Hand scrambles vs Computer scrambles

Faz

Former Clock NR Holder
Joined
Apr 24, 2008
Messages
4,250
Location
Melbourne, Australia
WCA ID
2009ZEMD01
YouTube
Visit Channel
Hey,

I currently use computer scrambles printed off a sheet.

I use jnetcube too.

This morning i hand scrambled it myself and got a 27.53!! Yay!

I am not sure whether to add this to my pb's as it was an easy cross (3 moves)

Should i or should i not make this my single pb.
 
Well, for me, hand scrambles are always about 30-60 moves with really bad cross and usually bad first F2L pair. When I do the Sunday contest, I'm surprised at my times and how easy the cross is.
 
I usually scramble untill I don't immediately see any connected pairs. For the most parts I find easier 2x2x2 blocks on computer scrambles.
 
I just scramble randomly.... I don't care if there are blocks... I use them even. But usually I don't count my hand scrambles...(got all my records with com scrambles anyway) usually I never make a hand scramble with over two pieces in the right place.
 
I find that my hand scrambles are biased towards simple triggers and i end up missing some faces (i don't like doing L or D' turns)
 
Hi :-)

This is also an old topic. When hand-scrambling it is easy to fall into the habit of just doing "semi-random" moves. One does a few good scrambling moves, but often one just repeats them unconsciously. This may lead to poor scrambles even with a high turn count. On the other hand cautious hand-scrambling may be very good. One deliberately breaks any good blocks forming by visual inspection. A good hand scramble is comparable to deliberate fully random compouter based scrambles (permutation- not turn based). Ie a computer scramble where you set a random state and find how to set it up. Not the other way round by just doing random turns. There has been a few analysis done that indicate a much higher turn count should be used for normal computer scranbles to achieve truly random scrambles.

My personal opinon is that we only need fair scrambles not "hard" scrambles - as far as competition goes anyway.

-Per
 
If you can solve the cube in 7 moves each time, why aren't you under 5 seconds yet?

Anyway computer scrambles and hand scrambles are about the same, but I've noticed that good hand scrambles tend to have fewer blocks because I consciously try to break them. This is especially true on 5x5, where computer scrambles give me 2x2 blocks of centers quite often, but I very rarely have them when I hand scramble. It's neat stuff.
 
Some of you noted that hand scrambling misses face turns, so I suggest doing double layer turns (that's what I do). Also, I personally practice with hand-scrambles using visual inspection (breaking blocks/roundabouts/similar patterns). Computer scrambles are too slow for me and i occasionally do L instead of L' or so.
 
When I hand scramble, I probably favor certain faces so I also rotate the entire cube at random every 10 or so twists about x or y or z, all the while not looking at the cube. I make sure I do 180 degree, CW and CCW twists. I stop after maybe 40 turns or so, and do not care what the cube looks like as I go to solve it. If I intend to record my times (like for Sunday Contest) I use the computer scrambles provided.

Obviously for 4x4 and 5x5 you need some double layer twists. (I am focusing on 3x3 these days.)
 
Back
Top