[Help Thread] Petrus Discussion and Help thread

[Cyrus C.]

Here's a quick example.

Scramble (Blue Front, Yellow Top): R U R' F' L' U2 L2 R' F L' R2 U2 R'

First thing I notice is the 2 yellow stickers facing me at FL and FU. I orient these With L' U L.
Next I see the yellow sticker facing the right at UR and the yellow sticker facing me at FU. I solve these with R' F R.
Finally I see a blue sticker facing up on the U face at LU, I know that it's pair must be with the red sticker at LF. I do a U' to set up, and L' U L to solve.
I know I'm done, since I've done 3 triggers, meaning I've solved six edges, the maximum that'll ever be unoriented.

 

[Jaycee]

Wow. That helped a lot! I may not recognize quickly, but I'm beginning to actually understand what I'm doing. T_T

Holding the two free sides on F and U and only fixing 2 at a time really helped.

Average of 12: 35.22
1. 39.90 R' U' F D' B2 R U' L2 U F' U F D F U L2 R2 U2 D2 L2 D' U' F2 D U
2. 40.94 R U L2 R' B U' D' F B2 U D' L' U' D' F' R' F' R2 D F2 D L2 R' U2 D
3. (41.81) F2 D2 U' F D2 L' U2 R2 L' F L F' B U2 F' L F' B' U R F L' B' U' B2
4. 38.83 D2 L' D2 F' R U' B2 D R2 U L' D' B2 R L2 B2 L2 B U2 D2 L' F' L' R F'
5. 29.56 R F2 U F' R' B R' L2 D2 R' F B2 L2 D2 U2 F U2 B F D2 U F B L' B
6. 32.59 R2 D F L2 D F2 U2 F' D2 U L2 D2 F' R U2 F' R2 U' F' U2 D' B D2 L2 B'
7. 32.62 U2 B' L' B2 U D L2 U' F' U2 B R' B2 L R2 U' L' R2 D' R U2 F' R' U2 B2
8. 35.30 B2 L' D' B2 U L' F R' B' L2 D2 R2 F' L D U R B2 F U' B2 U2 D' F' D'
9. 36.14 L' B' D R B' F2 U' R2 U F B' D2 L2 R' U2 F L2 U R U D' B2 U' L' D2
10. 37.65 D2 U' R2 B' U2 F B2 R D' B' R B2 U R D2 F2 L2 F B' R D F R2 B' R2
11. 28.67 U2 B' R' D2 B F' L F' B2 D' L D2 B' L2 B F' D2 U' R2 L2 F B' U2 D' B
12. (28.01) B L' F2 U' L' R B' R L B' D R L B F2 R' L' D B2 R2 L2 B' F' D B2

Thanks, Cyrus.

~Jaycee

 

[Jaycee]

29.99 Average of 50.

I'm done for today.

 

[pi.cubed]

Mini bump here, but I think It's okay as this is supposed to be a thread for general Petrus discussion.

I've been thinking about the number of 'fast' Petrus users recently and am wondering how many sub-20 Petrus users (with Petrus as their main method) there are on Speedsolving.
A few I've thought of:
ErikJ
Johannes (I don't think he cubes anymore, but he was sub 15 (or at least close to).
Cyrus C
Charlie Cooper
Sn3kypandaman

Any others?

And a few questions for you sub-20 Petrus users:
How long did it take you, from starting Petrus seriously, to becoming sub-20 consistently?

And something I've noticed in my blocks and the example blocks posted in this thread - most of the time my 2x2x2 block takes less moves than my expansion. But most of the examples I've seen here have the 2x2x3 with less moves than the 2x2x2. Do you guys generally have less moves on your 2x2x2 step or your expansion step? Also, do you have any tips for the 2x2x3 expansion step?

I only just started practising Petrus seriously a week or so ago, and atm I'm still just practising my blocks and stuff untimed. I'm using Johannes' Just another Rubik's Cube Solver to scramble the cube and use the optimal solver to find which 2x2x2 block is the best out of the 8? possibilites (I'm colour neutral) and how many moves it takes. Then I try to work out how to solve it in that many moves. If I cannot find the optimal solution in around 5 minutes then I check the solution. Then I work out what made that 2x2x2 more efficient than the others. I'm making a word document with a list of stuff to check and look for during inspection of the 2x2x2 block.
Then I just finish the cube, taking my time with the remaining steps.
Does anyone have other practice tips/ideas/techniques for a fairly nub Petrus user (If it makes any difference, I average(d) ~24 with CFO(O)P).

Thanks heaps in advance.

 

[Johannes91]

Johannes (I don't think he cubes anymore, but he was sub 15 (or at least close to).

He doesn't, but he can still get sub-16 averages easily.

How long did it take you, from starting Petrus seriously, to becoming sub-20 consistently?

Six months for my first sub-20 average. Probably one more for consistency.

Concentrating on getting sub-20 quickly might be counter productive, if you eventually want to be sub-15. You can waste a bunch of moves at every step and average sub-20, but later that will become more important. So keep your move counts in mind, even when speedsolving.

And something I've noticed in my blocks and the example blocks posted in this thread - most of the time my 2x2x2 block takes less moves than my expansion. But most of the examples I've seen here have the 2x2x3 with less moves than the 2x2x2. Do you guys generally have less moves on your 2x2x2 step or your expansion step. Also, do you have any tips for the 2x2x3 expansion step?

Rather than comparing them to each other, it would seem more meaningful to me to compare them to optimal or "good" numbers. How many moves do you typicall use for each step? For 2x2x2, you want to stay below 8 moves every time and use around 6 moves on average. It's possible to average just below 5 moves but that would be really hard. For the expansion, I don't know. It's by far the hardest step for me.

 

[pi.cubed]

He doesn't, but he can still get sub-16 averages easily.

Haha lol. Btw, thanks heaps for your solver which has optimal solutions for various Petrus steps. It is extremely helpful!

Concentrating on getting sub-20 quickly might be counter productive, if you eventually want to be sub-15. You can waste a bunch of moves at every step and average sub-20, but later that will become more important. So keep your move counts in mind, even when speedsolving.

Thanks.
What kind of movecounts would be good to aim for to be sub-15? If I am trying to get into good movecount habits now, what should I aim for?
Also, what kind of time breakdown do you think would be good for sub-20/sub-15? (So I know what to work on the most)

Rather than comparing them to each other, it would seem more meaningful to me to compare them to optimal or "good" numbers. How many moves do you typicall use for each step? For 2x2x2, you want to stay below 8 moves every time and use around 6 moves on average. It's possible to average just below 5 moves but that would be really hard. For the expansion, I don't know. It's by far the hardest step for me.

Thanks.

Because at the moment I am not speedsolving, but rather concentrating on improving my blockbuilding & lookahead, I have been aiming for a 6 move limit on my 2x2x2 blocks. On some scrambles I find this very difficult so I check on your solver and follow along with the solution.
Once I get into using a 15 second inspection, do you think it would be more reasonable to have a 7 move limit (I'm colour neutral)?
I'm focusing more on the 2x2x2 atm so I don't really know what my movecount for the expansion is, but I think 8-9 moves on average. I'm expanding in whatever direction looks easiest.

I saw in a Petrus example solves thread from a while ago that you suggested using 30 moves for F2L (I assume that means first four steps) in a speedsolve. The fact that I have only recently started would obviously make a difference here, but I find that I can only achieve this on very lucky scrambles. Do/did you frequently average 30 moves for the first four steps in a speedsolve?
I guess that could be a breakdown of something like: 2x2x2: 6 moves 2x2x3: 8 moves Bad edges: 6 moves. Wait that only leaves 10 moves for step 4. :S

Thanks for your reply!

Also, to anyone else reading this, please don't feel these are directed only to Johannes. The more replies the better!

 

[Ezy Ryder]

Petrus is not my main method but I had an average of 12 Sub-20 with it and I think I'll try to get better once I'll get a Sub-20 average with ZZ and maybe also Roux. Oh, and I started practicing it about three days ago. But not too much.

 

[mariano.aquino]

Cyrus: "Finally I see a blue sticker facing up on the U face at LU, I know that it's pair must be with the red sticker at LF"

Cyrus, how do you tell, on the go, that the egde HAS to have a red sticker....??

thanx in advance...=)

 

[oll+phase+sync]

I guess that could be a breakdown of something like: 2x2x2: 6 moves 2x2x3: 8 moves Bad edges: 6 moves. Wait that only leaves 10 moves for step 4. :S

2x2x3 the average migth be smaller than 8
Bad edges: at most 5 moves avg
And god's numerber for <R,U> Step 4 really is 10 or 11 moves - but I don't know how to get there while speedcubing.

I didn't really calculeate averages for myself, but my regular Petrus F2L averages are above 30 moves.

 

[Cyrus C.]

Cyrus: "Finally I see a blue sticker facing up on the U face at LU, I know that it's pair must be with the red sticker at LF"

Cyrus, how do you tell, on the go, that the egde HAS to have a red sticker....??

thanx in advance...=)

First of all because I've solved all the other edges. Also, all the other stickers besides that one and the one at UF are solved, so process of elimination. I probably would've just glanced to the side and seen the blue sticker, though. Or I could've figured it out at some point while solving other pieces.

 

[pi.cubed]

Thanks oll+phase+sync. Do you use Petrus as your main method? For some reason I have in my mind you use Roux. Not sure why.

I feel kinda silly posting these questions again, especially when they are only on the previous page, but only one of them was replied to before... I reckon that is probably because the post was to long and daunting, so I'll just post like the 'most important' questions here, as a tl;dnr thing.

Q1. What kind of time breakdown should I aim to achieve for a sub-20 (let's say 18) second average? My idea for a good time breakdown: Entire 2x2x3 block: 7 seconds, Bad edges: 2.5 seconds, Finish F2L: 3.5 seconds, OCLL+PLL: 5 seconds.
How does that sound?

Another, more general question: Does anyone have any practice tips/ideas/techniques? (If it makes any difference, I average(d) ~24 with CFO(O)P) Particularly focusing on the blockbuilding atm.

Thanks. Hopefully I'll get a reply soon

 

[chris w]

Q1. What kind of time breakdown should I aim to achieve for a sub-20 (let's say 18) second average? My idea for a good time breakdown: Entire 2x2x3 block: 7 seconds, Bad edges: 2.5 seconds, Finish F2L: 3.5 seconds, OCLL+PLL: 5 seconds.
How does that sound?

Another, more general question: Does anyone have any practice tips/ideas/techniques? (If it makes any difference, I average(d) ~24 with CFO(O)P) Particularly focusing on the blockbuilding atm.

Thanks.

Hi, I'm not really a petrus user, sometimes every now and then, maybe round 18ish avg i think, havent done an average but round there. Your time breakdown seems pretty reasonable. I wouldnt say bad edges should take quite as long though also a short efficient solution for the 2x2x2 + extention could be less than 7secs. as for the LL, just drill the algs and recognition times will begin to decrease.(all algs sub2 and you could have around a 4s LL) For blockbuilding just work on taking the least amount of moves as possible. I dont know how many they should take on average, but probably not more than 15? not sure really.

 

[oll+phase+sync]

Thanks oll+phase+sync. Do you use Petrus as your main method? For some reason I have in my mind you use Roux. Not sure why.

I'm not sure if I have a main methode ... my averages with Roux are maybe the most stable (an fewest moves), but to get the best single of a day Petrus and even CFOP offer more lucky cases to me.

Petrus also is not only the fist methode I learned (just 2 algs) it also has tons of variation options (EJF2L, Phasing, Pseudoblocks, Heise F2L , COLL, Last 3 corners, Last 5 Edges) to name some , I always come back.

At the moment I complete my ELL - let's see if it's usefull afterwards.

 

[Noahaha]

This is a very specific question, but hopefully someone knows an answer. I have been in the habit of using Lars' 4b tricks to complete the F2L (http://lar5.com/cube/s4b.html). I noticed however, that in his "A" section, there is a case missing, and I was wondering if anyone had a good solution.

The missing case is when the top face of the corner piece is the same color as the top face of the edge piece, but the corner piece is placed differently than in the sixth example on that page.

I feel like Lars must have just made a mistake because the second and fifth cases are the same.

The goal is to place the two pieces in the FR slot with the BR slot solved and the 2x2x3 block on the bottom left.

 

[pi.cubed]

This is a very specific question, but hopefully someone knows an answer. I have been in the habit of using Lars' 4b tricks to complete the F2L (http://lar5.com/cube/s4b.html). I noticed however, that in his "A" section, there is a case missing, and I was wondering if anyone had a good solution.

The missing case is when the top face of the corner piece is the same color as the top face of the edge piece, but the corner piece is placed differently than in the sixth example on that page.

I feel like Lars must have just made a mistake because the second and fifth cases are the same.

The goal is to place the two pieces in the FR slot with the BR slot solved and the 2x2x3 block on the bottom left.

I can't visit Lar's website, so I'm trying to work out what the case is from your description. However, I don't think there's quite enough information. Also, by 'top face' i am assuming you mean the sticker that is on the top face of the cube. Is that correct?

Where specifically are the pieces in question located on the cube? For example UFR and UL or something like that.

Some possible solutions since I'm not sure on the location of the pieces:
R U' R U2 R2 U' R2 U' R2
U' R' U R' U2 R U' R'
U' R' U2 R' U2 R U' R'
and the other case is the case where they are already connected in a corner edge pair.
The 'algorithms' above start with the corner needing to be solved at UFR. Do the inverse of the alg to get the case.

 

[insane569]

Is there any good tutorial that someone could recommend? Also, what is the idea behind orienting bad edges? Does it matter if you just do OLL + PLL for the last layer? Basically I think it can be simplified by building a 2x2x2 block and then expanding it to 2x2x3, then slotting in 2 pairs and an edge so that the first two layers are completed. Then finish with LL. I'm too used to CFOP and LBL methods that I probably don't even know what I'm talking about so correct me if I'm wrong.

Not sure about a tutorial. EO is done many different ways. The way i do it is i switch out bad edges 2 at a time since im not all that good at it. If you wanna use oll+pll for last layer you can but that wouldn't be full petrus. Thats how i solve one handed. cross, 2 F2L pairs next to each other(sorta like a 2x2x3 block) EO then finish F2L using R and U moves(can be done extremly fast) and then LL using 2 look OLL and full PLL.

 

[JasonK]

Pretty sure what Sahid is describing is FreeFOP - blockbuilding F2L, OLL, PLL. I thought the EO step was Petrus's defining feature. There's nothing wrong with skipping EO and doing it FreeFOP style, I just wouldn't call it Petrus.

I'm not a Petrus user so I could be talking complete nonsense, that's just the impression I've got

 

[Noahaha]

I can't visit Lar's website, so I'm trying to work out what the case is from your description. However, I don't think there's quite enough information. Also, by 'top face' i am assuming you mean the sticker that is on the top face of the cube. Is that correct?

Where specifically are the pieces in question located on the cube? For example UFR and UL or something like that.

Some possible solutions since I'm not sure on the location of the pieces:
R U' R U2 R2 U' R2 U' R2
U' R' U R' U2 R U' R'
U' R' U2 R' U2 R U' R'
and the other case is the case where they are already connected in a corner edge pair.
The 'algorithms' above start with the corner needing to be solved at UFR. Do the inverse of the alg to get the case.

Sorry I wasn't more specific. The "A" cases on Lars's site are the ones where the corner and the edge are separate in the top layer. The one case that he does not cover is when the corner is at UFR and the edge is at UL. Both have the same color sticker on the top face (you were right about that). What I have been doing is going R U' R' U' and then doing the sequence for that situation which is R U2 R' U R U' R'. Unfortunately this is 11 moves long whereas the rest of the sequences in that section are 7 (or 3) moves long. What I am looking for is a seven move solution to that case.

Also, I'm a little confused about your examples because none of them seem to be for step 4b of Petrus (The positions they solve do not have the BR, BRD and DR pieces solved).

 

[immortalchaos29]

Sorry I wasn't more specific. The "A" cases on Lars's site are the ones where the corner and the edge are separate in the top layer. The one case that he does not cover is when the corner is at UFR and the edge is at UL. Both have the same color sticker on the top face (you were right about that). What I have been doing is going R U' R' U' and then doing the sequence for that situation which is R U2 R' U R U' R'. Unfortunately this is 11 moves long whereas the rest of the sequences in that section are 7 (or 3) moves long. What I am looking for is a seven move solution to that case.

Treat it the same as the last of the "A" cases. Only difference is a double face turn after "hiding" the corner, which is why I think it wasn't included. So R U2 R' U' R U2 R' should solve it.

 

[Brest]

Sorry I wasn't more specific. The "A" cases on Lars's site are the ones where the corner and the edge are separate in the top layer. The one case that he does not cover is when the corner is at UFR and the edge is at UL. Both have the same color sticker on the top face (you were right about that). What I have been doing is going R U' R' U' and then doing the sequence for that situation which is R U2 R' U R U' R'. Unfortunately this is 11 moves long whereas the rest of the sequences in that section are 7 (or 3) moves long. What I am looking for is a seven move solution to that case.

Also, I'm a little confused about your examples because none of them seem to be for step 4b of Petrus (The positions they solve do not have the BR, BRD and DR pieces solved).

The algorithms listed on the "Step 4b tricks" are by no means a definitive list; there are many more cases for each corner position. It seems with that page that Lars is trying to display some tricks for specific cases and not trying to cover every possible case.

If this is the case you are describing

these are both optimal solutions
U' R U2 R' U2 R U' R'
U' R U2 R' U' R U2 R'