• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

3-Style Corner Alg Generator

Keroma12

Member
Joined
Jul 28, 2010
Messages
656
Location
Vancouver, BC, Canada
WCA
2010MATT02
YouTube
Visit Channel
(Updated on 2018 05 08: see bottom of this post for new version)
(Update on 2019 01 22: related new tool, Corner Twist Finder, see bottom of post for link/details)


Link: http://csclub.uwaterloo.ca/~krmatthe/corner-search.cgi

From the page itself:

"This page should generate almost all regripless 3-style algs for a given corner case, with setups no longer than the provided length (maximum 8). Direction is not considered, so you may need to invert some algs. Regripless is not entirely well-defined, so a few of the generated algs may be a bit awkward, and some algs which you feel are regripless may not appear. Contact me if you have an alg that you think should be appearing but is not, and vice versa.

For each alg, the length is displayed (counting, for example, U D as a single move, R2 as two moves, U2 D' as two moves), and the algs are roughly sorted from best to worst, though this sorting isn't particularly great. Ideas for improving the sorting are welcome.

Disclaimer: I'm not claiming that the best alg for a case is always regripless."

Last updated: 2017 12 02 (see bottom of actual page for a more accurate date)

~~~~~~~~~~~~~~~

Update 2018 05 08: There is a new, much improved version now.

Link: http://csclub.uwaterloo.ca/~krmatthe/3-style-corners.cgi

The search depth now specifies how many setup moves allowed in axial turn metric, direction of the case is accounted for, the case you searched is in the url and hence shareable, and most importantly, it now finds lots algs which have a have a minor regrip in the setup (so not just regripless).

From the page itself:

"This page should generate most nearly-regripless 3-style algs for a given corner case, with setups no longer than the provided length, in axial turn metric (maximum 4, which takes several minutes). You can select if you only want regripless algs.

For each alg, the length is displayed (counting, for example, U D as a single move, R2 as two moves, U2 D' as two moves), and the algs are roughly sorted from best to worst. There is a * before the alg if the setup has a regrip. Ideas for improving the sorting are welcome."

~~~~~~~~~~~~~~~

Update 2019 01 22: Corner Twist Alg Finder

Link: http://csclub.uwaterloo.ca/~krmatthe/corner-twist.cgi

This new page uses the above page to find two 3-style algs which twist corners. It takes a long time (several minutes) at maximum search depth (see text below). If you want a search depth of 3 for a 2-twist, you could do 6 individual searches where you specify which third corner must be involved in the 3-style cases.

From the page itself:

"This page should generate most nearly-regripless twist algs for a given 2-twist or 3-twist corner case that are made from concatenating two 3-style cases, with setups no longer than the provided search depth in axial turn metric (maximum 2 if specifying 2 corners, maximum 3 if specifying 3 corners; these may take a couple of minutes). You can select if you only want regripless algs.

For each alg, the length is displayed (counting, for example, U D as a single move, R2 as two moves, U2 D' as two moves), and the algs are roughly sorted from best to worst. There is a * before the alg if one of the setups has a regrip. Ideas for improving the sorting are welcome.

Type in the location of the U or D sticker on the twisted pieces. For a 2-twist, either leave the third spot blank or type in another solved corner (starting with U or D) to only search for comms involving that corner."

The page itself will be updated, I probably won't update this post much about it.
 
Last edited:
U

Underwatercuber

Guest
Link: http://csclub.uwaterloo.ca/~krmatthe/corner-search.cgi

From the page itself:

"This page should generate all regripless 3-style algs for a given corner case, with setups no longer than the provided length (maximum 8). Direction is not considered, so you may need to invert some algs. Regripless is not entirely well-defined, so a few of the generated algs may be a bit awkward, and some algs which you feel are regripless may not appear. Contact me if you have an alg that you think should be appearing but is not, and vice versa.

For each alg, the length is displayed (counting, for example, U D as a single move, R2 as two moves, U2 D' as two moves), and the algs are roughly sorted from best to worst, though this sorting isn't particularly great. Ideas for improving the sorting are welcome.

Disclaimer: I'm not claiming that the best alg for a case is always regripless."
Cool idea but I think it still needs a bit of improvement before it can get really nice algs. I will still be making my own but I might check up occasionally on this to see if it generates a better one.
 

Keroma12

Member
Joined
Jul 28, 2010
Messages
656
Location
Vancouver, BC, Canada
WCA
2010MATT02
YouTube
Visit Channel
Cool idea but I think it still needs a bit of improvement before it can get really nice algs. I will still be making my own but I might check up occasionally on this to see if it generates a better one.

From asking various people who have great algs for each case, they've typically said that 40%-55% of their algs are ones which would appear here.

I agree that there are lots of cases where this won't give the best alg. There are lots of cases where you can regrip while doing other moves, which makes the alg almost essentially regripless. The point of this wasn't to find the best alg for each case, just list all regripless algs (and roughly sort them).

Did you have a particular bit of improvement in mind? I'm happy to look into improvements.

Thanks for the feedback!
 

Keroma12

Member
Joined
Jul 28, 2010
Messages
656
Location
Vancouver, BC, Canada
WCA
2010MATT02
YouTube
Visit Channel
Now featuring 50% more comms! (literally)

It does rarely display the same comm twice, in different formats. I'll fix that (and some other key problems) soonish.

Edit: Fxed those things. No plans to change anything in the near future.
 
Last edited:

Keroma12

Member
Joined
Jul 28, 2010
Messages
656
Location
Vancouver, BC, Canada
WCA
2010MATT02
YouTube
Visit Channel
New Page 2019 01 22: Corner Twist Alg Finder (also in the original post now)

Link: http://csclub.uwaterloo.ca/~krmatthe/corner-twist.cgi

This new page uses the original 3-style corner alg finder to find two 3-style algs which twist 2 or 3 corners. It takes a long time (several minutes) at maximum search depth (see text below). If you want a search depth of 3 for a 2-twist, you could do 6 individual searches where you specify which third corner must be involved in the 3-style cases.

I don't think it's very useful for 2-twists, after having tested it out (for hours on my computer with bigger search depths) and compared to standard algs. It might be useful for 3-twists. Or it might be useless.

From the page itself:

"This page should generate most nearly-regripless twist algs for a given 2-twist or 3-twist corner case that are made from concatenating two 3-style cases, with setups no longer than the provided search depth in axial turn metric (maximum 2 if specifying 2 corners, maximum 3 if specifying 3 corners; these may take a couple of minutes). You can select if you only want regripless algs.

For each alg, the length is displayed (counting, for example, U D as a single move, R2 as two moves, U2 D' as two moves), and the algs are roughly sorted from best to worst. There is a * before the alg if one of the setups has a regrip. Ideas for improving the sorting are welcome.

Type in the location of the U or D sticker on the twisted pieces. For a 2-twist, either leave the third spot blank or type in another solved corner (starting with U or D) to only search for comms involving that corner."
 
Top