New 3x3x3 Method: Renders

**theZcuber** · 02-18-2012

Michael, how do you make it so it will generate orientation algs? I can never seem to get it to work, as it says it's incomplete (I only have a single color filled in on an edge)

**qqwref** · 02-18-2012

I don't think Cube Explorer will let you fill in only one color on a cubie. Every cubie has to be either totally filled or totally empty for it to work.

**teller** · 02-18-2012

Originally Posted by theZcuber

Michael, how do you make it so it will generate orientation algs? I can never seem to get it to work, as it says it's incomplete (I only have a single color filled in on an edge)

You have to hold down the CTRL key while filling in the U color, and make it look something like this:

**theZcuber** · 02-18-2012

Originally Posted by teller

You have to hold down the CTRL key while filling in the U color, and make it look something like this:

Thanks! I knew that there was a way, because I've heard of people generating OLL algs with it.

**Hodari** · 02-18-2012

Originally Posted by mariano.aquino

it's incredible. I pretty sure belt method is the one method with the largest number of authors all around the world (I include myself, I discovered it after learning how to solve sq1 ...^^). We are all happy to have discovered something new, only to find we're just another lamb in the field =P

If you think about it though, it probably makes sense that this would be a popular one. If you want to start creating a new method, the logical place to do so would be by looking at existing methods and looking at what the easiest and hardest parts of it are. Since most people use Fridrich, that would be the method they will look at trying to improve on. I think it would be fair to say that for most people, the easiest parts are PLL and OLL. Lot of algorithms to learn but once you do know them, you can do this step very quickly and without much thought and there are 2-look versions which are still decently fast. The cross isn't too hard either though mostly because it is preplanned during inspection. For most people though, F2L is the hardest part and the one which takes the most time.

So it makes sense that someone looking to create a new method would think that something which eliminates F2L and possibly the cross as well and relies more on OLL and PLL might be a good idea. To do this though, you need to solve the middle layer first, then separate the remaining pieces into their respective layers. Then orient and permute both layers. Unfortunately, this sounds good but doesn't actually work especially well since you run into problems like 1 edge incorrectly flipped on each layer or 2 twisted corners. Fixing all these requires too many steps and it ends up being too slow. So people might try playing around with what order all the steps are done in hoping to find some way to avoid these issues, leading to a number of different variations on belt methods.

And of course, the second reason that belt methods are so popular to invent is...while they sound good in theory, none of them really do work that well(at least none so far). They end up being good for when you just want to try something different, but not fast enough to actually be useful for speedsolving. So they remain obscure enough that a lot of people won't have heard of them until AFTER they go through the "trying to invent a new method" phase.

**miotatsu** · 02-18-2012

personally I think belt methods have potential to be fast, I use a system that is similar to LCBM personally. http://www.speedsolving.com/forum/sh...miliar-to-LCBM