• 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!

*No More PLL* 3x3 Rubik's Cube Method

Joined
Feb 27, 2014
Messages
946
YouTube
Visit Channel
No More PLL 3x3 Rubik's Cube Method (decided on a name xD).
*NMP* is a speedcubing method which uses 4c instead of PLL. You make your cross differently, solve F2L (the same way you do CFOP), COLL and 4c.
It has all the advantages of CFOP (doesn't handicap the CFOP flow), is more efficient than CFOP on average (I'm getting 45-47 stm solves), and the speedsolve can be done pauselessly since 4c is easily predicted, and during F2L, you can predict your corners case (CP). You can also achieve the same TPS that you get with CFOP, using this method and lookahead is similar to CFOP lookahead (so you won't need to re-learn lookahead)

Why is this method good?

This method provides an alternate finish to CFOP which has a lower move count, yet still allows for fast TPS and lookahead/prediction ability. Solving with PLL gives rise to bad cases like F/V/Y/N perms which you’ll never get with this approach. Plus, the worst 4c cases are 8 moves which just requires 8.08 TPS to sub 1. Since, these cases are all alg based (like U perm), you can achieve much faster TPS than 8 TPS. Let’s take a detail look.

3 Last-Step Methods

CMLL + LSE
This is standard for Roux. Avg movecount: 10 CMLL (computed from avg of all the algs in my cmll sheet + 0.25 AUF) + 13 LSE (assuming 8 EOLR/EOFB and 5 4c, essentially the most advanced LSE there is) = 23 STM total

OLL + PLL
This is standard for CFOP. Avg movecount: 10.3 OLL (algdb OLL sheet + AUF) + 13.7 PLL(using all of the fingertricky good algs(15 move G perms, RU U perms, J perm setup N perm, T perm setup F perm etc)) +2.25 AUF = 26 Moves

COLL + 4c
Standard for this method. Avg movecount: 11.2 COLL (taken from algdb + AUF) + 6-7 4c (0.25 move for AUF before M2, 1 move for initial M2, 1 move for AUF, 4 move avg for 4c) = 17-18 STM total

Not only is the proposed COLL+4c approach more efficient than CMLL+LSE and OLL+PLL, but COLL+4c can reach high TPS too, as COLL is recognize case, and spam TPS to solve. Predict 4c and you can pauselessly transition to 4c and spam MU once more. Since, 4c is pretty algorithmic (it can have 3 move solution, or a 5 move solution), you can spam TPS here like you do for U perms. MU U perms average 7 STM and 4c also averages 7 STM. U Perm can be done around 0.6 on average, and 4c can also be done around that that.

Overall, this method has its merits and can be equally fast as CFOP. You can use your current CFOP techniques, for this method (like X-Cross, CN, etc) and 4c is consistent and pretty short on cases, which is predicted with relative ease and solved like MU U perms MU algs.

Example Solve (more in the doc)
L2 B2 L2 D2 L2 F' R' F2 D R' F2 R U' R' D' L' R
z2
F' U' R D U' M' U2 l D // Mixed XX-Cross
y' R U2 R' U2 y' R' U R // F2L 3
U2 R U' R' // F2L 4 + EO
x' R U R' D R U' R' D' x y // COLL
U2 M2 U M' U2 M' // 4c
34 Moves STM

*NMP* Method Document: http://bit.ly/idknamemethod

Thank you :)


Edit: Bolded out certain texts.
Edit 2: Added proper move count for OLL/PLL and COLL after counting and verifying using python program. Thanks to Tao Yu for verifying :)
 
Last edited:

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
You're being slightly disingenuous with your movecounts for both OLL/PLL and your proposed method as the wiki gives optimal movecounts (there is simply nt way that COLL is shorter than CMLL) and I think you should be doing LS+LL in these movecounts as that's where the *IN* method starts to deviate in approach from standard CFOP (ignoring cross which for all intents and purposes is the same)

You should probably add all AUFs to in each method as well (before COLL, CMLL etc). Also, your method will force a rotation half the time either during or just before LL which none of the other methods do so it does interupt the flow of methods.

E: having read back this post it seems that I'm being too negetive. The method is interesting and could be an alternate way to solve and get very fast time though I'm not sure it is better than Roux (movecountwise) or CFOP (for ability to just spam tps) as the numbers seem to be a but off.
 
Last edited:
Joined
Feb 27, 2014
Messages
946
YouTube
Visit Channel
You're being slightly disingenuous with your movecounts for both OLL/PLL and your proposed method as the wiki gives optimal movecounts (there is simply nt way that COLL is shorter than CMLL) and I think you should be doing LS+LL in these movecounts as that's where the *IN* method starts to deviate in approach from standard CFOP (ignoring cross which for all intents and purposes is the same)

You should probably add all AUFs to in each method as well (before COLL, CMLL etc). Also, your method will force a rotation half the time either during or just before LL which none of the other methods do so it does interupt the flow of methods.

E: having read back this post it seems that I'm being too negetive. The deadline is intestine and could be an alternate way to solve and get very fast time though I'm not sure it is better than Roux (movecountwise) or CFOP (for ability to just spam tps) as the numbers seem to be a but off.
I see. Can you please help me with the move count averages? Thank you!
Yeah, it will force rotations 1/2 times after F2L, however that isn't as bad as you think as I do those rotations as d and d' moves, which isn't that bad. Probably will need getting some used to if a person uses this method. I personally think that *IN* is on par with traditional CFOP (if not better) and slightly inferior to Roux in terms of move count. In terms of TPS, you can average same TPS with Cross and F2L on both methods. OLL vs COLL is debatable but 4c vs PLL, 4c has the upper hand. The whole of Last Pieces with *IN* is just alg spam without any pauses (think of it as CMLL + EO, as you can predict EO+position of pieces during CMLL). OLL/PLL can be done close to pauseless execution, but its not entirely pauseless. COLL + 4c in this regard has the upper hand. TPS with COLL + 4c vs OLL/PLL should be similar, but I think we should be comparing times/splits rather than TPS/moves here.
 

gogozerg

Member
Joined
May 30, 2006
Messages
155
WCA
2004ROUX01
- Maybe you can optimize more the first step, by letting the "M" centers (the last four) random. Fixing them is a job for "4c". If they are misaligned by a quarter turn, invert the edge orientation pattern while fixing U-EO with the last pair.
- Even more freedom in the first step if you consider putting UF/UB in DF/DB, not just UL/UR.
 
Joined
Feb 27, 2014
Messages
946
YouTube
Visit Channel
- Maybe you can optimize more the first step, by letting the "M" centers (the last four) random. Fixing them is a job for "4c". If they are misaligned by a quarter turn, invert the edge orientation pattern while fixing U-EO with the last pair.
- Even more freedom in the first step if you consider putting UF/UB in DF/DB, not just UL/UR.
Fine additions sir :D Using the concept of mis-oriented centers and solving F2L with inverted edge orientation, will allow more efficient first cross. Plus, with the option for UF/UB and UL/UR, we have even more freedom during the initial stages of the solve. Inspection is also used wisely now with your approach. Thank you so much for the advice and improvements :)
 

gogozerg

Member
Joined
May 30, 2006
Messages
155
WCA
2004ROUX01
If I remember correctly, the technique you proposed before shares the same last steps (3/+4/) as the one above:
1/ Solve two L/R 1x2x3 blocks.
2/ Bring UL/UR (or UF/UB by the way) to DF/DB while orienting the 6 edges.
3/ Solve U corners with COLL.
4/ Solve the prepared L6E.

The main difference here is the Fridrich friendly way of making pairs, and 2/ relies on [M, U].
It's probably a matter of personal taste, but I think I prefer the first method.

What's the best in your opinion, do you think the recognition time before 2/ would make it slower (but as you said before, it offers some advantages for 3/ early recognition)?
Hard to tell I guess.
 
Last edited:

Cube4Life

Member
Joined
Jul 3, 2017
Messages
6
Wow...Really cool method! Low alg count, flows well, low move count. Just the right combination. You should make a page on this on the speedsolving wiki and get more people to use it. It's genius!!
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
:You have to do a y/y' after F2L/COLL if you're not at the right angle.
I've already mentioned this
How would 4c be better than EPLL? Let's say we did F2L + COLL + EPLL... I mean it seems a lot easier to just do that
4c is better than EPLL; it's less moves and in general easier to predict.

E: also what gilles said
 

Palmtop Tiger

Member
Joined
Feb 2, 2018
Messages
14
Location
Hamburg, germany
I've toyed around with this idea for a bit but i did it in a ZZ solve. So basically building an eo line with UR and UL or UB and UF. Apart from having more choices in building the eo line i noticed that phasing (solving relative edge permutation during inseration of the final pair) is super easy to do because you only need to look at the top layer to recognize it. Unfortunately it doesn't eliminate long LSE step 4c cases (but a few ones that are hard to recognise). I think i should mention this. maybe someone finds a legitimately good use for phasing in this method.
 
Joined
Dec 26, 2017
Messages
133
Location
Class
YouTube
Visit Channel
Strange cross?

No, it isn’t weird at all IMO, it’s just like a normal cross except with different yellow pieces in DF and DB, which very similar to CFOP.

So far, after using this method for 3 days, my PB’s are as followed
1/mo3/ao5/ao12/ 23.03/29.33/30.56/33.45

Definitely a great method, will be switching, the cross is quite optimisable, has great finger tricks, but still remains as a great low-move count method.
 

Duncan Bannon

Member
Joined
Sep 9, 2017
Messages
1,817
Location
Here
No, it isn’t weird at all IMO, it’s just like a normal cross except with different yellow pieces in DF and DB, which very similar to CFOP.

So far, after using this method for 3 days, my PB’s are as followed
1/mo3/ao5/ao12/ 23.03/29.33/30.56/33.45

Definitely a great method, will be switching, the cross is quite optimisable, has great finger tricks, but still remains as a great low-move count method.

PM me how your doing. It would be nice to see your progress. I'm interested in the method as well.
 
Joined
Dec 26, 2017
Messages
133
Location
Class
YouTube
Visit Channel
Yes, just build your eo-line with the up-right and up-left edge. Alternatively you can use up-back and up-front ( up-front and up-back is only a bit harder to recognise during the 4c step )
Yes, that’s what I do!

I build the EO-Line, except on left and right, so I skip an alg. Trust me, it doesn’t take long to get decent at. I average 7 moves for EO line in this method.
 
Top