critically about your method, try to poke holes in it even before showing others, and either patch those holes up or try to come up with another idea instead. You can come up with something interesting, but it needs to have something unique going for it.
3) Have you put in the work?
With how many methods have been proposed over the years, many great ideas have come to light, and countless others have proven unoriginal, weren't great in the first place, or have otherwise faded into obscurity. In order to set yours apart, you need to prove that your method is worth paying attention to. Lay out the steps clearly, describing how they're done and what they're meant to accomplish. Give move count estimates for each step along with algorithm counts. Describe the pros and cons, and how the idea compares to similar ones. Give example solves, or better yet, real speedsolves you yourself have done after practicing your method. If you don't care about your own method, if you aren't invested enough to put in this work, then it will be difficult to convince others to.
As I mentioned, method development is awesome. I really enjoy seeing what others come up with, and there have been some really interesting ideas to come out of this forum. But as someone who's been on this site for the last decade, I've become frustrated seeing the same ideas proposed time and time again, seeing methods whose utility haven't been properly thought through, and seeing halfhearted suggestions without much time or effort behind them. You should pursue this sub-hobby, but make sure you're approaching it with the correct mindset. So go out there, put in some work to make something interesting, and I'll see you in the new method ideas thread!
To begin: new method and subset ideas should not be given their own individual threads. The vast majority of them belong in the official New Method / Substep / Concept Idea thread (linked here). Once a method has been thoroughly developed, has been peer-reviewed in that thread, and has been confirmed to meet the three criteria below, then it can perhaps be moved to its own thread. It can be annoying to others to see an idea given its own thread and popping up in the "Latest posts" section of the site, only to find out that the method is a duplicate of one that's been proposed many times before. Now that I'm a mod, I may start moving posts in these new threads to the New Method thread.
But before posting even to the linked idea thread, there are three questions you should be asking yourself:
1) Is it original?
When designing a method, it's important to make sure you aren't accidentally repeating something that's been done before. The first post in the new method idea thread I linked has a small list of commonly-proposed ideas, but there exist many other niche methods and subsets. One thing that's important to keep in mind is that in general, the less a method/subset deviates from existing popular methods, the more likely it is to be a duplicate idea. CFOP but skipping a cross edge only to insert it later while solving EO, or combining the blockbuilding of Roux with the last layer of CFOP, or skipping a D-layer corner in Petrus and using the slot to solve something else later—these are all very close to existing methods and have been proposed many times, and are usually slower than just doing the popular version of the method.
Another way to decide if your method is likely original is by asking how obvious it seems. For example: starting a solve with making the 4 F2L pairs, then solving corners of the last layer, then solving the M and S slices? This seems like an obvious deviation from something like CFOP, and in fact has been proposed by many different people. I myself "discovered" this method years ago, thinking it new and exciting, when it fact it was just a duplicate of the PCMS method. If the method doesn't take much effort to think up, then unfortunately, you are probably not the first to think of it. This is one of the consequences of a hobby being around for decades: most of the good ideas are taken.
2) What makes your method unique?
For 3x3 alone, there are already hundreds of methods, subsets, alg sets, and other concepts out there. If you're making something new, it's important to consider the why. What does your idea bring to the table that others don't? Making something original isn't hard—I can think up a new 3x3 method right now: solve the corners of the D layer, then the edges of the U layer. Solve edge orientation, then finish the corners, then solve the last 8 edges with commutators. This method is almost certainly original, but it isn't good. It's hard for beginners, and it's slow for intermediate and advanced solvers. It doesn't have particularly few algorithms, and it isn't especially good for BLD or OH solving. If you want your idea to go somewhere, there has to be something special about it, something that sets it apart from what is already commonly used. Think critically about your method, try to poke holes in it even before showing others, and either patch those holes up or try to come up with another idea instead. You can come up with something interesting, but it needs to have something unique going for it.
3) Have you put in the work?
With how many methods have been proposed over the years, many great ideas have come to light, and countless others have proven unoriginal, weren't great in the first place, or have otherwise faded into obscurity. In order to set yours apart, you need to prove that your method is worth paying attention to. Lay out the steps clearly, describing how they're done and what they're meant to accomplish. Give move count estimates for each step along with algorithm counts. Describe the pros and cons, and how the idea compares to similar ones. Give example solves, or better yet, real speedsolves you yourself have done after practicing your method. If you don't care about your own method, if you aren't invested enough to put in this work, then it will be difficult to convince others to.
As I mentioned, method development is awesome. I really enjoy seeing what others come up with, and there have been some really interesting ideas to come out of this forum. But as someone who's been on this site for the last decade, I've become frustrated seeing the same ideas proposed time and time again, seeing methods whose utility haven't been properly thought through, and seeing halfhearted suggestions without much time or effort behind them. You should pursue this sub-hobby, but make sure you're approaching it with the correct mindset. So go out there, put in some work to make something interesting, and I'll see you in the new method ideas thread!
. i study all the alggs to the l8e (11-15 moves per case )and when i bout to post i see ts
