LBL Middle Layer Cases? [Ineffecient, Yet Possible]

[IamWEB]

I was watching a video on YouTube about speedcubing, and an idea somewhat randomly popped into my head about the LBL method.

After the first layer, you have to individually insert the 4 edges of the middle layer. I was thinking, what if you double could multi-slot, but with edges only? And on, not just 2, but 3 or 4 edges at once? Of course, the algorithms would likely grow longer with the more edges you're trying to insert ( I say likely because of possible special cases), but it's definitely possible.

I know this would probably not be efficient, but it's still a possible method for solving the middle layer on LBL.

Could this actually be of good use (A least when you 'double-slot')? How many cases would there be for inserting x amount of edges? How long are the algorithms.

I'm really not sure how to go about testing this entire thing, but I'm sure it was worth mentioning, as it struck a sudden interest in changing how the solve goes.

Never know the potential of it until you try, although some could prove me wrong there if need-be. The question is, where to start? And what do you think of this? (by this, I don't mean whether you think it's for speedcubing or not; it obviously doesn't beat F2L, but it's a change from what's already done)

 

[nitrocan]

For that, you need to study the cube a lot and make a lot of ways to insert the edges. You will need algorithms etc.

The LBL method is just a get-ready for Fridrich. If you want to get better at F2L, just use Fridrich F2L.

 

[gavnasty]

I thought the LBL method is meant for beginner's who want to learn as little as possible yet still be able to solve a cube.

I guess that idea sounds possible, but I don't think it's reasonable.

 

[Ellis]

I'm really not sure how to go about testing this entire thing, but I'm sure it was worth mentioning, as it struck a sudden interest in changing how the solve goes.

Don't test it. And even though something has struck an interest in you to change how a solve goes, which is good, it doesn't automatically make it worth mentioning.

If you're doing a full first layer before moving on to the second layer edges, why not do something like leave the last corner free and use the D layer to quickly insert edges. Or you could even start inserting middle layer edges right after the cross and leave the last one empty so you can insert all the corners.

Or.... I dunno if this would work but you might be able to try putting each corner next to its edge before you solve it. However you might have to learn like, zero algorithms for this, oh also don't forget add zero algorithms for some of the harder cases.

You can solve the cube however you want, but why mention every way to do it even if its incredibly inefficient?

 

[IamWEB]

It's because if the algorithms weren't extremely long, and recognizing cases wouldn't take 'forever' (something to be found out through testing), I would actually show this to some using a different beginner's method, before teaching them F2L.

I myself use Fridrich F2L and am learning more and more LL algs, but I wanted other's opinions and an idea I had.

Yes, I won't test it, but it's only because I'm still not sure exactly how yet (other than studying how many cases I can find). If I could I would, but I would only post results if I find something that would make this better than LBL. Chances might be slim, but as far as beginners go, do you think it's work a shot to test?

 

[Ellis]

but as far as beginners go, do you think it's work a shot to test?

Honestly, no. Even if there were no algorithms to learn it still wouldn't be more efficient than keyhole. Taking out a correct corner to pair it with an edge and reinsert it is 7-8 moves. The most turns you'd save with a double slot is probably one or two, which means that you'd still be doing 25-30 moves just for the middle layer edges. With keyhole the first 3 edges are 4-5 moves each. and very little recognition time is needed. It's not a good beginners method, it's not a good intermediate or advanced method. If you want to spend your time on it and test it, then go for it, but no one is going to want/need/use it.

 

[daniel0731ex]

i actually got a idea of correcting SL permutation before, and im occseasonally using the two alg i invented myself

im still working on finding algorithms that could correct SL orientations



R2 U2 R2 U2 R2 <- a simple algorithm invented by an asian retard (oxymoron?)

E2 L2 E2 L2 i just simply took it from the P06 case alg, not completely 'invented' by myself

Honestly, no. Even if there were no algorithms to learn it still wouldn't be more efficient than keyhole. Taking out a correct corner to pair it with an edge and reinsert it is 7-8 moves. The most turns you'd save with a double slot is probably one or two, which means that you'd still be doing 25-30 moves just for the middle layer edges. With keyhole the first 3 edges are 4-5 moves each. and very little recognition time is needed. It's not a good beginners method, it's not a good intermediate or advanced method. If you want to spend your time on it and test it, then go for it, but no one is going to want/need/use it.

well, but if you just simply correct the permu/orient of second layer with a short algorithm you'll find it pretty efficient

 

[Escher]

R2 U2 R2 U2 R2 <- a simple algorithm invented by an asian retard (oxymoron?

what?
message too short.

 

[IamWEB]

R2 U2 R2 U2 R2 <- a simple algorithm invented by an asian retard (oxymoron?

what?
message too short.

It swaps the 2 middle edges of the R face, and the front and back edges of the U face.

Perform it again using L2's, and it's like an E Perm of the middle layer.

Performing another U2 at the end of the original algorithm does was F2 D2 F2 D2 F2 does, because it's the same thing skewed around.


Unless you're talking about the 'retard' thing.

 

[somerandomkidmike]

I was watching a video on YouTube about speedcubing, and an idea somewhat randomly popped into my head about the LBL method.

After the first layer, you have to individually insert the 4 edges of the middle layer. I was thinking, what if you double could multi-slot, but with edges only? And on, not just 2, but 3 or 4 edges at once? Of course, the algorithms would likely grow longer with the more edges you're trying to insert ( I say likely because of possible special cases), but it's definitely possible.

I know this would probably not be efficient, but it's still a possible method for solving the middle layer on LBL.

Could this actually be of good use (A least when you 'double-slot')? How many cases would there be for inserting x amount of edges? How long are the algorithms.

I'm really not sure how to go about testing this entire thing, but I'm sure it was worth mentioning, as it struck a sudden interest in changing how the solve goes.

Never know the potential of it until you try, although some could prove me wrong there if need-be. The question is, where to start? And what do you think of this? (by this, I don't mean whether you think it's for speedcubing or not; it obviously doesn't beat F2L, but it's a change from what's already done)

I think it could be an efficient method. You actually can use some ELL algorithms, and other 3 cycles. I did this for a while, and I think it's pretty efficient. Considering that now I take 7 seconds for my first layer, I think the first 2 layers could be done easily in 14 seconds for me. I think the first layer could get down to 4 seconds for an average, then the middle layer edges down < 3 seconds, the last layer I'm not sure about. What does Eric get for an LL?

 

[rachmaninovian]

I was watching a video on YouTube about speedcubing, and an idea somewhat randomly popped into my head about the LBL method.

After the first layer, you have to individually insert the 4 edges of the middle layer. I was thinking, what if you double could multi-slot, but with edges only? And on, not just 2, but 3 or 4 edges at once? Of course, the algorithms would likely grow longer with the more edges you're trying to insert ( I say likely because of possible special cases), but it's definitely possible.

I know this would probably not be efficient, but it's still a possible method for solving the middle layer on LBL.

Could this actually be of good use (A least when you 'double-slot')? How many cases would there be for inserting x amount of edges? How long are the algorithms.

I'm really not sure how to go about testing this entire thing, but I'm sure it was worth mentioning, as it struck a sudden interest in changing how the solve goes.

Never know the potential of it until you try, although some could prove me wrong there if need-be. The question is, where to start? And what do you think of this? (by this, I don't mean whether you think it's for speedcubing or not; it obviously doesn't beat F2L, but it's a change from what's already done)

I think it could be an efficient method. You actually can use some ELL algorithms, and other 3 cycles. I did this for a while, and I think it's pretty efficient. Considering that now I take 7 seconds for my first layer, I think the first 2 layers could be done easily in 14 seconds for me. I think the first layer could get down to 4 seconds for an average, then the middle layer edges down < 3 seconds, the last layer I'm not sure about. What does Eric get for an LL?

]

OLL/PLL would take...3-5s

 

[Dirk BerGuRK]

Never know the potential of it until you try, although some could prove me wrong there if need-be. The question is, where to start? And what do you think of this? (by this, I don't mean whether you think it's for speedcubing or not; it obviously doesn't beat F2L, but it's a change from what's already done)

Off the top of my head, it doesn't seem efficient as you state, but I feel people are too quick to dismiss on this forum. Where to start you ask? Figure out how many cases you think you are willing to memorize (depending on the number of edges you want to insert at a time), use a cube solver to generate algs, find out move counts, and time yourself on a few solves with cases you know. Takes a lot of effort to know for sure, but thats half the fun. I might suggest using a block building approach for the first layer instead of simply building a cross then inserting corners (or vice versa). With time I believe it is easier and uses fewer moves, but I have no proof. In addition, you could save yourself moves by not fully completing the first layer (leaving out a corner as in keyhole). This also makes the block building approach easier. Anywho, I always encourage new ideas, and I haven't heard this one before. Good luck.

 

[Brett]

Hm, an odd combonation of ideas could be.

1. Petrus 2x2x3
2. Place last "cross" piece, and the 2 final corners. One corner can be unoriented.
3. Use an alg to place the last 2 edges (could be 2gen? Would be fast)
4. Use an alg to orient the final corner, and all corners of the LL.
5. PLL.

Idk, I'm bad at figuring out how many potential cases there would be.

 

[Kenneth]

If we ignore the possibility to use AUF (it complicates things), the first edge may be in 2*8 places (including solved) so it got 16 cases. Add a second edge and it has got 7*2 = 14 cases, the two together makes 16 * 14 = 224 cases.

All four would make (8*2)*(7*2)*(6*2)*(5*2) = 26,880 cases

 

[joey]

Also, recognition would be absolutely horrible. I think you should forget about this

 

[Lord Voldemort]

I think the reason people are so critical is that it's much easier to progress to Fridrich F2l and if you really want algorithms you can learn multislotting...

 

[somerandomkidmike]

Also, recognition would be absolutely horrible. I think you should forget about this

I averaged 20 seconds for the F2L for using this.... it's not horrible.

 

[IamWEB]

Also, recognition would be absolutely horrible. I think you should forget about this

I averaged 20 seconds for the F2L for using this.... it's not horrible.

You should upload a video solve.

 

[riffz]

This sounds kind of pointless. LBL is used by beginners, so what's the point in making it more complex by adding more algs? Fridrich F2L is faster anyway.

 

[IamWEB]

This sounds kind of pointless. LBL is used by beginners, so what's the point in making it more complex by adding more algs? Fridrich F2L is faster anyway.

I'm aware the Fridrich F2L is by far the better method, the point is to expand knowledge on the cube and ways to solve it.