How to know if an algorithm is good?

[Argothman]

I'm starting to memorize PLL algs from the wiki, but how do I know which are good? There is the element of how they feel, but anything r/u/l/m isn't a problem and some of the other moves are still fine. Also, are M move algs better than their longer counterparts?

 

[cyoubx]

Really depends on the cuber and the cube. It depends on YOU for which algorithms feel comfortable, and it depends on the cube to see whether M moves are worth it (bad for locky cubes). It's just preference.

 

[cubefan4848]

When I was learning PLL from the wiki I looked at the moves I liked doing most which were R and U moves. So I found a PLL with mainly R and U moves and memorised. So of them though are a lot longer than the others and I probably wouldn't learn them.

 

[FatBoyXPC]

It really all boils down to cubing style. This should have been put in the one answer question thread, though. Most of the common PLLs are the first couple algorithms under each PLL. Go watch the youtube videos for PLLs, a lot of the more popular (and even not so popular) cubers have videos of them on their youtube.

My best advice is to find an alg that suits you and find finger tricks for it. If you can't find any alg that's suitable keep playing with the different algs, from different angles, different grips, etc. I believe it's Teller who said the easiest way to find finger tricks for an alg is to do it backwards.

Edit: Ouch, double ninja'd!

 

[cyoubx]

Edit: Ouch, double ninja'd!

Nah, you weren't really ninja'd, you brought up a good point about looking for videos on YouTube. I know Erik has a video on the PLL's he uses, and it helped me somewhat.

 

[iasimp1997]

1. Find an algorithm.
2. Perform it several times.
3. See if you like it.

And please use a title with a question mark in it so we don't think that this thread is a tutorial.

 

[cmhardw]

The only thing that should matter is average execution time. Time each alg for an average of 12, or even better an untrimmed average of 10, and pick the fastest one. Keep in mind that the algs will all get better as they merge more into your muscle memory. Generally, though, the fastest one right from the start will be the fastest one overall.

Don't go by how the alg feels. This is important. It can certainly happen that an alg that feels "cooler," or "faster," can actually be slower to execute than another alg that doesn't feel as nice.

--edit--
renamed the title slightly for clarity

Chris

 

[Joker]

Yes, it doesn't matter how good an alg feels. I like some algs better than others for feel, but they are slower. Get sort of good at a few algs, and time which one is fastest.
Don't go by feel, it doesn't help your time.

 

[Cyrus C.]

When I was learning PLLs I narrowed them down based on how good they looked and how move count friendly they were. The 3 or 4 that looked the best, I then tested to see which were best.

 

[Edward]

You just know man. Flow+Movecount.

 

[Lorken]

I find it easiest to just find a website with relatively new algorithms and just learn all of them, that way, it is a lot easier to learn because most of them are similar or really finger trick friendly. I learnt mine from bluenkightcubing.com and a few PLL's from badmephisto's site.

 

[FatBoyXPC]

I guess one argument that could be made is look at what the top 10 cubers use for PLLs,and use those. You know all those PLLs are definitely sub2able.

 

[Anthony]

I guess one argument that could be made is look at what the top 10 cubers use for PLLs,and use those. You know all those PLLs are definitely sub-1.5able.

fixed.

 

[oprah62]

I'm not trying to advertise or anything, but 90% of my algs come from speedcubing101.com

 

[Ranzha]

I generated a lot of my worse PLLs, like my Y-perm and Z-perm.
You can find a lot of cool algorithms for cases--you just need to find out which ones are easy to execute, and have a relatively low movecount.
What happens a lot of the time is one learns an algorithm, uses it for awhile, and find out it sucks. It really sucks. The person then either:
a.) tries to improve their times on that algorithm;
b.) learns a different algorithm from a site;
c.) generates a new algorithm.

My Y-perm was REALLY slow. It's the one a lot of people use (F R U' R' U' R U R' F' R U R' U' R' F R F'). I was getting 2.5s with it.
So I generated one using CubeExplorer (run a Google search if you want to learn more), and I found:
R2 U' R2 U' R2 U y' R U R' B2 R U' R'.
I now get 1.7 with it.

 

[FatBoyXPC]

Ranzha, I use that Y perm, and I my PB is 1.5 or lower I can constantly sub2 that Do you do the B2 with your right ring finger and do two separate pulls? That's the only easy I I was able to do it. I love how nice that R2 U' R2 U R U flows though, if only you could show me a better way to do that B2

 

[The Puzzler]

I generated a lot of my worse PLLs, like my Y-perm and Z-perm.
You can find a lot of cool algorithms for cases--you just need to find out which ones are easy to execute, and have a relatively low movecount.
What happens a lot of the time is one learns an algorithm, uses it for awhile, and find out it sucks. It really sucks. The person then either:
a.) tries to improve their times on that algorithm;
b.) learns a different algorithm from a site;
c.) generates a new algorithm.

My Y-perm was REALLY slow. It's the one a lot of people use (F R U' R' U' R U R' F' R U R' U' R' F R F'). I was getting 2.5s with it.
So I generated one using CubeExplorer (run a Google search if you want to learn more), and I found:
R2 U' R2 U' R2 U y' R U R' B2 R U' R'.
I now get 1.7 with it.

I was looking at your signature and saw your tshirts. I might buy some.

 

[KboyForeverB]

FAST TO DO OR, in my opinion, has no Bs, Ds or even maybe Fs in the MIDDLE of the alg

 

[RCTACameron]

I'm not trying to advertise or anything, but 90% of my algs come from speedcubing101.com

Right after Anthony posts. Ironic.

Probably try each alg for a case and do and average, and use the alg that gets you the fastest, as Chris suggested. If you can't decide, then see what other cubers choose by looking at their websites.

 

[Rpotts]

I don't think you can really know til you practice it/figure out how to do it right. Everyone performs algs slightly differently, sometimes you just have to learn to do execute it more fluidly. Like I used to suck at OLL 50 [r' U r2 U' r2' U' r2 U r'] but now that I learned how to air jeff the f*ck out of it i can sub1 it. : D