• 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!

The Meyer Thread (4x4 Reduction Sub-Method)

Mollerz

Swag Overlord
Joined
May 17, 2011
Messages
1,204
Location
Surrey, England
WCA
2011MOLL01
YouTube
Visit Channel
So I've been meaning to post this for a long time, I've had the idea of making Meyer a better method but it's taken my a while to consider a few things. Meyer is just a variation on reduction similar to Yau that instead of solving cross during reduction you solve the first block instead.

I hope there are a lot of roux solvers that want to help develop this method as I think there is a lot of potential with it. Basics of Meyer for those that don't know: 2 Opposite Centres, First Block, Last 4 Centres, Edges, 3x3 finish. The main drawbacks that existed before were things like pure OLL parity algs were bad, but there are some really good ones now, and that leaving the UL edge unsolved messes up flow through edge pairing. I have a solution to that which makes Meyer more similar to Yau which involves solving the DR edge with orientation to UL. During edge pairing this stays here and you can see where all your edges for the second block end up so going from edges to second block flows really nicely.

Here's a video of me rambling for 10 minutes and doing a couple of walkthrough solves. I hope it's clear enough but if you have any simple questions feel free to ask them on the video. And hopefully we can turn this thread into something nice with a lot of information relating to Meyer!


Go explore and contribute!

P.S. This thread is dedicated to Isaac Walters, without whom this would never be possible.
 
Last edited by a moderator:

TDM

Member
Joined
Mar 7, 2013
Messages
7,006
Location
Oxfordshire, UK
WCA
2013MEND03
YouTube
Visit Channel
What pure OLL parities do you use? I use:
Rw2 F2 r U2 r U2 x U2 r U2 r' U2 r U2 r2 U2 R2 x'
Though it's probably faster to just look at the D layer before CMLL and then do a normal OLL parity.
 

Mollerz

Swag Overlord
Joined
May 17, 2011
Messages
1,204
Location
Surrey, England
WCA
2011MOLL01
YouTube
Visit Channel
The video is up, the parity algorithms are in the description of that.

3-flip OLL Parity: (Rw' U' Rw' U2') (Rw' U2 Rw' U' Rw U2) (r U2 Rw U' Rw' U2 Rw' U2') (Rw' U' Rw')
Pure OLL Parity: r U2 r U2 r' U2 r U2 l' U2 r U2 r' U2 M r' U2 r'
 

ottozing

Platinum Member
Joined
Aug 13, 2011
Messages
3,289
Location
Canberra, Australia
WCA
2012MCNE01
YouTube
Visit Channel
Mollerz, how do you feel about doing just Lucas parity/some normal OLL parity alg before CMLL? I feel like it leaves a decent amount of room to force better CLL cases, or force good 4 flip EO meaning you can do COLL>ezpz EO (Unless it's a bad COLL ofc). Do you think using both this approach along with OLL parity after CMLL would also be something to consider?
 

Mollerz

Swag Overlord
Joined
May 17, 2011
Messages
1,204
Location
Surrey, England
WCA
2011MOLL01
YouTube
Visit Channel
Mollerz, how do you feel about doing just Lucas parity/some normal OLL parity alg before CMLL? I feel like it leaves a decent amount of room to force better CLL cases, or force good 4 flip EO meaning you can do COLL>ezpz EO (Unless it's a bad COLL ofc). Do you think using both this approach along with OLL parity after CMLL would also be something to consider?

Yes absolutely, but it changes the way people lookahead from SB -> CMLL hence the suggested approach. Although I don't see any problem doing CMLL first and doing the pure 3-flip alg since it's very fast anyway. By doing OP after CMLL I think the solve as a whole flows better but of course that is subjective.
 

Mollerz

Swag Overlord
Joined
May 17, 2011
Messages
1,204
Location
Surrey, England
WCA
2011MOLL01
YouTube
Visit Channel
I want to emphasize that this isn't necessarily a tutorial, just some ramblings and a couple of walkthroughs. I might make some more walkthroughs and go over the M-slice pairing a bit more since this seems to be the weakest part of the solve from my point of view.
 

Ross The Boss

Member
Joined
Jun 9, 2012
Messages
970
Location
Ontario Canada
WCA
2012SVEN01
YouTube
Visit Channel
how do you feel about mixing up the order of opposite centers and first block construction? my fastest solves (i dont use this method, but i still use this step) seem to be those where i ignore the opposite center and just build the first block right out of the gate. i suppose that this helps with recognizing and finding the fb pieces because you are not having to switch back and forth from one colour to another in your mind (if you want to have some more freedom to work with second center pieces, then you can always just build FB square, then do the opposite center). but as with most things, this is case sensitive. building opposite centers first can be fine as well.
and while we're on blocks, i (almost) always build the first center and an edge of FB at the same time. i usually do this by making two lines, both out of two FB center pieces and 1/2 of an edge. im sure most people do this, but if you dont, it's somethign to think about.

ps, i didnt watch the video (i might later) so sorry if im just spewing out reiterations.
 
Last edited:

cuber8208

Member
Joined
Oct 21, 2012
Messages
61
Location
Zeeland, Netherlands
WCA
2012GOOD02
YouTube
Visit Channel
Have played around with it for a few solves now. Still getting used to M-slice edge pairing but once that is fixed I think that it will be okay. How are you recognising for OLL parity though without being able to see the DB edge? Also, I like that for FB you can align a lucky corner and dedge and then slice the last dedge into place, quite fast if your inner slice moves are fast :)
 
Joined
Apr 23, 2010
Messages
1,391
Location
Scotland, UK
WCA
2009SHEE01
YouTube
Visit Channel
DP: R2 (2r2' x) U2 2L U2 2R U2 M' U2 2R U2 2R' U2 2R U2 2R2 U2

Quite nice IMO, then you can do LSE to leave pure OP or DP at the end. Makes sense to me to take advantage of an edges-only last step to never do 2 parity algs in one solve.

Edit: I did a few solves over lunch btw, kinda cool even though I'm rubbish at Roux. I personally went for putting any paired dedge into UL, don't know if that's bad or not.
 
Last edited:

Mollerz

Swag Overlord
Joined
May 17, 2011
Messages
1,204
Location
Surrey, England
WCA
2011MOLL01
YouTube
Visit Channel

I think that by mixing up the order the flow is bad. By being able to do your first block after 2 opposite centres helps the flow of the solve a significant amount. If your goal is to be as fast as possible with it, I think it is best to learn to get good by doing the block after the first two centres.


This is really interesting... I'd love to see more walkthrough solves. Especially if it was done on a livestream so we could talk to you while you do it. *wink wink nudge nudge*

I'm at a tricky point with streaming, regarding time and internet. I'm trying to figure it out :)

Have played around with it for a few solves now. Still getting used to M-slice edge pairing but once that is fixed I think that it will be okay. How are you recognising for OLL parity though without being able to see the DB edge? Also, I like that for FB you can align a lucky corner and dedge and then slice the last dedge into place, quite fast if your inner slice moves are fast :)

As said before you just have to look at the D layer, I think the CFOP equivalent is having to look at 3/4 sides of the U layer for PLL recognition, to see if you have PLL parity, and not being able to do 2-side recognition.

DP: R2 (2r2' x) U2 2L U2 2R U2 M' U2 2R U2 2R' U2 2R U2 2R2 U2

Quite nice IMO, then you can do LSE to leave pure OP or DP at the end. Makes sense to me to take advantage of an edges-only last step to never do 2 parity algs in one solve.

Edit: I did a few solves over lunch btw, kinda cool even though I'm rubbish at Roux. I personally went for putting any paired dedge into UL, don't know if that's bad or not.

Sweet alg, this could also come in useful. It definitely came to mind the idea of leaving parity until the absolute latest point and always doing it in one algorithm, but I haven't explored that so much.

Also there is a very particular reason for putting the DR edge in UL instead of any paired edge. From my experience doing any paired edge basically leaves your SB in a potentially awful situation, where by there is a high chance after edge pairing you will have to look for the DR edge along with another corner and edge, or both "pairs", giving you more things to look for. By putting DR there, with orientation, you automatically know exactly where DR is before you've even begun edge pairing, and it will be there every time. I've done a lot of testing and when I've done this the Second Block is always far more efficient, and I don't have to look for as many pieces. It's really hard to explain but I really think this is the best thing to do, I definitely recommend trying it.
 

waffle=ijm

Waffo
Joined
Jun 9, 2008
Messages
3,511
Location
Waffletopia
WCA
2008MANA02
YouTube
Visit Channel
You can't. You have to look at the D layer.

you can if you try hard enough from the start of 3x3. (tracking is love. tracking is life.)
And when you complete SB, you can already determine the number of bad edges (if odd then there's exactly 1 bad edge) and the best part is you don't even need to flip an incorrect edge.

more useful algs!

Justin Harder's 1Flip Double Parity r U2 r U2 l' U2 r U2 r' U2 M U2 r2 U2 r U2 r'
 

TDM

Member
Joined
Mar 7, 2013
Messages
7,006
Location
Oxfordshire, UK
WCA
2013MEND03
YouTube
Visit Channel
you can if you try hard enough from the start of 3x3. (tracking is love. tracking is life.)
And when you complete SB, you can already determine the number of bad edges (if odd then there's exactly 1 bad edge) and the best part is you don't even need to flip an incorrect edge.
Yes, but that's effort.

Well it's not if you're good at it, but I only ever track pieces, rather than what's in what location, so I'm not used to that type of look ahead.

More useful algs!
Normal PLL parity is [Uw2 Rw2 U2: r2]
PLL parity on D is [Uw2 Rw2 U2: l2]
 
Top