[3x3] How I Think I Should Approach on Breaking Sub-20 Barrier

[Fishcake]

Currently, I average around 25 seconds and I would like to improve it and hopefully break the 20 seconds barrier in 5-6 months. Here is what I think I should do to improve my time. Please comment on my thoughts, whether it's right or wrong. They are all based on my opinion, and I have never tried them. So that before I start practicing on them, I want to make sure I am going the right way.

Time Breakdown

Cross: ~3 seconds
F2L: ~12 seconds
OLL: ~ 4 seconds
PLL: ~5 seconds

The Cross

According to my time breakdown, I average around 3-4 seconds. However, sometimes I take more than 5 seconds to solve the cross. This usually happens when I get a lot of bad edges which costs me up to 10 moves. To top that, bad crosses like this also affects my F2L because I get a little amnesia after solving the cross, which increases my F2L by around 2 seconds.

So I think to avoid these bad crosses, I have to become color neutral. Currently I solve with blue is my cross color (because I started learning from Macky's site and he uses blue). IMO, by being color neutral, I can choose to start with the color that has the most good edges/requires the least move to solve, and thus avoiding the bad cross cases entirely. I'm not really sure if being color neutral helps in Fridrich. If it helps, then I will not hesitate to be one, even though it will probably take a long time to get use to, considering the fact that it will also affects my F2L recognition. To top that, I think being color neutral will help me in the future when I start doing X-Cross (or maybe I should start learning it right now?) because it requires block building skills, and I know that being a color neutral is essential in using block building methods (Roux, Petrus, etc).

The F2L

There is not much I can say about how I think I should improve F2L other than practice. With the average of around 12 seconds on F2L, I can say that it takes me around 3 seconds to solve each pair. So basically, I need to shave off 1 second from each pair. I think I am able to improve myself by learning alternative algorithms for solving F2L pairs from different angles, since some cases are really awkward to solve unless the cube is rotated, which slows me down (and sometimes I drop my cubes because of cube rotations ). Furthermore, IMO I should start using look ahead as an 'active tool' (I forgot where I heard this phrase from..), that is multi-slotting. Multi-slotting adds extra moves in order to save up more moves in solving the next pair. Since I always hear the advice of 'going slow and look ahead' for F2L, I think it would be better to cut down moves first to reduce time, not going faster, which actually makes it hard to look ahead.


The OLL and PLL

I combine the OLL and PLL section because I think they have something in common. They both depends on finger tricks and cube. I think just by buying a better cube and training my finger tricks, I should be better at them, although there are other things I have in mind about them. Currently I'm using a store-bought, even though I have a DS, because it looks like this now:

Obviously it needs to be restickered (it's painted) because recognition is so hard (one of the blue edges is actually 90% white right now).

Currently, I'm using 2-look OLL because I couldn't be bothered to learn 50+ algorithms for 1-look OLL. I'll do that after I break the sub-15 barrier (wow, so optimistic ). So the time I take to solve OLL depends on the cases I got. I want to ask you guys about doing the first look OLL (getting a cross on top). Should I avoid the 'dot' case entirely by last layer manipulation through F2L, or just don't care about it. Because currently I'm using an 11-move algorithm to solve the 'dot' case, compared to just 6 moves for the other 2 cases (the 'L' and the 'straight line' case).

Nothing much to say about PLL. It's all about the cube, finger tricks and the right algorithm. There are cases I love (U perm) and I hate (E perm). If I have the right combination of this 3, I should be good to go (tell me if I'm wrong).


Conclusion

That's all I have got to say for now (that's a lot of stuff ). I actually have a lot more to say, but it's getting late. So, I hope you guys can give your opinions about how I think I should break the sub-20 barrier. Maybe some of the stuff are just completely useless to learn, or it's too early to learn it.

 

[JLarsen]

The thing most people don't understand I think is practice. Now I know that being told to practice can get annoying after sometime, but eventually you'll see that you're over analyzing it. People ask so many questions on how to improve, but in the end, all that matters is that you practice, keep with it, and you will improve. Sadly this kind of realization came for me after sub 20. I would leave it at what are the parts of my solve I should work on, then keep the weak sections in mind and focus a little more on them. I don't use fridrich personally so someone help this guy out, tell him what part of the solve he might want to work on and he'll be golden. Good luck on getting sub 20.

 

[DAE_JA_VOO]

I think you're getting a little TOO advanced too soon. I'm averaging around 21-22, and my best avg is 20.44, and I don't do any multi-slotting, nor am I colour neutral. Furthermore, my slowness is because my look ahead needs to improve MORE. Using your blue cross, a good look ahead, 2 look OLL, and 1 look PLLs, you should EASILY be able to get sub 20. No need for anything more.

Like Pandaman says, practice is key. Go SLOW, look ahead, and practice practice practice.

 

[Fishcake]

I think you're getting a little TOO advanced too soon. I'm averaging around 21-22, and my best avg is 20.44, and I don't do any multi-slotting, nor am I colour neutral. Furthermore, my slowness is because my look ahead needs to improve MORE. Using your blue cross, a good look ahead, 2 look OLL, and 1 look PLLs, you should EASILY be able to get sub 20. No need for anything more.

Thank you for your opinion! Hmmm...looks like I shouldn't worry about those stuff now. Maybe when I want to sub-15, I'll come back to them. But do you think that OLL and PLL depends on your cube? I think this is when the cube's corner-cutting ability, lock-resistance and pop-resistance come into play because you don't really need to look ahead anymore (maybe you need it between the 2 OLLs) and just blindly execute an algorithm at the fastest speed using state-of-the-art finger tricks.

People ask so many questions on how to improve

IMO, I think they ask those questions to know how to practice. I solve my Rubik's quite often, like hours a day. But I don't really think just solving it is considered practice, and expect my time to slowly go down. I need some guidelines (or something) to follow. Or, maybe I'm wrong...

you're over analyzing it

Hehe, sorry, I just can't help it. I'm a nerd

 

[DAE_JA_VOO]

I think you're getting a little TOO advanced too soon. I'm averaging around 21-22, and my best avg is 20.44, and I don't do any multi-slotting, nor am I colour neutral. Furthermore, my slowness is because my look ahead needs to improve MORE. Using your blue cross, a good look ahead, 2 look OLL, and 1 look PLLs, you should EASILY be able to get sub 20. No need for anything more.

Thank you for your opinion! Hmmm...looks like I shouldn't worry about those stuff now. Maybe when I want to sub-15, I'll come back to them. But do you think that OLL and PLL depends on your cube? I think this is when the cube's corner-cutting ability, lock-resistance and pop-resistance come into play because you don't really need to look ahead anymore (maybe you need it between the 2 OLLs) and just blindly execute an algorithm at the fastest speed using state-of-the-art finger tricks.

Well, you still need to look ahead when doing last layer algs. I always look ahead while doing OLL. Otherwise once you've finished OLL, you need to do the looking

While a great cube and good finger tricks do aid, I personally feel that better technique is key. Give any sub-15 solver a crap cube and he'll STILL be much faster than you and I, simply because he has better technique.

 

[rahulkadukar]

Dude please resticker the cube. I think you will save a second there only

 

[gpt_kibutz]

Ok First of all I ave to say that I am not sub 20 yet. (I average around 20 and 21).
It was 5 or 6 monts ago when I was at 25's just like you. I think the only key here is to pratice. Definitely you DO NOT have to be color neutral (i think you will take longer to be used to neutrality). Just watch WR holders. I do not think that Erik or Yu are color neutral.
I suggest that you forget about multislotting. If you learned to make F2L intuitively you might look at the F2L cases and compare their solution on the internet to your solution. Sometimes they are just the same. I think that you can average sub20 without difficulty by using your own F2L algorithms and no multislotting (Because it will also take you time to get used to it).
Now, IMO (and not to be offensive) but your LL sucks. (Mine too). I think that averaging around 5 seconds to do a PLL is jus too much (I average around 3 seconds to do the PLL which is also too much).
Just practice to lower your PLL and do not learn OLL. (Maybe the dot cases only, but that only if you like...)
There are some guys who average around 16 and they know only the cross OLL's.
I guess thats it.
Remember, PRACTISE!

 

[mcciff2112]

I don't see anything wrong with learning the OLLs as long as you take it slow. Don't try to memorize them all in one week. I've been at it for a while, a good average for me is just under 20 seconds. I know a little more than half the OLLs and I've been learning for about 6 months now. Take it slow, learn them gradually. Also, about being color neutral, it really isn't necessary. I find it helpful sometimes to be opposite color neutral (I solve white cross, or occasionally yellow cross) because you still use the same four colors on the sides. Being color neutral can help, but recognition is extremely difficult to get used to. Just keep practicing like you are, you will eventually get there.

 

[Bomber]

This might be a fun way to look at at how to break this so-called 'barrier'. Treating other times as barriers, like 30 seconds or even 1 minute, work out what you did to get past them. I'm sure you did nothing special when you first started solving, you probably didn't think about what you had to do to get past the upper boundaries of solving. So why think too hard about breaking this particular barrier? Yes, everything is appears to have become a lot more advanced but don't let that confuse you. Learn the basics and work on the foundations of solving the cube. That means practice.

 

[JLarsen]

People ask so many questions on how to improve

IMO, I think they ask those questions to know how to practice. I solve my Rubik's quite often, like hours a day. But I don't really think just solving it is considered practice, and expect my time to slowly go down. I need some guidelines (or something) to follow. Or, maybe I'm wrong...

you're over analyzing it

Hehe, sorry, I just can't help it. I'm a nerd

Really, there isn't any special way we practice. We solve the cube. That's it! Even for sub 15 people, we don't analyze stuff very much at all. We might through out a "Hey I have a 12 sec f2l and a 20 sec avg is that fine" kind of deal but we just solve. If I have a bad section in my solve I might do a little bit of only that section to bring it up to par, but more than anything else, perfection will come with time. For me, that would be 2 years of solving.

 

[Daniel Wu]

Dude please resticker the cube. I think you will save a second there only

Haha. If I used that cube, my averages would go from ~18 to ~28.

 

[Neroflux]

Your plls are bad.

 

[PCwizCube]

Yeah, I agree, you might be going to advanced stuff too soon.

I average 16-17 seconds and I use one color cross, no x-cross, and I only know like 15 OLLs (LOL). My cross is like 2-3 seconds, F2L is finished at around 10-11 seconds, OLL finished at like 13.5-14.5 seconds, PLL finished at 16-17 seconds.

I think your cross is pretty decent, yeah sometimes I get really bad times because sometimes I get lots of bad edges. I guess you could go with color neutral, I haven't really tried it so I'm not going to say much about it. But I can usually get a good cross with one color.

F2L
Possibly learn algorithms for hard cases, but I wouldn't go with that. You should worry about look ahead. You've probably heard that a million times but you probably don't know what it really means. When you look ahead, solve the first pair. When you're solving the ENTIRE first pair, look at the second pair you are going to solve. Then during the ENTIRE second pair, look at the third pair you're going to solve, etc. While you are doing this, don't stop turning the cube at all, always keep on moving the cube at a CONSISTENT speed. IF you have to do one turn per second, then by all means do that for a consistent turning speed (really, do that, if you need it use a metronome to help you). Trust me, look ahead will help you improve your times, even when you're sub 15. Getting good at it can improve your times GREATLY!

OLL
I don't know about you, but I'm lazy and busy so I don't have really that much time to do OLL. I might bother to learn some now that it's the summer, but I'm still interested in F2L, getting better at look ahead. I'd rather spend a week getting better at F2L then spend like two weeks learning all the OLL algorithms (learning all the OLLs will probably shave only 2-3 seconds off your times). Either way, you'd probably improve the about same amount of seconds (in your situation).

PLL
Yeah, 5 seconds might be a little slow. If you haven't already, learn Full PLL and try to get a PLL Time Attack below 80 seconds or something like that.

For improving your times, I would work on the steps in this order.
1. F2L
2. PLL
3. OLL or Cross

And.... get a better cube lol. It seems like your storebought is worse than your Diansheng (Type E), and you have a painted DianSheng, which isn't that great. Get a better cube, or at least get some Cubesmith stickers (they can last almost a year if you practice like an hour a day), because your storebought stickers will wear off soon too.

 

[iggyzizzle]

Ok.

1. Don't try any crazy cross stuff until you can consistently get 8 or less moves per cross. It's possible. Try messing with cube explorer for help.
2. Learn to look ahead much better in F2L. I need to work on this a lot as well.
3. Full OLL DOES help GREATLY contrary to what many people believe. Shaving 3 seconds is very good, especially once you get sub-20.
4. Your PLLs need MUCH love and repetition. Sit down and repeat the PLLs until your fingers fall off, come up with different finger tricks to make you faster, look up different finger tricks to make you faster, practice them like crazy.

If you can learn full OLL and get all PLLs sub-2, then you will shave a good 5 or maybe even 6 seconds off your solve time. That's sub-20. Woo! And you haven't even mastered F2L or cross yet.

 

[PCwizCube]

3. Full OLL DOES help GREATLY contrary to what many people believe. Shaving 3 seconds is very good, especially once you get sub-20.
4. Your PLLs need MUCH love and repetition. Sit down and repeat the PLLs until your fingers fall off, come up with different finger tricks to make you faster, look up different finger tricks to make you faster, practice them like crazy.

If you can learn full OLL and get all PLLs sub-2, then you will shave a good 5 or maybe even 6 seconds off your solve time. That's sub-20. Woo! And you haven't even mastered F2L or cross yet.

Haha you do make a good point, but I wouldn't take that approach that way, just because I'm lazy and don't have that much time

I guess you could learn the additional 40+ OLLs, recognize them, remember them, get fast at them. To shave off 2 seconds your OLL stage (not 3, 4-2=2) it'd probably take you a long time to get really good at the OLLs. I don't have time for that lol (well maybe I do now since its the summer)

You said your storebought was bad and Dianshengs suck at cutting corners so PLLs might be a little harder to get good at. It's possible though but I would think it'd be harder.

Bleh. I don't know about you but doing all that from your situation would probably take me like a month or more. You'd be learning all lot of algorithms and practicing them a lot with recognition and stuff so that's a lot of work.

I'd rather go with the F2L approach of looking ahead but iggyzizzle's idea is not bad, just that having a bad cube might be hard for you.

 

[Fishcake]

but your LL sucks

Your plls are bad.

Yeah, 5 seconds might be a little slow.

Haha, I do realize that. I know full PLL, and I learned it from a website by some cuber (I forgot) which came with videos, and I visually memorized them (not the notations). I think I have to find algorithms that suit me, and oh, a better cube, and lots of practice. I'll try to get a DianSheng 222 from a mall tomorrow (hopefully it's there), because people say it's a whole lot better than the 333.

Don't try any crazy cross stuff until you can consistently get 8 or less moves per cross

I solve the cross with 8 or less moves most of the time. Except when I get bad crosses. You know, a lot of bad edges. And in these cases, I can only track 3 of the edges, and I just search for the last one after solving the other 3.