Older cubers discussions

[kbrune]

Right now I'm very focused on accuracy... without it, speeding up isn't quite as useful.

I'm just memoing the wrong targets. Recall is fine. But I'll flip the F/L (in speffz) edge, or something.

Really irritating. I've had up to 3 flips in a solve. One of those might have been a missed flipped edge... unlikely all 3 were.

Funny thing is: I'm not twisting corners. It is all edges. Now, I do use audio edges, and that means I am under time pressure. But still.

What's your average for bld?

yes very irritating. I make mistakes like that as well. But for me it's mostly missing flipped edges because I can't seem to do the simple math while trying to retain memo. I also mess up on cycle breaks a lot. I'll forget to include the target right before the new cycle piece. Also aggravating to me. Somehow while looking for a new target on a break. I forget to include the current letter.

 

[kbrune]

In other news, I spontaneously forgot how to do a PLL out of the blue. I'm not even 100% sure which one. I've temporarily had issues and blanked here and there but only for a few minutes, never chronically like this.

How long has it been since you learned full PLL? It still happens to me once in a while. I'll blank very briefly on an OLL or PLL. I've actually been meaning to drill my OLLs once or twice for a while now. I find there's many that don't come up often for me. Once and a while it'll cost me half a second or so to get it or i'll just do a random OLL to change the case.

 

[pglewis]

I'm probably greener than either of you so mostly commiseration.

I have a lot of difficulty knowing if I have all the edges covered. When I get a success it's ususally because there aren't any cycle breaks lol

I still track with fingers. I've been experimenting with just remembering what I've hit on the M slice so I don't have to use any there. If three have been done on a face and I hit a cycle break I'll often go for the 4th and just remember that face is done, freeing up the fingers that were tracking there. I have not been effective at all doing any counting in an attempt to track a checksum.

Recall is fine. But I'll flip the F/L (in speffz) edge, or something.

I get this a fair bit and it seems a fairly common error. Thats no help but misery loves company.

Funny thing is: I'm not twisting corners. It is all edges.

I think I'm the opposite. The primary challenge of edges for me is tracing and running out of fingers. Corners are easier in that respect but I have a 50% greater chance of mucking up and memo-ing the wrong sticker.

One of my current focus points is immediately visualizing which face the target sticker is, first thing. If I put my finger on it first I run the risk of just memo-ing the sticker I'm touching... which isn't always the correct one.

 

[kbrune]

I'm probably greener than either of you so mostly commiseration.

I still track with fingers.

I get this a fair bit and it seems a fairly common error. Thats no help but misery loves company.

I used to track with fingers all the time. As I sped up I stopped doing it. I may have to use that again. That or i'll have to make it a point to use the "remember what's done on one face" trick.

The thing is. I feel like my brain doesn't have enough RAM to do all these things at the same time! lol
I've been able to speed up by dropping some sub routines. But in the end I don't have accuracy. At all.
I'm going to have to think about some adjustments and stick to the method for a while to see if I can get some accuracy with it.

 

[pglewis]

How long has it been since you learned full PLL? It still happens to me once in a while. I'll blank very briefly on an OLL or PLL. I've actually been meaning to drill my OLLs once or twice for a while now. I find there's many that don't come up often for me. Once and a while it'll cost me half a second or so to get it or i'll just do a random OLL to change the case.

I went for full PLL early, had them mostly learned before I was consistently under a minute so they're well established. I've never drilled PLL outside of just solving but I'm approaching that point.

Lowest probability OLLs are: Runway (1), Highway (55), Streetlights (56), and Mummy/H (57) all at 1/108 and Checkers (20) at 1/216. All those are on my "drill them specifically" list. Double Sune/H-OCLL is also 1/108 but is well established for me... the challenge on that one is getting used to recognizing COLL for it due to the low frequency. I've only been at full OLL for about the past six months and all the L edge cases are still rather slow for me on the recognition side.

 

[mitja]

I'm probably greener than either of you so mostly commiseration.



I still track with fingers. I've been experimenting with just remembering what I've hit on the M slice so I don't have to use any there. If three have been done on a face and I hit a cycle break I'll often go for the 4th and just remember that face is done, freeing up the fingers that were tracking there. I have not been effective at all doing any counting in an attempt to track a checksum.



I get this a fair bit and it seems a fairly common error. Thats no help but misery loves company.



I think I'm the opposite. The primary challenge of edges for me is tracing and running out of fingers. Corners are easier in that respect but I have a 50% greater chance of mucking up and memo-ing the wrong sticker.

One of my current focus points right now is immediately visualizing which face the buffer sticker is, first thing. If I put my finger on it first I run the risk of just memo-ing the sticker I'm touching... which isn't always the correct one.

When I trace edges I use fingers to cover L and R layer, but not cover the middle layer pieces. Just keep track if all 3 are already traced. It is easier on fingers and don't miss a side edge. At the same time I keep focus on middle 3 with my mind. It works for me.
I mostly use numbers for flipped corners. I memo them on the end. Each corner has numbers 1-8 for CW and 10-80 for CCW. For edges, i dont use numbers, just both letters. It could be done with number too ( 1-12) . For parity, I add a letter I never use in memo, that way it reminds me that I have parity.

 

[pglewis]

The thing is. I feel like my brain doesn't have enough RAM to do all these things at the same time!

That's exactly how it feels to me. I'm big on doing a lot of sighted practice, partially because I'm bad at keeping up practice and constantly have to re-drill execution and exercise my memory again. If I'm coming off a couple week hiatus I might do edges sighted just 2 or 3 pairs at a time. It becomes more like flash cards for basic tracing, assigning speffz, and execution without as much memo multitasking anxiety, plus I get quick feedback on any issues. Two or three days of that and stretching it out as I get more comfortable/confident seems to do wonders for me, but I'm still a 5-6 min guy on a good day.

When I trace edges I use fingers to cover L and R layer, but not cover the middle layer pieces.

Yeah, that's the same thing I've been trying with very promising results so far.

 

[JanW]

My bld record isn't very impressive right now either. About 23% success rate this year. Most solves fall within the 2:30-3:30 range, but all too often there's one mistake. Not sure if I should slow down to improve accuracy, or keep pushing for faster memo and execution and hope that accuracy eventually improves.

Regarding tracing or counting, I've never done either. I don't even remember how many edge targets there should be. In a way I go by feel. When I hit the buffer piece, I can tell if my memo is not long enough already.

Lately I've often been flipping edges first thing during execution. Usually I still run through edge memo one more time after closing my eyes, so I can just as well spend that time flipping stuff.

 

[newtonbase]

Unfortunately tracing well is mostly down to practice. You can start with the number of pieces expected and adjust it depending on cycle breaks and flips. Eventually you'll just get a feel for if it's right but everybody makes mistakes whatever their level.
I get a few wrong now as I switch UB and UL for parity avoidance. This messes up the count a bit. I need to work out how to allow for it.
I had a scramble today with 4 flipped edges but I only saw 3 so went to flip the buffer with them, did the wrong piece and ended up with 2 non buffer pieces flipped. Not a good solve.

 

[mark49152]

@mark49152
Was it just a slow day for you? Or did something throw off your memo rhythm?

I have a lot going on this year so have done less practice and my concentration is often weaker. This attempt was predominantly a memo fail. Maybe 5 out of those 8 missed cubes were simply blacked out memo. I'm not that unhappy with the result though, because I knew I was stretching my limits and it was good to get an official 20 attempt done within the hour.

I get a few wrong now as I switch UB and UL for parity avoidance. This messes up the count a bit. I need to work out how to allow for it.

Are you sure? I think the count is the same because you're still solving 12 pieces. So targets = 11 + breaks. In fact it's easier because targets must always be even and that can help you figure out what to look for at the end. For example if you finish on an odd number of targets when you know you have 2 pieces left, you know they must be swapped, not flipped or solved. (Disclaimer: I haven't been doing it long.)

 

[pglewis]

... It messes up F2L the same way every time so I'm consistently making the same mistake... more than a half dozen times the past 3 days. I'm fairly certain it's Ga ...

Yeah, it's the primary Ga listed on algdb, R2 u R' U R' U' R u' R2 y' R' U R

For some reason I now have the urge to change that first R' U R' U' to R' U R U'

 

[newtonbase]

I have a lot going on this year so have done less practice and my concentration is often weaker. This attempt was predominantly a memo fail. Maybe 5 out of those 8 missed cubes were simply blacked out memo. I'm not that unhappy with the result though, because I knew I was stretching my limits and it was good to get an official 20 attempt done within the hour.


Are you sure? I think the count is the same because you're still solving 12 pieces. So targets = 11 + breaks. In fact it's easier because targets must always be even and that can help you figure out what to look for at the end. For example if you finish on an odd number of targets when you know you have 2 pieces left, you know they must be swapped, not flipped or solved. (Disclaimer: I haven't been doing it long.)

I could do with spending more time thinking about these things rather than just doing them. Thanks for saving me the effort

 

[Mike Hughey]

G perms - I hate them. I forget them and have to relearn them at least once a year. I finally learned some new ones this year that are much faster, but they don't work so well for big cubes. And I'm sure I'll forget them even faster with these new ones than I ever forgot the old ones.

Tracing edges for 3BLD - I still use fingers too. If I don't, and I get multiple cycles, it almost always takes me way too long to find the missing cycle(s) if I don't use fingers. Which is sad, because certainly it slows me down.

@mark49152 - For me it seems like it's more often true that if I have an odd number of edges with corner parity, it means I memorized the corners wrong and there's really no parity and I have to start over memorizing again. :-( It also usually means a 2:30 to 3 minute solve. :-( :-(

 

[mark49152]

@mark49152 - For me it seems like it's more often true that if I have an odd number of edges with corner parity, it means I memorized the corners wrong and there's really no parity and I have to start over memorizing again. :-( It also usually means a 2:30 to 3 minute solve. :-( :-(

For me the usual cause is that I forgot to swap UB-UL at all . But I do think that consistently having an even number of targets will be useful once I get fluent enough to be able to exploit it without too much thinking.

 

[Mike Hughey]

Okay, now that you mention it, I realize I do that a lot too.

 

[kbrune]

I vote that we boycott G perms. We all refuse to do them until the cube gods give us 4 new fingertrick friendly PLLs

 

[mark49152]

The French G perms are quite nice.

 

[h2f]

I like Gperms. I use Valk's versions.

 

[kbrune]

I like Gperms. I use Valk's versions.

I've never seen those. They are nice. I'm not sure if I can perform them fast. I'll have to try them out.

 

[h2f]

I'm not sure if I can perform them fast.

I'm sure you can. As one said - I use one Gperm: its inverse, its mirror and the inverse of the mirror.