[Help Thread] FMC Discussion

[xyzzy]

can someone who's actually good at fmc rate this 31 (eo + blockbuilding)
View attachment 33227

1. I looked at 2 EOs (not more because knowing me I would DNF due to time limit) and chose the better one, but they both gave 8 moves to 2x2x2 so I just chose the second one hoping for a better solution
2. Pseudo 2x2x3 bc why not, I also had a choice on F2L-1 between two variations of the same goal, I chose the one that makes a block to give me 5c
3. Making a real effort to find good insertions, I was able to cancel 5 moves in total giving me a 31

If that's your entire working paper, check like maybe 4 times as many things.

Also extending 222 to 223 can be done in three directions, but if you have EO already done, then only one of the directions is "good" (the two free faces don't break EO) and the other two directions are "bad" (one of the two free faces will break EO). It's not necessarily a bad idea to extend in the "wrong" direction, especially since sometimes that's the only reasonable option you have, but it's significantly harder to work with.

Also also, your skeleton is 20 to 5c, not 19 to 5c.

 

[precompsubwr]

If that's your entire working paper, check like maybe 4 times as many things.

Also extending 222 to 223 can be done in three directions, but if you have EO already done, then only one of the directions is "good" (the two free faces don't break EO) and the other two directions are "bad" (one of the two free faces will break EO). It's not necessarily a bad idea to extend in the "wrong" direction, especially since sometimes that's the only reasonable option you have, but it's significantly harder to work with.

Also also, your skeleton is 20 to 5c, not 19 to 5c.

For some reason I check things really slow when it comes to fmc, idk why
Also mb I checked the skeleton over and still thought it was 19 at the time

Also are you suggesting I should do eo after the 2x2x3 or if it's just a noticing

 

[tseitsei]

Technically wouldn't 3 be leaving people with less cubes at a disadvantage? And realistically shaving off like 10 seconds every time after the third you scramble wouldn't be entirely detrimental.

Yes, but 3 is quite a reasonable number. Most cubers already own 3 cubes anyway. Or if that don't, it will be quite affordable to buy 3 cheap cubes. Compared to like 50 cubes.

When you get better at fmc you will check lots of stuff on NISS and thus solve and scramble The cube propably tens of times during the solve so it can actually add up to several minutes. And it is not even that important how significant the advantage is but rather the fact that it would give people with more cubes/money an advantage. That is not something that good rules of a fair competition encourage.
Of course it will always happen to some degree (newer more advanced more expensive speedcubes) but it is definitely not something we would want the rules to encourage more than is absolutely needed.

 

[tseitsei]

can someone who's actually good at fmc rate this 31 (eo + blockbuilding)
View attachment 33227

1. I looked at 2 EOs (not more because knowing me I would DNF due to time limit) and chose the better one, but they both gave 8 moves to 2x2x2 so I just chose the second one hoping for a better solution
2. Pseudo 2x2x3 bc why not, I also had a choice on F2L-1 between two variations of the same goal, I chose the one that makes a block to give me 5c
3. Making a real effort to find good insertions, I was able to cancel 5 moves in total giving me a 31

31 is a reasonable result with blockbuilding+insertions if that is about your average. Looks like a reasonable solution. You clearly already know how to use NISS and do insertions effectively. If you want to keep improving with blockbuilding you just need to check a lot more stuff.

But if you actually want to get better at FMC you should learn DR+HTR. Sub30 global average is very fast and MUCH easier to achieve with that compared to blockbuilding. And sub25 average with that method is also completely reasonable to reach in less than a year.

 

[precompsubwr]

31 is a reasonable result with blockbuilding+insertions if that is about your average. Looks like a reasonable solution. You clearly already know how to use NISS and do insertions effectively. If you want to keep improving with blockbuilding you just need to check a lot more stuff.

But if you actually want to get better at FMC you should learn DR+HTR. Sub30 global average is very fast and MUCH easier to achieve with that compared to blockbuilding. And sub25 average with that method is also completely reasonable to reach in less than a year.

I'm still in the process of learning htr, I also know how to do dr but it's really inefficient rn

 

[CornerTwisted]

Yes, but 3 is quite a reasonable number. Most cubers already own 3 cubes anyway. Or if that don't, it will be quite affordable to buy 3 cheap cubes. Compared to like 50 cubes.

When you get better at fmc you will check lots of stuff on NISS and thus solve and scramble The cube propably tens of times during the solve so it can actually add up to several minutes. And it is not even that important how significant the advantage is but rather the fact that it would give people with more cubes/money an advantage. That is not something that good rules of a fair competition encourage.
Of course it will always happen to some degree (newer more advanced more expensive speedcubes) but it is definitely not something we would want the rules to encourage more than is absolutely needed.

However people who are less into the event often will need less time overall and anybody who is sub 30 or so will likely have more than 3 cubes.

 

[tseitsei]

However people who are less into the event often will need less time overall and anybody who is sub 30 or so will likely have more than 3 cubes.

More than 3 but less than hundred. We have to draw the line somewhere and 3 is as reasonable number as any. Most people have 3 cubes so not a problem. Having no limit would be a problem since then people would try to get an advantage by bringing tens of cubes for an attempt.

EDIT: And the time save achieved is not totally irrelevant I don't think. I average like 16s for speedsolve right now. So without inspection and casually solving propably like 20+s. And let's say I solve and rescramble the cube 15 times in an attempt (seems like a reasonable estimate to me). That is already 5+ minutes that I would save if I had 20 cubes for an attempt

 

[Imsoosm]

What are some reasonable splits for eo, rzp, and dr for a beginner at the method? And also for advanced solvers?

 

[xyzzy]

What are some reasonable splits for eo, rzp, and dr for a beginner at the method? And also for advanced solvers?

Values below are subtotals, not splits; if you use more moves in one step, that eats into your budget for later steps.

Beginner: EO in 5, RZP in 6-7, DR in 15
Advanced: EO in 4-5, RZP in 5-6, DR in 10-11

Also, these are not hard rules (especially for "experts"). You'll have to adapt based on what the scramble is and what you've managed to find.

My experience with DR is that it has a very nonlinear improvement curve; once you're good at finding solutions for a step, you can just skip doing all the hard cases for the next step by finding something easier instead. For example, if you have only one 4c4e RZP, then you're forced to get that one RZP to DR, which can be difficult (and take many moves), but if you have ten 4c4e RZPs, then you can just choose the easier ones to focus on. In this sense, there's a huge difference between having one RZP in 5 moves versus having ten RZPs in 6 moves each; the latter may seem worse purely from a splits perspective, but is actually much better for finding good DRs.

 

[porkynator]

Reconstructions of my 21.33 NR mean, third place at Euros 2024.
Plain text version: https://sebastiano.tronto.net/speedcubing/euros-2133.txt

1.
Scramble: R' U' F L' B2 U D2 L' F2 L F L' B2 D L2 B2 D B2 U' R2 F2 R' U' F

Spoiler

Solution: U2 L D2 L2 U2 D' R2 D2 F B' L2 F' U2 D F U' D F' L D2 (20)

U2 L (D2 L') //EO (4/4)
(F) //RZP 3-3 aka 3c6e (1/5)
(D' U F' D' U2 B [3e]) //DR 4a1 (6/11)
(R2 D U2 L2 D2) //All but 3 edges (5/16)
[3e] = B' F L2 B F' D2 (6-2/20)

This scramble was a nightmare unless you found the right DR, in that case it was really easy. This EO was the only 4-mover I had, I missed (F) F D' B and its variation with F'. There were a total of 14 linear 5-movers, 58 if you include NISS which is extremely few. The DR trigger is also non-standard, but I am familiar with it.

I found this solution at around 20 minutes, and then spent most of the attempt trying to improve on it. There were many slice moves - the "corner skeleton" of this solution is ony 12 moves. So I am quite bummed that I missed Jan's 18. It could have been found like this:

[3e] = S' R2 U2 S U2 R2 (8-6/18)

Or by slicing away the L2 B F' D2 in my final solution:

L2 B F' D2 R2 D U2 = D2 R2 F' B D (-2)

Skill issue I guess. I am not too mad about this because in the fantasy world where I found this replacement then Radek surely does too, and he still gets ER + 1st place while my 20.67 would just be NR + podium.

Good job Jan for finding an amazing solution, well deserved

I had a few other DRs in 12 on this scramble, nothing worth noting. Literally tunnel visioned on the one good thing, easily the right call.

2.
Scramble: R' U' F L F' U2 B' R' U2 F2 U2 F2 R2 D2 B' D' U2 B2 F2 R' D R' U' F

Spoiler

My working sheet is a bit of a mess but I think this is what I submitted:

Solution: D L' F' R B F2 L2 B2 L2 U' F2 U2 L2 U2 R2 B2 R2 U' L2 F U' F' U D (24)

D L' F' R //EO (4/4)
B (D' U' F U F') //DR 4a2 (6/10)
(L2 U) F2 L2 B2 L2 U' //HTR (7/17)
F2 U2 L2 U2 R2 B2 R2 //Finish (7/24)

A boring solve. I was not feeling good during this attempt, both mentally and physically: I was stressed about wasting my first chance in ages to get a good mean at a big comp, it was too hot in the room and I needed to pee throughout the attempt. From this point of view, I am actually happy that I could pull out a somewhat average result.

I could have easily found this 23 from My DR:

(L2 D) F2 R2 B2 R2 D //Basically my HTR, I tried this one too (7/17)
F2 U2 R2 F2 L2 F2 //Finish, not that hard to see but I missed it (6/23)

And there was also a 22 with 4qt, not easy at all IMO:

F2 U B2 R2 U2 L2 D' R2 D' F2 B2 U {F2} (13-1)

I had this other DR that I worked on quite a lot:

(U F' R D') //EO (4/4)
(B' R' B' L B') //DR 4a3 (5/9)

And in the last minute I found this:

L (R2 U2 L) F2 B2 D2 B2 R //HTR (9/18)
F2 U2 R2 B2 U2 R2 //Finish with suboptimal leave slice (6/24)

Which I thought could be sliced to 23, but It does not look like. Anyway, I was close to this 23 that I think Tommaso found:

L (L2 D2 L') U2 B2 L //HTR (7/16)
F2 U2 L2 U2 F2 D2 F2 //Finish (7/23)

I had multiple ways to find a 23, if not for skill issues and a subpar mental / physical state. But after finding out about other people's result, I knew I still had a shot for podium.

3.
Scramble: R' U' F U2 F2 U2 B2 F' U2 F2 L2 R D' B R' L2 D U2 B D F' R' U' F

Spoiler

Solution: D R F B' U R U2 B2 D' R2 D U' F2 R' F2 R2 L' U2 L2 D2 (20)

D R F B' //EO (4/4)
U R //JZP (2/6)
U2 B2 U' * //DR 4b2, one of the many variations (3/9)
(D2 L2 U2 {L' B2 R}) //HTR (6/15)
(F2 B2) //Leave double slice (2/17)
{} = L R2 F2 R //Solve M slice (+1/18)
* = U D' R2 D U' B2 //Solve E slice (6-4/20)

Although not quite as bad as the first scramble, this scramble is not good either *unless* one finds the good DR. And once again I was blessed that I found it early on, and I wrote down my 20 before the 20 minutes mark. Regardless of what everyone else did, at this point I had already reached my goal of showing that all the work I put in the last months has been useful, and that I can still get an NR.

Once again the solution is very slicey (15 move corner skeleton), so I tried to look for a simplification not to repeat the show of the first attempt. If I submitted the 20, Kai "just" needed a 19 to beat me and sneak on third place, while Jay and I think Louis-Marie needed an 18. In the end there was a 19, but I honestly would have not seen it, even if I decided to try the righ 4qt HTR - why trying to insert a 2e2e in 18 when I already have a 20? For reference, here is the 19:

D R F B' //EO (4/4)
U R //JZP (2/6)
* U2 B2 U //DR 4b2 (3/9)
(R U2 R' F2 L' B2 U2 R) //HTR with fixed slice (8/17)
(F2) //2e2e (1/18)
* = B2 U2 B2 U2 B2 U2 (6-5/19)

Congrats to Jaye for finding this sub20, very cool solution!

As people pointed out, the first U after EO changes a 4-2 aka 4c4e RZP into another 4-2. I did not even think about it, not sure if I was dumb or smart

I had a bunch of other promising RZP from the 3- and 4-move EOs, but nothing worth noting. This is the only other DR I had, again with 3c6e RZP:

R F D' B' //EO (4/4)
U //RZP 3c6e (1/5)
R U' L R2 D //DR 2c4 (5/10)

I started working on HTRs from (L' U2 L') but then moved back to findingDRs and found the good stuff.

Very weird scramble set: 1 and 3 and one good thing and I was lucky to find it, while scramble 2 is your average scramble where there is a bunch of decent stuff but nothing stands out. I am somewhat surprise few people got sub23 on it, and not at all surprised that I panicked and got a 24.

I practiced a lot this year - almost 200 solves since January, more than 100 of which after FMC 2024. But I was very frustated because I could not get convincing results at home. For reference, I recently got new PBs for Avg25 (23.00), Avg50 (23.39) and Avg100 (23.70), but my last 25 solves were back to the 23.9 range. Almost every attempt felt like I was randomly smapping things and that I always missed something stupid. Definitely not what I wanted after practicing DR for 5 years. So going into Euros I was not expecting to do well at all, although I was hoping to get something good, like a sub23 mean, and then perhaps close my FMC adventure with a positive note instead of ragequitting.

Obviously I am very happy with this result. I was able to get an NR and major championship podium 5 years after reaching my peak. I am also incredibly happy to hear many people saying that they are happy for me. Not just congratulating, multiple people said they were cheering for me and that they are *happy* for me to get a good result. This fills me with joy

FMC changed a lot with the DR revolution, and I am not sure I have the right mindset for it. It feels like a completely different event. My solves rarely feel more skill-based that luck-based, and when I can use JZP / AR + double slice shenanigans (like in the 3rd attempt at Euros) is one of the rare times when the opposite it true.

So, am I going to quit now? I definitely do not want to keep practicing like I have done recently, with 1-2 solves per day where I constantly feel frustrated for missing solutions that seemingly everyone else finds. I also want to get back into blindsolving, which is likely going to kill off most of the free time I have been dedicating to FMC.

But I am still going to compete of course. I still enjoy discussing FMC with other people and sharing what I know. In the future I may work on making nissy more usable and maybe try to develop some new theory - AR + double slice solving might be the DR killer, and who knows if I'll ever understand how to make Corner First consistent. (Although on this last part I suspect DR is the way to make "Corner First" consistent, becuase the trick is "aligning edges and corners" so that solving corners with HTR drags edges along.)

Up to you whether you call this quitting or not.
At least it is not a ragequit

 

[precompsubwr]

Up to you whether you call this quitting or not.
At least it is not a ragequit

I just consider it changing main events, congrats on the NR too

 

[Pyjam]

So, am I going to quit now? I definitely do not want to keep practicing like I have done recently, with 1-2 solves per day where I constantly feel frustrated for missing solutions that seemingly everyone else finds.

Maybe it’s time to explore FT Octahedron FMC.

 

[ottozing]

4x4 FMC UWR of 51 moves (Previous being 53 by Trangium)

B' r2 R' L2 F2 R' U2 r B L2 f' D2 F u' R' B2 D2 L B2 L U2 L' U' f' D R f2 F u2 F' r f2 B' L2 R' F' L2 f2 D' F'

U2 R Bw B2 Lw’ // HTR F2C 5/5
R B Dw // HTR C 3/8
~
Rw2 // 2E 1/9
(Uw2 Fw2) // 4E 2/11
Fw2 // 5E 1/12
R2 D2 R D Rw2 // 7E 5/17
U F2 B2 U’ F’ L Bw2 // 1 slice from redux 7/24
D’ L R D R2 D2 R’ Dw2 // Redux 8/32
~
(F') B R2 F // EO 4/36
L' U' L2 F2 D // DR 5/41
(B2 L B2 L B2 L) // HTR 6/47
(U2 B2 U2 R2) // Solved 4/51

Notes:

Time taken was about 3h10 min with 2-3 breaks (only 30 mins spent on 3x3 stage to get the 19 xd)
Scramble isn't random state unfortunately, but this was an intentional choice since random state scrambles currently spoil too many aspects of the redux stage. In the future I'd like to do attempts on scrambles that are both random state & sufficiently padded since random move scrambles are a lot luckier on average

Sub 50 soon

 

[trangium]

4x4 FMC UWR of 51 moves (Previous being 53 by Trangium)

B' r2 R' L2 F2 R' U2 r B L2 f' D2 F u' R' B2 D2 L B2 L U2 L' U' f' D R f2 F u2 F' r f2 B' L2 R' F' L2 f2 D' F'

U2 R Bw B2 Lw’ // HTR F2C 5/5
R B Dw // HTR C 3/8
~
Rw2 // 2E 1/9
(Uw2 Fw2) // 4E 2/11
Fw2 // 5E 1/12
R2 D2 R D Rw2 // 7E 5/17
U F2 B2 U’ F’ L Bw2 // 1 slice from redux 7/24
D’ L R D R2 D2 R’ Dw2 // Redux 8/32
~
(F') B R2 F // EO 4/36
L' U' L2 F2 D // DR 5/41
(B2 L B2 L B2 L) // HTR 6/47
(U2 B2 U2 R2) // Solved 4/51

Notes:

Time taken was about 3h10 min with 2-3 breaks (only 30 mins spent on 3x3 stage to get the 19 xd)
Scramble isn't random state unfortunately, but this was an intentional choice since random state scrambles currently spoil too many aspects of the redux stage. In the future I'd like to do attempts on scrambles that are both random state & sufficiently padded since random move scrambles are a lot luckier on average

Sub 50 soon

Does it work though? Let me know if I'm missing something, but it seems like the centers are still scrambled at the end.

As far as I know, with indistinguishable centers you can't easily use NISS without weird things happening.

Regardless, it's exciting to see more people interested in moving 4x4 FMC forward!

 

[ottozing]

Does it work though? Let me know if I'm missing something, but it seems like the centers are still scrambled at the end.

As far as I know, with indistinguishable centers you can't easily use NISS without weird things happening.

Regardless, it's exciting to see more people interested in moving 4x4 FMC forward!

Hmm, seems like you're right. I guess I falsely assumed that since the redux worked with NISS that the 3x3 stage wouldn't be an issue. UWR still stands at 53 then!

NISS will no longer be used until 3x3 stage for my attempts I suppose

 

[ottozing]

Hmm, seems like you're right. I guess I falsely assumed that since the redux worked with NISS that the 3x3 stage wouldn't be an issue. UWR still stands at 53 then!

NISS will no longer be used until 3x3 stage for my attempts I suppose

Wasn't quite able to tie the 51, but at least it's still a UWR improvement!

Spent a few more hours than last time, mostly because it took me more than 2.5h to find the HTR centers solution I ended up using lol

I'm not at all convinced I entered redux anywhere close to optimally after the first 16 moves, but so be it... Tried so many ways to do the 3 cycle first hoping for an easy one slice finish but nothing was good (best was 30 to redux+PLL parity with this approach)

B' R2 u2 D2 U2 F2 u' R2 r' F' U2 R f2 u L' r' U2 B' r' D' R B L r B' D2 L' D2 F L' F' u U2 D F D2 u2 r' D' f'

B2 Rw’ F2 Uw // HTR F2C 4/4
R’ B Rw Fw2 Rw U2 Lw // HTR C + complete U/D C 6/10
~
U2 D R’ F’ Dw2 // 1 slice to redux + 3 wings 6/16 (!!!)
B2 R2 B’ D L’ D2 B’ R2 F2 Dw2 // AB3W 10/26
Lw2 L2 U L2 U’ Lw2 // Redux 6/32
~
({R D2 R'}) // EO 3/35
F2 U F' U B2 U // DR 6/41
L2 [F' R2 F' // HTR 4/45
L2 U2 B2] // Leave double slice 3/48

[] = F2 B U2 F' D2 L2 F B // Leave slice 8-6/50
{} = (L' R2 F2 L R2) // Solved 5-3/52

So annoyed that +4 was optimal from the double slice I'm not even convinced I got optimal from the DR but I don't want to spend more than an hour on 3x3 stage ever since I'd like to fit some practice in y'know

5x5 sub 100 next, then maybe sub 50 4x4 after (probably need to fish for a sub 30 redux to have a good chance)

 

[yes cubing]

I'm trying to get better at FMC, so I tried to watch the skeleton tutorial by JPerm, but it made no sense to me. Does anyone have any tutorials so I can get gooder

 

[Fred]

I'm trying to get better at FMC, so I tried to watch the skeleton tutorial by JPerm, but it made no sense to me. Does anyone have any tutorials so I can get gooder

This is my favorite:

It covers tons of techniques from beginner to advanced, so just skip to the parts that you want to learn.

 

[Puzzle Maniac]

A 34.00 average ranks you exactly 1000th in the world right now!

 

[abunickabhi]

Super interesting video explaining HTR - 2QT