Seconds Per Piece

[elimescube]

So I'm coming to the end of another semester: Final projects are looming and I have way too much to do. So what do I do? Play with some cubing stats of course.

I've very recently started practicing 4x4x4 again, a puzzle I have not taken seriously in years. I thought it might be interesting to compare my other official times (especially 5x5x5, since I've practiced it so much more) to my official 4x4x4 times on a "seconds per piece" (spp) basis. Then I thought I might as well compare my results to the current WR times to see how my event focus compares to the focus of the top cubers in each event.





It doesn't surprise me that my spp for 4x4 is almost as slow as for my 5x5. It did surprise me a little that there is also a really small difference between the spp for the 4x4 and 5x5 WR's though (0.629 and 0.652 respectively).

I also had a look at the percent difference between my PB spp and the WR spp (which I acknowledge is no different than finding the percent difference in just time) and found, to my horror, that my events ranked best to worst compared to the WR's are 6x6x6 (by a long shot... :l), 7x7x7, 4x4x4, 5x5x5, 3x3x3, 2x2x2. Conclusion: I am not very good at cubing even at events I practice.

I'd like to see if anyone in the community can tell us some ways to use ssp for interesting analysis.

 

[qqwref]

After I finished solves of 12x12x12 through 20x20x20, I did an interesting seconds-per-piece analysis of my computer bigcube records: http://www.speedsolving.com/forum/showthread.php?31692. I noticed some interesting things, like how my seconds/piece increased slowly until about the 11x11, then made a huge jump and didn't increase much more after that. I've since improved some of the records, and now the graph up to 11 is smoother, with a jump to 1.3x for 12 and 13, and then 1.8x-2.1x for the remaining cubes.

 

[Stefan]

I can't read those tiny pictures, do you have larger versions?

 

[elimescube]

After I finished solves of 12x12x12 through 20x20x20, I did an interesting seconds-per-piece analysis of my computer bigcube records: http://www.speedsolving.com/forum/showthread.php?31692. I noticed some interesting things, like how my seconds/piece increased slowly until about the 11x11, then made a huge jump and didn't increase much more after that. I've since improved some of the records, and now the graph up to 11 is smoother, with a jump to 1.3x for 12 and 13, and then 1.8x-2.1x for the remaining cubes.

I did not realize how outrageously good you were at computer cubes. That's just insane...
I like how your seconds-per-piece drops down again slightly for 19x19 and 20x20. Do you just solve those more often? In truth I was hoping to see something similar between my 4x4 and 5x5 times, or maybe even a bigger deal like a drop in seconds-per-piece from 6x6 WR to 7x7 WR. No such luck.

I can't read those tiny pictures, do you have larger versions?

 

[qqwref]

I did not realize how outrageously good you were at computer cubes. That's just insane...
I like how your seconds-per-piece drops down again slightly for 19x19 and 20x20. Do you just solve those more often? In truth I was hoping to see something similar between my 4x4 and 5x5 times, or maybe even a bigger deal like a drop in seconds-per-piece from 6x6 WR to 7x7 WR. No such luck.

When I posted that topic, I had only done the 12x12 through 20x20 on the keyboard sim (qCube) once each. I did a few much earlier, slower solves on mouse simulators, but they took much longer. I did the solves in order, though, and I think I was getting used to the amount of handshifting involved by the very largest ones.

It's pretty interesting to me that the 4x4x4 and 5x5x5 solves are so close to the same sec/piece. I expected to see a constant increase... neat stuff. Perhaps it's because the 4x4x4 WR could be substantially improved (given people's best times in practice) whereas the 5x5x5 WR is very good compared to unofficial times?

 

[elimescube]

It's pretty interesting to me that the 4x4x4 and 5x5x5 solves are so close to the same sec/piece. I expected to see a constant increase... neat stuff. Perhaps it's because the 4x4x4 WR could be substantially improved (given people's best times in practice) whereas the 5x5x5 WR is very good compared to unofficial times?

I know that there were two major factors that influenced my decision to practice 5x5 more than 4x4. I hated how the parities could ruin a solve with an awesome reduction. I'm sure this has influenced at least some other peoples devotion to odd cubes vs. even cubes. I also saw an opportunity to do better in 5x5 on a global scale, which I'm sure you noticed in your rankings back in early 2008, pre-V-Cubes.

I agree with your thoughts on comparing unofficial capabilities with competition performance... 5x5 WR is just revoltingly far ahead of 2nd place, while 4x4 WR has much closer runners up.

 

[Goosly]

It's pretty interesting to me that the 4x4x4 and 5x5x5 solves are so close to the same sec/piece. I expected to see a constant increase... neat stuff. Perhaps it's because the 4x4x4 WR could be substantially improved (given people's best times in practice) whereas the 5x5x5 WR is very good compared to unofficial times?

I blaim that on the parity. I think 7x7 WR can 'easily' be improved, if there would just be a better cube on the market. (Hopefully ShengShou will produce a 7x7)
When we take that in account, the pps for 7x7 compared to the 6x6 will be relatively the same as 5x5 to 4x4. Which indicates it can indeed be blamed on the parity.

 

[elimescube]

I blaim that on the parity. I think 7x7 WR can 'easily' be improved, if there would just be a better cube on the market. (Hopefully ShengShou will produce a 7x7)
When we take that in account, the pps for 7x7 compared to the 6x6 will be relatively the same as 5x5 to 4x4. Which indicates it can indeed be blamed on the parity.

That's possible I'm sure. However one of the things that slows 7x7 tps is definitely its awkward size, which probably won't be easy to improve. I wouldn't mind seeing a ShengShou 7x7 on the market either as long as they go pillowed too.

 

[qqwref]

It might be possible to make a smaller, non-pillowed 7x7x7... but the pieces would be pretty small. It would not be easy to prevent it from breaking.

 

[Cheese11]

It might be possible to make a smaller, non-pillowed 7x7x7... but the pieces would be pretty small. It would not be easy to prevent it from breaking.

I personally don't think it's worth it though...

 

[qqwref]

Hey, if it improves people's times, it's definitely worth it.

 

[Cheese11]

Hey, if it improves people's times, it's definitely worth it.

Ohh, I thought you meant that it would be hard to keep from breaking while solving...
I'mm pretty much stupid.