• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

42- Method Proposal

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
So some of you have probably heard something about this method before. I've been working on it for a while and I've spoken to a few people about it. Let's go through what this method is:

The simplest version of this method takes 42 moves on average using 42 algorithms (also something to do with Deep Thought :p)- this is discounting any beginner variations

  1. First Block
    • Exactly the same as Roux, no fancy CP, EP, CO, EO, BLD tracing etc, just a standard 1x2x3 in DL
  2. Second Block Square+ 1 oriented LL corner
    • The second block square is again exactly the same as the in Roux. For beginners you will need to build a block in BDR and orient any LL corner and place it in UBR then do an R.
    • For more advanced solvers, the block could be built directly with the R move off.
    • The corner orientation will frequently be skipped (just under 1/3 of the time)
  3. Conjugated CMLL/BT-redux
    • This sounds scary but it really isn't. all you do is recognise what case you have and execute an algorithm exactly as you would with COLL, CMLL, TTLL etc. You don't even need any weird tricks for cycle recognition or anything. You don't have to know a thing about Commutators or Conjugates. You just recognise by looking at the sticker in RDB and then what CLLesque pattern of orientation and what opp/same colours there are.
    • If you know all the algorithms for any CxLL, you already know all the necessary algorithms for this step so it is simply a matter of learning new cases to associate with that algorithm.
    • Conjugated CMLL/BT-redux cases (Word version)
    • Conjugated CMLL/BT-redux cases (PDF version)
    • *BT-redux is a slightly more general and advanced
  4. Last 7 Edges (L7E)
    • This step is very similar to LSE in roux and there are many ways to do this. my personal favourites are:
      • EO+FR: This is quite similar to EOLR with Roux where you combine 4a+4b so you solve EO for all edges while solving the FR edge. This essentially reduces the step down to oriented LSE which is fast and easy.
      • 2opp EO: this could be considered slightly more advanced than the previous but only fractionally more. You solve any 2 edges which are opposite each other (so whatever slots are in ULUR or UFUB (or even DFDB but i'm not sure I can recommend this).
    • Of course you can always do this algorithmically like in LMCF or WaterRoux but I don't find much improvement (at least movecountwise)
    • I will be releasing an "alg" set for EO+FR or 2opp EO (or both) soon.
  1. Ergonomic movesets
  2. Very low movecount (especially for the number of algs)
  3. Very low algorithm count (especially for the number of moves)
  4. Relatively simple recognition
  5. Easy obvious extension (especially compared to Roux)
  6. Can look proportionally further into the intuitive part of the solve than most methods
  7. Almost every benefit Roux has
  1. Multiple cases per algorithm can be difficult to get used to.
  2. Lower chance of lucky solves do to more of the cube being solved in final steps where lucky cases are more beneficial (if you can really call this a disadvantage as it leads to much more consistent solves)
  • If you use TCMLL then you don't even have to orient the corner with SB
  • Algorithms for each "set of cases" are not optimised for this method and just use CMLLs/COLLs from other methods so there is room for further optimisation
  • Any potential Roux techniques can be applied to this method
  • Nonlinear blocks may be more practical to implement than in Roux or FreeFOP etc because there is less to blockbuild at a time
  • NMB is much easier than with Roux as the recognition presented above is not as affected- this mean that you can easily build any of the 4 possible SBsquare which will make it easier to plan further into the solve when a solver becomes reasonably proficient at FB
  • It is possible that the solver could use F/B moves rather than L/R moves which could lead to better blocks or better L7E etc
  • This could also be a reasonably effective method for 2x2

 
Last edited:

greentgoatgal

Member
Joined
Feb 6, 2017
Messages
734
Location
US
Maybe I'll learn because I love the whole 42/Big Thought bit :p Sticking with CFOP as my main though ;)
 

efattah

Member
Joined
Feb 14, 2016
Messages
711
Seems like a nice compromise method, but I see a problem with the later step, solve FR and orient the other edges. Crafto and I tried that in one of our WaterRoux variants, and there was no way to recognize edge orientation. Solving DR (rather than FR) and orienting the edges can be done (180 degree offset), but the 90 degree offset of the FR edge confounds the edge orientation recognition. Perhaps you found a way around that-- I would love to learn more if you have solved it. There were some workarounds, like forcing the FR edge into a specific location/orientation then examining the other pieces, but it was cumbersome...
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
Seems like a nice compromise method, but I see a problem with the later step, solve FR and orient the other edges. Crafto and I tried that in one of our WaterRoux variants, and there was no way to recognize edge orientation. Solving DR (rather than FR) and orienting the edges can be done (180 degree offset), but the 90 degree offset of the FR edge confounds the edge orientation recognition. Perhaps you found a way around that-- I would love to learn more if you have solved it. There were some workarounds, like forcing the FR edge into a specific location/orientation then examining the other pieces, but it was cumbersome...
There will be a video detailing this as well as quite a few other things about the method (keep checking the playlist to see when it goes up) that I will be posting in the next few weeks (I tried to get the bare bones resources out there before announcing but I will be fleshing out and making more over the coming weeks).
 

Arc

Member
Joined
Jan 27, 2017
Messages
63
Location
Washington
Do you think it would be too difficult to use this on a case by case basis with Roux?
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
Do you think it would be too difficult to use this on a case by case basis with Roux?
Do you mean solve to F2B-1, see if you know the case and solve otherwise continue to full F2B?

Because if so then yes. My reasoning is that you have some set up for this case and if you focus on recognition then it wouldn't be very easy to continue the lookahead. This was one of my reasons for considering this a method seperate to roux.

@efattah what do you mean by this being a "compromise" method?
 

Arc

Member
Joined
Jan 27, 2017
Messages
63
Location
Washington
Do you mean solve to F2B-1, see if you know the case and solve otherwise continue to full F2B?

Because if so then yes. My reasoning is that you have some set up for this case and if you focus on recognition then it wouldn't be very easy to continue the lookahead. This was one of my reasons for considering this a method seperate to roux.

@efattah what do you mean by this being a "compromise" method?
No I was thinking just when the situation arose where you got the setup for free while solving the first square of SB.
 

Reed Merrill

Member
Joined
May 14, 2017
Messages
111
Location
Canada
Looks pretty cool! It's nice that it's so similar to roux. Is advanced 42 your main method now, or do you want it to be?
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
Looks pretty cool! It's nice that it's so similar to roux. Is advanced 42 your main method now, or do you want it to be?
Atm roux is still my main method. I am in the process of learning the cases for vanilla 42 (so far I've done the T and O sets) and it is definitely possible I will switch in the near to mid future especially if my times are the same with both methods.

I will also be releasing recognition videos for each set of 6 cases as I learn them (so expect the T cases soon).

The recognition is near identical to 3 sticker recognition in CMLL so it's actually pretty easy- you only have to take into account U layer stickers as well.
 
Last edited:
Top