• 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 35,000+ people from around the world today!

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

Optimal Edge Execution [theorycrafting, discussion]


Feb 23, 2016
San Jose, California
These all assume you perform M2s with either a pinky->ring, ring->pinky, L l'M', or R' rM' move. This is because these are the M2s that free your index (as opposed to something like ring middle, which traps your index and frees your pinky), meaning it is quicker to execute comms because there is no pause between the M2 and insert.

If you do the cycle DF UB FR, with a lefty M2 (again, one of the aforementioned ways), it is regripless, because you do the insertion with your right hand. If you do the cycle DF UB FL, there is a small regrip, because you have to insert with the same hand with which you do your M2. If you time these two comms you should hopefully see a difference of 0.1-0.2s.

This got me thinking, is it optimal to learn the mirrors for every comm? For instance using a righty M2 flick (or an R' rM') for DF UB FL, making it regripless. These comms should theoretically be faster to execute (assuming you practice them), because of the lack of a regrip. However, there are a few issues I have encountered.

1) Transitioning from lefty M2s (or M' or M or whatever) to righty M2s between comms requires a regrip, which would potentially cancel out / defeat the point of using a regripless comm. For instance if you do DF UB DR, it sets you up to more quickly continue with a lefty M2 - because your fingers naturally position there at the end of the comm.

2) Point 1) is further complicated by the fact that due to edges' being symmetric, there are an (approximately) equal amount of comms that end in a spot where it is faster to continue with a lefty M2 as there are with a righty M2, meaning you wouldn't be doing any more regrips than you would if you only did lefty M2s.

3) Point 3) is even further complicated because the split between comms is not 50/50. If you only do a lefty M2:

The cycles DF-UB-FR and DF-FR-UB set up to another lefty M2
The cycle DF-UB-FL sets up to a righty M2 instead (naturally, as it is mirrored).

However, the cycle DF-FL-UB sets up to a lefty M2, because you end on the interchange meaning your left hand is automatically positioned to perform another M slice move. This implies that if you perform every edge comm with lefty M moves, there are slightly more comms that will end on another lefty M2, meaning you will be doing <50% regrips, whereas in the mirrored situation approximately 50% of your comms will have a regrip.

This makes it very complex to figure out the optimal way to do edge execution. I assume a true optimal method (ignoring 5style, floating buffers, etc) would be to practice mirrored fingertricks, and perform the comm based on the position your fingers were left from the last comm.

This however, requires very good thinkahead. Finally, you would have to individually check each case if it would be faster to perform the slower comm (where you don't have to regrip between the comms), or to regrip first between the comms, and to perform the mirrored, pauseless comm. This itself would be a feat because it is extremely hard to measure the differences between the two.

I've been sitting at my computer for about 10 hours now doing comms over and over, to try to determine the best way to go about this, and would greatly appreciate anyone's thoughts on the matter.

Daniel Lin

Woah this is waaaay too comprehensive

Imo it doesn't really matter because you as you said you either have to regrip in the middle of a comm or between comms.i do mirrors just because its natural for me( I'm comfortable with lefty and righty M2)
Also it should be exactly 50/50, cuz theyre mirrors right?

I think anoyher thing we could implement into 3style is learning multiple comms per case fordifferent angles. I actually use that for corners. Say you get the case UBL RUB DBL. Normally I do the alg you showed me R2 D R' U R D2 R' U' R2
But if I'm off by a z rotation (from executing the previous comm) i do R2 U2 R' D' R U2 R' D R'
I think avoiding rotation s is more important than avoiding regrips
Apr 23, 2010
Scotland, UK
I don't change what hand I use for the M2' for UB->FR and UB->FL, but I don't regrip either. I guess you never considered that option because of your turning style? I've always used my left hand for M slice stuff, and it works fine for me so I don't intend to learn to do it mirrored. Will try to make a short video later if there's interest in how I fingertrick them.

Edit (video was made quickly and not high quality but I don't care :p):
Last edited: