Sponsor Invitation
Snyder Cube Challenge

Tony Snyder's Cube Solving Hobby

Opinions Page

2 Million Solves
speedsolving wiki


As a first generation cube solver I have 40 years of experience solving the cube. I estimate that I have solved the cube 2.2 million times. As a result of these stats I tend to have a lot of opinions. Here at this page I list some opinions I have on things related to the cube.


This is generally a very frustrating issue with lots of ups and downs. When I was young it didn't matter that much, but as I got older, I noticed that the skin on my fingers produced less of the oils necessary to get that slight grip feel in a sticker that is necessary for a speed solve. Now the feel of a sticker tends to be way too slippery nearly all the time, kind of like grabbing a glass bottle with greasy hands, just can't hang onto it. At least for me this makes the sticker issue one of the most important issues in speed solving.

I would say that in optimal temperatures, optimal humidity, and optimal skin conditions, that the best sticker for speed solving is vinyl. This is because of the slight grip it has, like a layer of rubber, yet not quite that sticky. Vinyl is great if you need speed because you can grip it, and you can slip it, as needed.

PVC wears way way too fast. Granted, I solve the cube like a mad man sometimes several hundred times per day. And when I do this I wear out PVC stickers in roughly 4 days. Gradually chunks start coming off the edges until the stickers get smaller and smaller, and uglier and uglier. And replacing them every 4 days is a pain.

Cloth stickers are a total joke. They do not grip as well as vinyl, and yet they wear fast. Cloth generally turns white as the color rubs off, and sometimes they will peel out in layers. I would avoid these completely.

Stickerless cubes are the best given the right temperature. My Moyu is a bit too slippery on the outside at around 70F, and sticks just the right amount to my fingers at around 80F. I also think the pieces are more uniform and balanced with stickerless.

And, having experimented a lot with nutrition, I have discovered that an overly healthy dose of omegas each day tend to improve my finger oil issue to the point where I can recover most of the stick I had when I was young, especially if the temperature is at least 80F.

Cubesmith stickers are a big jump ahead in technology. Made of polycarbonate I can speedsolve non stop for a month and not see a bit of wear. However, after roughly 1-2 months there will be an occasional chip, and that's where things start to get bad.

Cubesmith stickers don't have the grip that vinyl has, they are smooth and slick. I tried textured and that was great for a few weeks until the texture wore off (a bit odd here, as you'd think polycarbonate would last longer than that). Then if that wasn't bad enough, when an edge of a Cubesmith sticker lifts up, it does not crumple harmlessly like vinyl or cloth. The polycarbonate sticker is razor sharp and very strong. It slices right through your fingers as you solve the cube.

Of course slicing through fingers is not all that pleasant, but it is the price to pay for the best sticker.

One way to not have to deal with stickers at all is to go with the Guhong cube, which is designed so that different colored plastic parts come together perfectly in forming the colors to each side. And I have found in practice that I get a better grip on this plastic than any of the stickers. In my opinion this is the best solution to the stickers issue (assuming I'm okay with the other qualities of the Guhong - see cube section below).

However, if I'm going to put up with stickers, then I go with Cubesmith textured polycarbonate, and replace them each month. And if I can't find these then I go with vinyl stickers.


I have tried a number of new and old cubes, plus built many cubes using parts from multiple cube brands. I'm continually experimenting to find ways to improve, and as a result my opinions will change from time-to-time.

For many years I considered the PGCO that I bought at Hobby Hive in 1981 to be the very best cube for speed solving. This cube though very stiff out of the box, could be shaved with a knife and lubed to obtain one of the smoothest operating cubes ever. Maybe too smooth, though I found with practice I could get used to that. The best tuned PGCO I ever had was so smooth I could tap a corner and watch the side rotate nearly 3 complete revolutions before stopping. It was not that great at corner cutting, but back then my fingers were accurate, so it didn't matter. I got my fastest average prior to 2004 with that cube: 15 seconds. It is also the same cube I solved in 11.5 seconds at the Chicago Cube-A-Thon in 1982.

For many years I considered the Rubik's brand cube one of the worst cubes, as I had practiced from the start using a PGCO, which had far better internal engineering. Plus Rubik's brand cubes required continual maintenance and would break easily. Then in roughly 2002 out came a much better cube. And even today, I like to hunt for Rubik's brand cubes from 2002 and 2003, as these seem a bit better than the ones since, and way better than the ones prior. I discovered that Long's Drugs stores everywhere carried this specific cube, so once my inventory of PGCO cubes were thoroughly worn out I started emptying out each Long's Drug store of their cubes. And then a few years later I discovered that contests no longer required Rubik's brand cubes, so I then started investigating the others.

It was not until I discovered the F cube that I switched again. This cube would cut corners better than the Rubik's brand cubes. Unfortunately, the center caps would pop off a lot, and if I glued them in I couldn't adjust the screws, leading to a bit of a catch-22. These caps once glued are on for good, because of the design which is better for gluing, they cannot be removed later to adjust the screws. Also, the F must be adjusted perfectly, otherwise it pops and grabs easily. However, once perfectly adjusted this speed cube has great potential. It cuts corners very well, it has predictable audible feedback, it is self lubricating (I didn't add silicon for about 2 months), and it requires less effort to turn than most cubes. However, the characteristics of this cube are quite different, so don't expect to get your best times on your first day with this cube - take your time getting used to it first. The large bevel on the edges produces different feedback for your fingers, and this may feel less certain at first.

Then a few months later I tried the Guhong. This is the first cube that seemed to correct my fingering problems. It's speed, balance and corner cutting were all excellent, plus it eliminated the frustration with stickers (see stickers section above). I also purchased a Lubix upgrade to this cube, and found that I could significantly improve my finger speeds. However, it did not work correctly with torpedoes (an innovation to keep edge pieces from popping). But this cube was so well balanced in characteristics (due to the superior manufacturing design and process) that I found the need for torpedoes greatly lessened.

Then I upgraded to a Magnetic MoYu WeiLong GTS V2 M. This cube has advanced features that are balanced across the board. Its much better on every aspect of speed solving. It is currently my best speedcube. Interestingly, it seems to require a lighter weight silicone lube. I'm still experimenting to get the weight right.

What Affects Speed?

Over the last few years tremendous advancements have occured in what's called modern fingering and look ahead. These advancements are responsible for the new much faster speeds that the top cube solvers around the world have attained.

Using the Fridrich system, modern fingering and look ahead may be optimized to a far greater extent, resulting in 10+ seconds savings on many cube solvers' times.

It appears that at this time no method comes close to the Fridrich method for speed. In fact very few can solve the cube in an average of less than 15 seconds using methods other than Fridrich.

However, the point that most people are unaware of is that Fridrich is at its limit, there is no way to improve it further. And yet other methods can still be improved. The question is whether or not there exists a method that can improve enough to overtake Fridrich.

As Fridrich averages 57 turns, and Snyder Method 2 averages about 40, it is clearly possible to save turns over Fridrich. However, it is probably not possible to overtake Fridrich look-ahead and fingering benefits. Therefore the challenge will be to reduce average turns to the point that in combination with improvements in fingering and look ahead, will produce the necessary net effect.

The goal of Snyder Method 3 is to produce a huge turn advantage combined with improved look-ahead, while simultaneously introducing a variant to modern fingering that works with this method.

I am convinced that Snyder Method 2 can come close to over taking Fridrich once software has been used to optimize the algorithms, the look ahead, and the fingering. And I'm convinced that Snyder Method 3 is capable of surpassing Fridrich in speed.

In the past I did not worry that much about these details. I had the passing thought that I could find ways to finger better, and to work on look-ahead methods, but I really didn't focus on speed that much since the 1980's.

Probably my best skill when I was young was fast fingering, it often came natural and I could at times burst to about 8-9 TPS (approximately one full algorithm in 1 second). However, I found it difficult to go this fast throughout a solve, as I was using my wrist a lot. Older fingering methods that use a lot of wrist action are now called "wristing". In contrast, newer fingering methods that reduce the number of times you have to use your wrist are called "modern fingering". The method I used in the past to burst to 8-9 TPS on some algorithms was quite a bit different from typical wristing methods, though it did use a lot more wrist action than modern fingering. I can remember times when my wrists would literally get hot to the touch after an hour or so of practice. Though I'm convinced my fingering method was nearly as fast as modern fingering I never made it work across the board with all algorithms, and therefore it only helped speed up the solves that utilized specific algorithms. However, I am currently convinced that it could be expanded to handle all algorithms, and to make possible consistent fast solves without turn sacrifices. And that this in combination with reducing average turns to 30 will result in faster speed solves. I'll fully document how to do this in Snyder Method 3.

Though most people tell me the problems I'm having with my fingers are due to aging, I'm convinced they are all fixable. If I do succeed in correcting these problems then I should be able to pick up about 5 seconds on my averages. Though I haven't timed myself yet, the Dayan ZhanChi has renewed my belief that this can be done with what feels like much faster solves.

There is also the issue of mental focus, which affects the time a cube solver takes at each look. Mental focus is necessary to correctly recognize the scenario and to produce the best algorithm for the next step, as well as to execute that algorithm. Better focus reduces mistakes and reduces the pause at each look. I have discovered that my mental focus tends to slip if I don't keep a balanced healthy lifestyle, and that this is more of an issue as I get older. Now it is very important for me to exercise each day and eat healthy foods and supplements. And even with all that I sometimes need to boost it more with energy drinks and caffeine.

Another factor affecting mental focus is the disadvantage that comes from evolving my method over time, and through an intense creative effort rather than pure memorization. Due to all this creative effort in the past, I often find myself distracted by too many thoughts during a solve. This causes more of a pause at each look. Oftentimes the advantages in using one algorithm over another are subtle, and result in an extra moment of indecision. This is one point where a cube solver that starts out learning an existing professional method will have an advantage over someone like myself who is used to continually weighing advantages of one algorithm over another during each solve.

I am also convinced that stickers play a big role in the speed of one's solves. I know for myself that as I get older my skin produces less of the oils necessary to get that slight stick that is necessary to a speed-solve (see Stickers).


I have several opinions about contests.
  1. They should include the Composite Challenge (CC) event. In my opinion this is the very best test of one's cube solving abilities. Once this event is added to the contests, cube solving should begin to achieve a much higher level of notariety, possibly someday matching chess.
  2. They should allow 20 seconds for study on the speed solve event, not 15. 15 favors too heavily the simpler solutions, making fast fingers a bigger issue than sophisticated algorithms. Or, like chess, there could be multiple events. I would suggest these: 1) a no-study speed solve, 2) a 20 second study speed solve (primary event), and 3) a 1 minute study speed solve.
  3. I notice there is a 1.5 meter rule, where people must stay at least that distance from a contestant during their solve, however, from the contest recordings I've viewed, this rule was clearly not observed. Instead there was constant traffic around the contestants. I know for me that would be an issue.
  4. More than just temperature, UV also seems to affect a cube a great deal. For example, I was outside one cloudy morning solving my cube, getting mediocre times over and over, then suddenly the sun came out, and within a couple minutes the cube started cooperating exceptionally well and my solve times improved substantially. This has happened to me several times now. It seems that UV affects cube solving so much that it should not be left to random chance at a contest. Otherwise people like myself will walk into a contest with a little UV box to keep their cubes in, and not share their secret with others, lol.

Turns to Mix a Cube

Back in the early 80's I practiced about 5 hours per day, from all this practice I had a good idea that a full mix would require upper teens to lower 20's in turns. And anytime someone would ask me I'd tell them my opinion. However, when I asked Ideal Toy Corporation how many turns were used to mix a cube at their contests, they said "13". And sure enough, when I was in the contest in 1982, I solved a cube that seemed like it was only mixed 13 turns (generally, you can tell due to the much easier block building stage, however, very few people used block building back then). So, when I learned this I quit mixing 20-25 turns and started mixing 13 turns. I wanted to practice for the official contest, not for a fully mixed cube. Therefore some of my early solve averages reflected this 13 turn mix.

In my opinion this cannot be considered an unfair advantage due to the simple fact that the official contests also mixed it 13 turns. Records at that time were established from cubes mixed just 13 turns.

Then later, as the official number went up to 18, I started mixing 18. Then later it went all the way up to 25. So I started mixing 25. Then I realized that mixing 25 was not enough, because a human turning through a variety of fingering patterns will not approximate the properties of randomness necessary to stop at the theoretical number of turns necessary. A computer using a pseudo-random number generator could stop there, but a human should go about 50% farther (in my opinion). So at this point I started mixing 35-45 turns each time.

And now, quite recently it has been proven to take only 20 random turns to fully mix a cube. So I now mix my cube at least 30 turns by hand (adding the 50%). And I also attempt to get a different mix each time I do it.

So in my opinion the final say on this should be to mix it 20 times by computer or 30 times by hand.

Questions? Send me an e-mail: tony@snydermind.com