• 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!

Mindcuber vs. Human Experiment?

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
So as an experiment for a science fair, I'm recording the speed of the MindCuber robot's solving ability and comparing it to human solving ability. I was hoping to have some participation from people who actually know how to solve the cube. If anyone is willing to participate, know that you don't need to be a genius at solving, honestly, you could take as long as you need, I just need some results.

Basically, you will need your own Rubik's cube and two stopwatches.
1. Scramble the cube, and don't look at it.
2. Start the first stopwatch as you begin to examine the cube, and when you are ready to start solving, start the second stopwatch.
Both should be running during the time you solve.
3. When you finish, stop both timers.

If possible, you could also count the moves as well (I can't solve Rubik's cubes so I don't know how difficult this is), whether you count as you solve or video record it and count later.

Attempt as many trials as you wish.

If you decide to try this, your results would look something like this:
Name (optional, for recognition during science fair)
Total time (this comes from the first timer): 5:06.19 (or whatever time shows up)
Solving time (from the second timer): 4:43.78 (etc. etc.)

I know this is a bit much to ask for, but this site was one of the only places I thought to turn to for this kind of demonstration. I do hope someone decides to try. For science! Thank you in advance! c:
 

Lucas Garron

Administrator
Joined
Jul 6, 2007
Messages
3,718
Location
California
WCA
2006GARR01
YouTube
Visit Channel
It would probably make the most sense to modify a web timer to record inspection.

How "accurate" does your data need to be?
You could get a lot of data from YouTube by just counting the inspection time of lots of videos (some timers will display it on the screen, often to the nearest second).
It might even be more useful to get data from solves thataren't done specifically for this data collection.
 

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
For accuracy, it doesn't exactly have to be to the millisecond, but most stopwatches do record to the second, and that is perfectly fine as well.

I have considered using solves posted by Youtubers as part of the overall data, since they are recorded to the millisecond.

I was hoping for a variety in statistics. For example, I have a few friends who will be recorded, and along with the Youtubers, and (hopefully) people from here, there is a good assortment there.
 

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
Preferably, from now to Thursday would be most favorable. I would appreciate your help, thank you. C:
 

Hypocrism

Member
Joined
Aug 4, 2012
Messages
316
WCA
2009ADLA01
Well most of us can tell you straight up what we average over 1000's of solves with 15s inspection. Would you prefer that...?
 

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
I do think that would help as well, though I'm mainly trying to collect the timings for single solves, rather that the averages of thousands.
 

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
Well, it is a 9th grade science fair, so I'm not exactly trying to be the most scientific, so to say. Besides, the point was to compare the single scores for the solves. For the Mindcuber scores, it is timed per solve, not as an average, so I thought it would only make sense to record the human scores the same way. You probably know way more than I do, though, so if you have any suggestions concerning the alteration of the procedure, I am open to any and all assistance. :)
 

Carrot

Member
Joined
Feb 9, 2009
Messages
1,910
WCA
2008ANDE02
YouTube
Visit Channel
a human solve requires more inspecting integrating IN the solve, while a machine would find the whole solution in one go. I think this is more comparable with blindfolded solving where the human can't look at the cube after starting the solve. (just like the machine)
 

DeeDubb

Member
Joined
Mar 9, 2014
Messages
1,432
Location
South Korea
WCA
2014WHIT07
YouTube
Visit Channel
You should have a cuber help you so you know more what you are looking for. If I were you, I would post a scramble (which means a list of moves that we can use on our cubes at home to fully scramble it) have the robot solve the same scramble, then we can post vids of us scrambling, timing our inspection and timing our solve. This will give you a much more scientific result, as we are all solving from a level playing field. Sometimes we get lucky scrambles that can make a considerable difference to our times, so if we all are working with the same one, it will give you a fair result.

Also, you are going to get a wide variety of human results. I'm a neophyte at speed cubing, so my solve will take considerably longer than someone who has been cubing for a while. I'm not sure if that's what you're hoping for, or if you just want to see the best possible human results compared to the robot.
 

Hypocrism

Member
Joined
Aug 4, 2012
Messages
316
WCA
2009ADLA01
Well, it is a 9th grade science fair, so I'm not exactly trying to be the most scientific, so to say. Besides, the point was to compare the single scores for the solves. For the Mindcuber scores, it is timed per solve, not as an average, so I thought it would only make sense to record the human scores the same way. You probably know way more than I do, though, so if you have any suggestions concerning the alteration of the procedure, I am open to any and all assistance. :)

Well, you have to explain what it is you're trying to find out first-what is your hypothesis or research question? Then I'll give you some ideas about the scientific basis.
 

demonpowder

Member
Joined
Dec 4, 2013
Messages
7
You should have a cuber help you so you know more what you are looking for. If I were you, I would post a scramble (which means a list of moves that we can use on our cubes at home to fully scramble it) have the robot solve the same scramble, then we can post vids of us scrambling, timing our inspection and timing our solve. This will give you a much more scientific result, as we are all solving from a level playing field. Sometimes we get lucky scrambles that can make a considerable difference to our times, so if we all are working with the same one, it will give you a fair result.

Also, you are going to get a wide variety of human results. I'm a neophyte at speed cubing, so my solve will take considerably longer than someone who has been cubing for a while. I'm not sure if that's what you're hoping for, or if you just want to see the best possible human results compared to the robot.

AAAHHH I wish I had seen this before I started. I'm a bit short on time, so I probably won't be able to redo the hour and a half I spent sticking randomly scrambled cubes into the robot and just timing. :/ This would have been so much better of an experiment. I assume it's too late now, though, so I'll just come up with some random impromptu stuff and make the board look cool. -__-
 

Antikrister

Member
Joined
Dec 8, 2012
Messages
61
Location
Sweden
a human solve requires more inspecting integrating IN the solve, while a machine would find the whole solution in one go. I think this is more comparable with blindfolded solving where the human can't look at the cube after starting the solve. (just like the machine)

Dude, just go with this. Just compare your robots result with that of the best blindfold solvers.
 

Bindedsa

Member
Joined
Sep 8, 2013
Messages
2,035
Location
New York
WCA
2014NURU01
YouTube
Visit Channel
Dude, just go with this. Just compare your robots result with that of the best blindfold solvers.
+ 1

As I side note, I assume you have now knowledge of cubing, let alone Bldinfolded cubing, so you might find this information useful.

A while back I wrote this tutorial [A basic blindfolded solving method] solves the cube piece by piece. We solve one piece at a time with edges, then solve one piece at a time with corners. In this we do not mess up any other part of the cube other than the pieces we want moved. This is done by some algorithms.
The whole post is here: http://www.speedsolving.com/forum/showthread.php?t=8653

Basically, you use sequences of moves to solve each piece by piece, swapping it from one place to another without disturbing the other pieces. More advanced blindfolded solving methods will swap more pieces at a time. Blindfolding solving is generally less efficient, in terms of moves, than other speed solving methods and even less efficient that computer solutions. The memorization process just tracks the order in which pieces should be cycled around and is rarely, more than 25 different swaps to remember, often using letters or words linked to specific places on the cube. If anything is wrong please correct me, like I said I don't blind solve.
 
Top