Since I use the Jessica (jiri) method (with the addition of non-matching pairs, and the occassional use of a working corner or edge), much of what is below relates only to the Jessica method. I am still progressing, and am currently a sub-30 cuber hoping one day to be a sub-20 cuber. These tips are only to be viewed as suggestions to try; you may find that some of them work for you, others against you, but at least try them all.
TIP 1: bad cubes a bad cube leads to bad times. If your cube isn't sanded and lubricated then you won't be able to perform to your optimum. Also, the more a cube is worked the better it performs.
TIP 2: active looking during F2L make sure that you actively look for the next pair of cubes that you wish to place. This has made the biggest difference to my times. By actively looking I mean not worrying about what is happening to the pair you are currently placing and not just hoping that the pieces you require will pop into view during the algorithm, but instead looking around the cube for the pieces you need.
TIP 3: F2L on bottom do the Jessica method with the F2L on the bottom. I started with F2L on the top and found it hard to find the pieces I required, swapped to F2L on the left (which is what Jessica herself does) and found that I could do the algorithms a lot faster, though still struggled to find the pieces for the next move, and finally swapped to F2L on bottom whence the pieces became much easier to find and I could still do the algorithms fast. The added bonus is that the cube is already in position for doing the LL (last layer) algorithms.
TIP 4: condition hands for speed practise each corner-edge placing algorithm on its own, trying to do it as fast as possible. The aim is to have the raw speed to be able to do each of them in less than 2 seconds which is needed to achieve sub-20 times.
TIP 5: cube at less than full speed when cubing for real don't do the algorithms at full speed as this doesn't give you enough time to spot the next pair and work out what algorithm you will be doing with that. Ideally you should be doing algorithms without a pause between them, which means that you need to take slightly longer placing a pair than it takes to find another pair.
Proof: If you solve an edge-pair extremely quickly, say, 1 second using triggers, but have moved the faces so quickly that you then have to look for the next pieces to place for, say, 3 seconds, then from when you first turn the faces to finishing the pairs takes 1+(3+1)x3=13 seconds. If you turn the cubes a little slower, say 3 seconds/pair and then going full speed (1 second) on the final pair, you may be able to spot the next pair before you have completed the current pair and would take only 3x3+1=10 seconds. So here, by turning the faces at triple the speed you actually take 3 seconds longer, and that is without even considering that when turning the faces faster you increase the risk of cube lockup which wastes even more time.
TIP 6: know your algorithms 1 practise each algorithm until your fingers know it. As long as your brain is involved in remembering the moves you cannot divert its attention fully to looking for the next pair.
TIP 7: know your algorithms 2 once you are doing an algorithm without thinking about it, learn where to look whilst performing it in order to find the next pair. Also, learn what it does to corners and edge, so that once you have found the next pair you know what position they will be in, which eliminates the pause inbetween pairs.
TIP 8: close your eyes it is too easy to think that you know a move, and that you are actively looking for the next pair, while in reality you have one eye tracking the current pair. If you are placing a pair whilst looking for the next pair then you need to be doing the current pair in the back of your brain without any visual clues. Try locating a pair, then placing with your eyes closed. Keep repeating this until you can place the pairs rapidly without looking and without making mistakes. If you can do the pair without looking then you are able to place the pair whilst using all of your eyesight to look for the next pair.
TIP 9: learn to bail sometimes it is hard to find a pair quickly. If you find you are suffering from cube blindness just start placing any cube in the F2L; the other cube in the pair may spring into view whilst you are doing it, and if not, at least you are one cube nearer completion.
TIP 10: be smooth if you turn the faces at a regular speed then it is easier to look around the cube and still know where you are when you look back. If you speed up and slow down you need to expend mental effort keeping track of what is going on, and if moves are done too fast you cannot see what is happening to the cubes whilst it is happening.
TIP 11: choose algorithms that retain visibility it may be easier to look for the next pair when using one algorithm rather than another; eg R'FRF' vs URU'R' where I find it easier to tilt the cube and look round it whilst using the second than the first. There may be a trade-off which involves doing more moves against retaining improved visibility eg L'U'LU L'U'LU L'U'LU L'U'L is longer than other algorithms that achieve the same effect, but it can be done rapidly whilst also giving great visibility of the cube.
TIP 12: minimise moves the F2L phase should be achievable in around 26 moves on average (excluding the cross) if attempting to do it in as few moves as possible, and a lot of time should be spent trying to minimise the number of moves used whilst doing the cube slowly. Try to figure out different ways of placing the pieces. Revisit moves that you have rejected in the past every once in a while as you may find that a move you didn't like now fits your hands really well. The minimising of the number of moves is a trade-off the other needs of reducing cube lockup and retaining visibility.
TIP 13: reduce cube lockup if you find that whilst doing a move that you often encounter lockup, then find another way of doing it. Cube lockups occur less the more comfortable you get with doing an algorithm at speed, so practise it a fair amount before you give up on the move. Swapping to a cube in better condition may also make the algorithm achievable without lockup.
TIP 14: take 'smoking' breaks at work you know those skivers who nip out for half a dozen 20 minute breaks a day? Nip out for your own breaks and get in your practise without damaging your health.
TIP 15: stay with your current pair if you spot a pair that is easy to place just as you are about to place a pair that is difficult to place, then stay with the pair you were going to place, otherwise you may spend longer thinking about swapping than actually doing it.
TIP 16: stop making mistakes if you continually waste time by making mistakes placing a pair, then you cannot achieve fast times. Practise doing the cube really slowly to make sure that you can identify where your problems are, and then eliminate them.
TIP 17: learn to use a working corner/edge do the cross and then 3 corners or 3 edges and then complete the F2L using the working edge or corner. The need to use a working piece crops up whenever an odd number of pieces has been placed by accident, so it is worth taking the effort to learn to do this fast.
TIP 18: learn to do non-matching pairs okay, so only I do this, and I'm not the fastest in the world, but try it if only to reduce the number of moves that you need when solving for minumum number of turns rather than for speed.
TIP 19: practise lots the only way to get good is to practise lots.
TIP 20: learn all the algorithms for the last layer this is advice that I should take myself. I do a 3-look LL, which can be fairly fast, but often isn't.
TIP 21: avoid needing to flip 4 edges in the last layer when placing the final pair you may be able to orient some/all of the edges - try to avoid needing to flip all 4 edges in the orientation stage as this are the longest/slowest algorithms. This is an idea Dan Knights implements. For me, this wastes more time working it out than it gains. I will revisit it when I am a lot faster.
TIP 22: practise each phase of the solution method on its own spend time concentrating on each phase on its own. For instance, practise the placing of pairs on a cube with the cross done already, and don't do the last layer. This way you can expend all your efforts on the phase you are trying to improve rather than dissipating energy on other phases.
TIP 23: stick with one cube colouring scheme I used to cube with a few cubes with different colouring schemes. It took a few minutes to adjust from using a cube with one colouring scheme to another, but once adjusted speed would be the same on all of them. I eventually stopped swapping cubes and seemed to improve by a few seconds by being familiar with the scheme on the cube being used. I'm not entirely convinced, but think that having just one scheme is beneficial.
TIP 24: always start with the same face starting with the same face each time means that you know the colours of the pairs you will be placing without having to think about it, hence will cube quicker. The downside of always choosing the same face to start with is that it will take an extra move or so to do the cross.
TIP 25: don't rotate the cube rotating the cube (so F becomes L or R) during the solving of the F2L places a burden on the brain and makes it harder to keep track of which slots need filling, plus the pieces are harder to see when the cube is twirling. The way I cube involves a lot of turning of the the bottom 2 layer at once (which is effectively rotating the cube), but I am trying to minimise that. Chris Hardwick does the same and has swapped to not turning the cube at all - it will be interesting to see how well he does.
TIP 26: learn LL algs from different starting orientations try to learn as many algorithms as possible from different starting orientation. Learning the mirrors avoids the need to do U2 to get to the orientation that you know.
TIP 27: recognise LL permutation from looking at 2 sides only the last layer permutation required can be recognised from looking at just 2 sides, which is quicker than looking at all the sides to determine what algorithm is required. Learn to recognise as many as possible from this minimal information.
TIP 28: rotate U not cube for starting LL permutation if the permutation is at an angle different to one that you know, rotate U to get to an angle you know in preference to rotating the whole cube in your hand.
TIP 29: don't start LL permutations with U turns learn the LL permutation without any initial U turns as you will find you will be doing and undoing U turns when you use the moves in practise. This doesn't contradict the previous tip - reread them carefully.
Beginner Intermediate Expert
4 edges in ~10 secs ~4 secs ~2 secs
delay looking for 1st pair ~5 secs ~2 secs 0 secs
4 pairs in ~25 secs ~20 secs ~11
F2L subtotal =40 secs =26 secs =13 secs
see edge orientation ~2 sec ~1 sec -
orient edges ~4 secs ~2 secs -
see corner orientation ~2 secs ~2 sec -
orient corners ~4 secs ~4 secs -
see orientation pattern - - ~1 secs
do orientation pattern - - ~2 secs
see permutation ~6 secs ~3 secs ~1 sec
permute all ~10 secs ~6 secs ~3
Last layer subtotal =28secs =18 secs =7 secs
TOTAL =68secs =44 secs =20secs
TIP 30: concentrate on your poor areas compare your performance in each area with those in the chart above, and work on any area that you feel that you are doing poorly. Do the phase in isolation to the others so that you can concentrate properly upon it.
TIP 31: perform in public to reduce hand shake competing against the clock causes the hands to shake from nerves; cubing infront of people is extremely nerve wracking so competing against the clock won't seem so daunting after that.
TIP 32: plan the solution of the cross before starting the first 4 edges can take a long time to solve when choosing moves on-the-fly as you solve, particularly if track is lost of one of the pieces. The best way is to plan out the solution beforehand looking at taking <8 moves. If you are a beginner then try to see in advance how to put at least 3 of the 4 edges in place, and work your way up to doing 4. Practise solving just this stage and do it very slowly.
TIP 33: eliminate delay between cross and first corner-edge pair the delay between completing the cross and starting on the the first corner-edge pair can cause a large loss of time. It takes a while to acquaint yourself with where the pieces are if the cross is done rapidly - if you do the cross slightly more slowly you will have time to find an appropriate pair by the time that you have completed it, which is what I do. The other solution to this problem is to plan the moves for creating the cross and first pair before starting. This is possible, but I find it too much hard work to implement.
TIP 34: learn macros/triggers for common situations in building the cross placing the pieces with macros/triggers will allow the cross to be placed extremely rapidly, preferably under 2 seconds. Learn ways to place the final piece quickly, and to do pairs of pieces.
TIP 35: twisting faces faster may not be quicker during the F2L phase it is best to remember that twisting the faces faster does not necessarily lead to quicker times; try to go slow enough that you can spot the pieces to be placed next.
TIP 35: choose moves that are easy for your hands if a move is awkward for your hands you won't be able to do it quickly. Practise will make those moves easier, but try to find algorithms that fit better even if they are slightly longer.
TIP 36: don't go stale revisit ideas and try new ones from time to time which may help you cube faster and your cubing will remain interesting. Remember that ideas that were not for you previously may suit you later.
TIP 37: carry a cube and moves sheet with you carry a sheet of the moves that you are learning, so that whenever you get a slack moment you can practise. It is better to learn in small chunks spread throughtout the day, then periodically in large chunks.
TIP 38: cube in good light this I think is important (and so does Jessica Fridrich), though it may seem a little strange. If you struggle to identify the colours of the pieces because of poor lighting conditions then you can't expect to place pieces rapidly.
TIP 39: make yourself comfortable hold the cube at a comfortable distance from your body.
TIP 40: take occassional breaks from cubing the occassional few days taken avoiding cubing can recharge your batteries, and also gives time for your subconscious to improve you. I have often been better after not cubing for a while than when cubing incessantly.
TIP 41: use a complimentary pseudonym this one from Ron, who reckons that using a complimentary handle on the yahoo club will be beneficial to your time - slowcuber changes to smoothcuber and broke 30 seconds, I mentally imagined I was slick_hands and broke 30 seconds, so it may be bizarre but has succeeded twice.
TIP 42: cube on the can another Ron contribution - do an algorithm 20 times every time you go to the toilet. I guess this can only be a sit-down session unless you are able to go one handed.
TIP 43: examine videos of fast cubers watching videos of fast cubers can give you good ideas of style, moves to use, and show you what is achievable.
TIP 44: keep cube stable cube sitting, with your elbows on your knees
TIP 45: don't muscle it do not use much pressure, cube light-fingered
TIP 46: cube with the tips of your fingers use your fingers as if you were playing a piano or guitar, not like you were having your prints taken at the police station. My slower algorithms all have my fingers flat against the cube.
TIP 47: think of your moves as a system not a sequence of moves to memorise to get any real speed the moves to apply in a situation have to make sense to a degree. One way of thinking is presented here... Suppose your corner is in the wrong slot. Start your move as if the corner is in the correct slot, and then once both the corner and edge are in the U layer (supposing F2L is the bottom two layers) then rotate U and finish in the correct slot.
TIP 48: retain visibilty when matching the centres after completing the cross once the cross is complete the centres can be matched in one of two ways, either rotating the bottom layer or rotating the top two layers. It is much better to rotate the bottom layer only as that way most of the cubies stay static allowing you to have a good look at them.
TIP 49: make the fourth edge of the cross easy before placing the third edge of the cross check whether the fourth edge would then be easy to place. If not, then often a simple turn of one face will put that fourth edge in position such that when the third edge is placed it moves the fourth edge to a place where it can be placed easily.
TIP 50: go slowly! bizarrely this is the most important advice that you will ever receive, but you won't believe it until the day it finally clicks. By going slowly it is easier to spot pieces and that leads to less and smaller delays and hence faster cubing. When you finally reach this state you will know what I mean.
TIP 51: try unobvious ways of placing edges when doing the cross suppose that doing the move R would place an edge, then consider doing R' and see if another edge is then easy to place with you being able to then place the original edge with R2.
TIP 52: consider placing edges in the wrong slot when doing the cross placing say the blue edge where the green edge should be may mean that the blue edge is in a convenient place when the green edge piece has replaced it in the slot, thus making the cross solution flow whereas without doing that you may find the particular setup you have is a number of disjointed moves.
TIP 53: use Gorilla Snot this is what drummers use to stop their sticks slipping out of their hands. If you find your cube is too slippery then this must be a good idea