elrog
Veteran
I know that EG uses 84 algs, but you can save moves by placing the pairs whever it is convenient. It usually saves 3 or 4 moves, and solving the top 4 corners while matching the pairs doesn't use any more moves than solving the top 4 corners normally.
I did have this idea for the beggining of a method in which you would orient all the edges during inspection and next use 16 algorithms like in guimond to orient the corners but preserve edge orientation.
There is also the way of solving corner orientation first as is done in the orient first method. I dislike this because of recognition for placing the middle edges.
I also had the idea of orienting edges, then orienting corners while placing the E edges in the E layer much the same way as F2L. You could also solve the top 4 corners orientation at the same time as placing the last E edge by doing F2LL algs that could possible be shortened.
Another approach would be to orient the edges and place the E layer edges in the E layer followed by algorithms to orient all of the corners. This uses quite a bit of algs and has recognition problems. The number of algs and the recognition could be reduced greatly by having 2 or 3 corners on one side already oriented. I actually heard of a pertus varient that purposely ignors the orientation of 1 corner in the F2L and solves it while orientating the top 4 corners. Prehaps these algs could be shortened and used in a different method.
As for the last 8 edges, you could also seperate out the edge circuits and then solve rather than seperating out the top and bottom colors to solve. I think seperating the top and bottom colors out would allow for better recognition of the last step. not including mirrors, solving all 8 edges once they have been seperated out from the top/bottom colors would take 20 algorithms or 10 if you use a 4 move parity algs that occurs 50% of the time.
One last thing before I go. It may be possible to make a method that orients all of the edges and corners, then places all of the edges into their circuits. You could then seperate out the top and bottom colors and do PBL. Solving parity in the edges circuits could either be done at the same time as PBL or seperating out the top/bottom corners. You could then finish the edges with 21 algorithms. Solving the parity in the edge circuits being done at the same time as PBL would triple the amount of PBL cases, but there are very few PBL cases. The only problem with this is that the number of PBL cases is increased due to not being able to AUF. I still think the number of cases for solving parity in the edge circuits at the same time as PBL would not take much more than 30 at its max. The only problem I see with this method is a recognition sysyem for seperating edges out into their circuits.
For the hardcore, you could possibly learn to place edges in their circuits while orienting corners, place edges into their circuits while seperating corners out, seperate the corners out while orienting them or solve the PBL case at the same time as seperating the top/bottom corners out and finish by solving the edge circuits and their parity. Solving the PBL case at the same time as seperating the corners out proabalby wouldn't be the best way to go because of recognition problems, and it would increase the alg count for this step and the last step by quite a bit.
I did have this idea for the beggining of a method in which you would orient all the edges during inspection and next use 16 algorithms like in guimond to orient the corners but preserve edge orientation.
There is also the way of solving corner orientation first as is done in the orient first method. I dislike this because of recognition for placing the middle edges.
I also had the idea of orienting edges, then orienting corners while placing the E edges in the E layer much the same way as F2L. You could also solve the top 4 corners orientation at the same time as placing the last E edge by doing F2LL algs that could possible be shortened.
Another approach would be to orient the edges and place the E layer edges in the E layer followed by algorithms to orient all of the corners. This uses quite a bit of algs and has recognition problems. The number of algs and the recognition could be reduced greatly by having 2 or 3 corners on one side already oriented. I actually heard of a pertus varient that purposely ignors the orientation of 1 corner in the F2L and solves it while orientating the top 4 corners. Prehaps these algs could be shortened and used in a different method.
As for the last 8 edges, you could also seperate out the edge circuits and then solve rather than seperating out the top and bottom colors to solve. I think seperating the top and bottom colors out would allow for better recognition of the last step. not including mirrors, solving all 8 edges once they have been seperated out from the top/bottom colors would take 20 algorithms or 10 if you use a 4 move parity algs that occurs 50% of the time.
One last thing before I go. It may be possible to make a method that orients all of the edges and corners, then places all of the edges into their circuits. You could then seperate out the top and bottom colors and do PBL. Solving parity in the edges circuits could either be done at the same time as PBL or seperating out the top/bottom corners. You could then finish the edges with 21 algorithms. Solving the parity in the edge circuits being done at the same time as PBL would triple the amount of PBL cases, but there are very few PBL cases. The only problem with this is that the number of PBL cases is increased due to not being able to AUF. I still think the number of cases for solving parity in the edge circuits at the same time as PBL would not take much more than 30 at its max. The only problem I see with this method is a recognition sysyem for seperating edges out into their circuits.
For the hardcore, you could possibly learn to place edges in their circuits while orienting corners, place edges into their circuits while seperating corners out, seperate the corners out while orienting them or solve the PBL case at the same time as seperating the top/bottom corners out and finish by solving the edge circuits and their parity. Solving the PBL case at the same time as seperating the corners out proabalby wouldn't be the best way to go because of recognition problems, and it would increase the alg count for this step and the last step by quite a bit.
Last edited:
