Moves per second is way more important than total moves

[PenguinsDontFly]

But I rotate and use PLL
Guess I'll never be fast
(Also I don't use Roux; I can't possibly be fast)

lol my propaganda

Part of the reason I think roux and petrus are better than CFOP is pure efficiency. Eventually, when cubers are reaching human TPS limits and have perfect lookahead, roux and petrus will win.

 

[Ranzha]

lol my propaganda

Part of the reason I think roux and petrus are better than CFOP is pure efficiency. Eventually, when cubers are reaching human TPS limits and have perfect lookahead, roux and petrus will win.

Petrus probably won't win because of the lack of popularity. I see ZZ making huge strides though in recent years.

 

[cubo]

I'm not sure where you are going with this...?

Wherever it takes us

There are three variables, not two. Move count, TPS, and pauses.

Completely agree, and thank you again for bringing pauses into the discussion.

It doesn't make sense to conflate pauses and TPS to give a TPS measure for the whole solve, because that measure doesn't help us understand how to improve.

Again, agree. Given that the far majority of pausing happens between algs, not during algs, I think we can and should make a clean distinction between move count (total moves taken), pausing (total time taken between algs) and TPS (total time minus pause time divided by total moves taken).

All three variables are important and you need to improve all three to get fast. It sounds like you're saying TPS is more important than the others, but again it's not helpful to take it in isolation like that.

Yes ... and no. I'm saying that mathematically, the most important factor and biggest contributor to speed improvement between my two example cases (100 second solve and 10 second solve) is the difference in TPS. And honestly, that's just maths, as far as I can tell. (Please show me I'm wrong I will be happy to concede the point if shown otherwise.)

Just to clarify a couple of things, here's what I'm NOT saying:

• I'm not saying TPS is the ONLY thing you must improve to get better times. Lowering move count and pauses will naturally also help.
  
• I'm not saying TPS improvement happens in isolation. Everything happens together. Learning new algs could lower move count and also help improve TPS if they are a smoother (faster) alg to perform, for you.

Typically we improve through phases. When very slow, we make big early gains by reducing pauses.

Agree, although perhaps still disagree on the amount of pausing at that early stage.

Then we learn better F2L/B and more alg sets and become more move efficient. TPS improves gradually all the time. Once you know a good amount of algs and get to around 20s, improvement becomes harder because TPS takes longer to improve and reducing pauses means looking ahead rather than just shortening recognition further. The variables start to conflict with each other because pushing TPS too hard makes look ahead more difficult.

So it's complicated and saying TPS is the most important thing over-simplifies the process.

Great description of the learning process and how it changes. Just to make the point again, I'm NOT saying "OK you're a 20 second solver and you want to get to 10 secs? Move those hands faster!" I am just observing that, overall, between 100 seconds and 10 seconds, hand speed makes the biggest impact.

Also, 70% is just a guess

I would love to get some more reliable data on this.

Improvement above 40-50s really is more about becoming competent than becoming fast.

What do you mean by "competent"?

There's no point comparing 100 second solvers to world <10 second solvers.

Well actually, that's my whole point And why not? 100 seconds is a nice round number, and is a point a cuber would reach fairly quickly after stumbling through the first, absolutely basic part of learning. Likewise 10 seconds is another nice round number representing a very advanced cuber, but not ridiculously advanced. Basically I just cut off the extremes of complete novice and world-class and got my 100 vs 10 second comparison.

And another really important factor is, as Mark said, that faster turning makes for harder lookahead, which is EVERYTHING. Look ahead is, without a shadow of a doubt, the most important thing in any current speedsolving method.

Why? Lookahead reduces pausing, but gains from reducing pausing, as I've previously demonstrated, is still not as important as gains from reducing TPS between the 100 sec and 10 sec solve.

There is a categorical, and very important difference between turning speed and 'tps', a shorthand we use for moves/time. There are plenty of solvers about 2 seconds slower than I am that are capable of faster burst, and some that pause less than me. Min-maxing the trade-off as accurately as possible is the idea ...

Just to clarify, are these "bursts" happening between different algs, or during a single alg? I assume the former, in which case you are basically describing the pauses ... and the balancing act between wanting to go faster (TPS) but also wanting to minimise pausing. But if upping your TPS gets you to the next alg faster, surely that gained time can be used to inspect once you get there - if needed, and perhaps it won't be - and you may still be ahead in the end?

Now in terms of actually getting better, priority thinking like the title isn't that useful either.

I was more making an observation rather than prescribing a training approach, which as a beginner, I don't feel qualified to do

Edit: Btw, only just noticed you said you were a beginner - it's really cool that you're thinking about stuff like this as a novice. Can't promise I'll respond that quickly but you're welcome to PM me if you ever want some advice on improving.

Thank you so much! It's great to find an amazing cubing community such as this one!!

It's pretty obvious TPS is mainly why we are seeing todays fast solves. The solve movecount is very similar to what it was 20 years ago, before finger-tricks and decent cubes, but the solve times are down to one third of what they were then.

Gracias senor, you have summed up my point beautifully. That, in essence, is all I am saying.

 

[GuRoux]

lol my propaganda

Part of the reason I think roux and petrus are better than CFOP is pure efficiency. Eventually, when cubers are reaching human TPS limits and have perfect lookahead, roux and petrus will win.

and there's huge popularity where pros have cubing as their full time job for several years before they even come close to breaking the world records.

some of the previous post talk about what percent of the solve are pauses. i think in general, people are about 50% pauses. it's not until sub 10 that you see 10-30% pauses.

 

[cubo]

Some new estimates, based on:

• 100 second solve, 100 moves, 50% pausing. TPS is therefore 2.00
  
• 10 second solve, 50 moves, 10% pausing. TPS is therefore 5.55

That means the improvement factors are now:

• TPS (x2.78)
• Move efficiency (x2)
• Pausing (x1.78)

So even with pausing at 50% initially, TPS is still clearly the biggest factor, about 40% more significant than move reduction.

As a further note, if we consider how you could possibly improve 10 to 5 seconds in this scenario, the gains would have to be almost all about improving TPS (about 70%) with the rest a mixture of move reduction/lucky solves? (25%) and pause reduction (5%).

This indicates to me that TPS is always the most important factor, and only becomes more so as you get lower times.

 

[mark49152]

So even with pausing at 50% initially, TPS is still clearly the biggest factor

How? You got 90 seconds faster, and 48 seconds of that was by eliminating pauses.

 

[AlphaSheep]

But if upping your TPS gets you to the next alg faster, surely that gained time can be used to inspect once you get there - if needed, and perhaps it won't be - and you may still be ahead in the end?

This is not true because you can't just go from paused to turning at full speed. There's a bit of a startup time, so while you may pause for half a second, it could actually cost you a full second on your total time. It would be better to turn at 75% speed to avoid the pause.

I think you're seeing a lot of disagreement in this thread because what you're saying goes against the experience of most fast cubers.

From what I've seen, the typical experience is this:

1. Learn to solve the cube, get to 60 seconds or so with a beginner method and a little practice.
2. Learn F2L and maybe 4 look last layer.
3. Start trying to solve faster. Times drop quickly with practice and improving fingertricks and TPS alone. This gets most people to 25-35 seconds, but then they plateau and stop getting faster.
4. Learn better methods - 3 look or even 2 look last layer. This can get people to 20 seconds. Many people get stuck here trying to break the 20 second barrier.
5. Learn to transition between steps, and recognise cases without having to pause. To do this, you actually have to slow down your turn speed. This is where I am at the moment, and it seems to get people to around 12-15 seconds.
6. Learn to predict further and further in inspection and see two or more F2L cases ahead so you absolutely minimise pause in the solve. This seems to be the key difference I've noticed between 11-12 second solvers and 8-9 second solvers.

I've met so many people with TPS far higher than mine stuck at step 3 because the think that TPS is everything.

I don't disagree that TPS is important - its a large part of what separates a 100s solver from a 30s solver, but the key difference between a 20s solver and a 10s solver is lookahead, minimal pauses, and efficiency, and TPS only plays a small role.

It's also worth noting that the reason for this is that TPS is far easier to improve on than anything else, especially for a beginner. It's the only thing that gets better with mindless practice. Lookahead and efficiency need you to actively concentrate to improve, but just mindlessly scrambling a cube helps with TPS. On top of that, having a higher TPS actually makes it harder to improve at the other things. The faster you turn, the harder it becomes to track pieces and see the next case. Honestly, I've cut my times by 20% just by turning slower.

And honestly, that's just maths, as far as I can tell. (Please show me I'm wrong I will be happy to concede the point if shown otherwise.)

I'll try take you up on that offer.

Let me try make an analogy. A bicycle can cover 10 km in 20 min (30 km/h), and a car can cover 5 km in 2.5 min (120 km/h). The halving the distance for the car contributes a 2x factor, whereas the speed contributes a 4x factor. At first glance this seems fine - you would conclude by your argument that the car being faster is more important than the distance in determining travel time.

If you think about it, that's obviously not a valid conclusion. I could increase the distance for the car, to say 200 km, but leave the distance for the bicycle the same. Now the bicycle will reach it's destination long before the car, so it appears that the distance is more important in this case. The correct answer is that travel time is a function of both distance and speed and neither is more important than the other.

To bring it back to cubing and your argument, you've noted that fast solvers typically have a faster TPS (bundling pauses in with the rest of solve time for simplicity) with a 5x factor but the move count only differs by a factor of 2x. You wrongly conclude that TPS is therefore more important, when as with the analogy above, neither is more important than the other.

The actual reason that the factor for TPS is higher is just because TPS is a lot easier to improve with very little effort. In fact, since TPS gets better without much effort, it could be said that its more important to concentrate on move efficiency.

Obviously there's a tradeoff. Which is why, as has already been mentioned a few times in this thread, most people use algs that are a couple moves longer than optimal, because they flow more smoothly and can be executed faster. This is something that pretty much every fast cuber is well aware of.

 

[shadowslice e]

You know, I think most of the arguments you're getting here is that people are disagreeing with your use of the phrase "more important"; improves most=/=more important.

 

[cubo]

How? You got 90 seconds faster, and 48 seconds of that was by eliminating pauses.

Although I can see how you got that number, I don't think it's mathematically correct because it tries to calculate an absolute figure but we are dealing in 3 intertwined factors. By that logic, move improvement has saved us 25 seconds (by halving the 50 secs of active solve time) and TPS has saved us 32 seconds (by going from 2 moves per second to 5.5 moves per second) for a total of 105 seconds saved!!

My calculation is based on the relative contribution of each factor to the overall x10 time improvement (which is just another way of saying our 10 second time is 10 times faster than out 100 second time). Our 3 factors, when multiplied together, must equal 10. That's my calculation above (it's actually a bit under 10 because pausing should be 1.8 not 1.78, I'll show why below - my first calc was made in haste).

Another way of considering the mathematical importance of each factor is to look at our 10 second solve and then imagine what would happen if each factor regressed to the beginner level and see which one causes us to lose the most time, as that will be the most important factor in this inverse scenario. So ...

If our 10 second solver lost all
their move efficiency and went back to 100 moves (while maintaining the better TPS and pause time) their total solve time would now be 20 seconds seconds (18 second solve plus 2 seconds pause time).

If our 10 second solver went back to 50% pause time, that would be a new total of 18 seconds (9 second solve with 9 seconds of pausing). This is slightly lower than the new time if we lost our move efficiency, meaning pausing is slightly less of a factor than move efficiency.

And lastly, if we went back to our old TPS of 2 moves per second (while keeping our improved move efficiency and pause time), our time blows out all the way to 27.8 seconds seconds (25 second solve plus 2.8 seconds pause).

Now you should be able to see the factors more clearly ...

Move efficiency blows us out from 10 secs to 20 secs, so it's a x2 factor.

Pause improvement blows us out from 10 secs to 18 secs, so it's a X1.8 factor.

And lastly, TPS, which if lost blows our 10 secs time all the way out to 27.8 secs, is a x2.78 factor.

Mark does that explanation make what I'm saying any clearer?

PS Alpha and Shadow I will respond in the morning, gotta hit the sack now.

Peace out all

 

[mark49152]

Mark does that explanation make what I'm saying any clearer?

Yes your reasoning is clear, but I don't agree with it. You've now created three hypothetical solvers who have regressed on each of the three variables alone. Except that your pausing solver hasn't fully regressed because he still pauses 41 seconds less than the 100 second solver.

I think you should go back and read the analogies about speed and distance above.

I'm curious about your experience. How long have you been cubing and how has your own progression been so far?

 

[Escher]

Just to clarify, are these "bursts" happening between different algs, or during a single alg? I assume the former, in which case you are basically describing the pauses ... and the balancing act between wanting to go faster (TPS) but also wanting to minimise pausing. But if upping your TPS gets you to the next alg faster, surely that gained time can be used to inspect once you get there - if needed, and perhaps it won't be - and you may still be ahead in the end?

Well, no, I actually mean max turnspeed, as in, someone who can execute a fixed sequence of moves faster than I can. Choosing to up your tps and inspect later makes me feel dirty, not from phobia but from experience. You are not just blindly anticipating and executing fixed patterns at the highest level (tbh neither are a lot of intermediates). You are making predictions, inferences and assumptions, and it's much easier to solve those individual questions across 2-3 seconds of turning - slow enough to strengthen or solidify existing theories and begin generating new ones, rather than blast out one seconds worth of high tps without updating any of your existing lookahead. You need to act just slow enough to collect information as it is being revealed to you.

Also I'm hot on paint these days and I made you another couple of pics.

Hope this helps,

Rowan

 

[JanW]

I agree with Mark that you improve the most by eliminating pauses and improving look ahead, not by improving TPS.

Your TPS theory makes the false assumption that the TPS of a beginner is dependent on how fast this solver is able to perform algs. Also, you seem to assume that TPS is uniform over the entire solve. It's not. A beginner using the method that ends with spamming sexy moves uses 27 or 52 moves only to orient the LL corners. If it's 52 moves, it will not take half the time of a 100 move 1-minute solve. Far from it. The 1-minute solver would most likely do this step at way above 5 tps.

Most of the solve time will be spent on F2L. TPS during F2L is not limited by how fast you can turn the faces of the cube. Your TPS is mainly determined by how fast you can find your pieces and work out how to insert them. As many have mentioned, aiming at high TPS at this stage is actually detrimental to your solving time, as it results in worse look ahead. If you use layer by layer, inserting the corners, then the edges should be 40 moves max. Add a badly planned 10 move cross to that and it's 50 moves for cross+F2L. At a very modest 2 moves/second, it should be done in 25 seconds, which is much faster than a 1-minute solver would manage. And the reason they can't do it in 25 seconds is not that they can't perform the algs at 2 moves/second. As long as you have a half-decent cube, anyone can do a R U R' insertion in <1 second without any training at all.

Like Alpha Sheep said, improved TPS comes automatically the more you solve. Focus on look ahead and learning more efficient algs will bring you much faster from 1 minute to 20 seconds.

 

[cubo]

You know, I think most of the arguments you're getting here is that people are disagreeing with your use of the phrase "more important"; improves most=/=more important.

Yes I know what you mean ... but the alternatives "most significant", "biggest impact" etc don't seem better.

Again: my observation is NOT about saying "this means you must focus on TPS above all else". That would be prescriptive, which as a beginner I don't feel qualified to be. It may actually be that you don't need to focus on TPS at all as you progress. I simply don't feel qualified to comment on that. Telling people how to get better is NOT my point here.

Rather, I'm simply observing and describing the mathematics involved in moving between two realistic scenarios and pointing out that, overall, TPS contributes most (i.e. is most significant, important, whatevs ) in the overall time improvement between those two scenarios.

Hi Alpha! My comments in bold in this quote ...

From what I've seen, the typical experience is this:

1. Learn to solve the cube, get to 60 seconds or so with a beginner method and a little practice. So TPS + lower pausing get you here?
   
2. Learn F2L and maybe 4 look last layer. OK that's move efficiency.
   
3. Start trying to solve faster. Times drop quickly with practice and improving fingertricks and TPS alone. This gets most people to 25-35 seconds, but then they plateau and stop getting faster. OK so you're saying efficiency + TPS take you from 60-30.
   
4. Learn better methods - 3 look or even 2 look last layer. This can get people to 20 seconds. Many people get stuck here trying to break the 20 second barrier. So then move efficiency takes you from 30-20.
   
5. Learn to transition between steps, and recognise cases without having to pause. To do this, you actually have to slow down your turn speed. This is where I am at the moment, and it seems to get people to around 12-15 seconds. So pause reduction takes you from 20-12.
   
6. Learn to predict further and further in inspection and see two or more F2L cases ahead so you absolutely minimise pause in the solve. This seems to be the key difference I've noticed between 11-12 second solvers and 8-9 second solvers. More pause reduction.

I don't disagree that TPS is important - its a large part of what separates a 100s solver from a 30s solver, but the key difference between a 20s solver and a 10s solver is lookahead, minimal pauses, and efficiency, and TPS only plays a small role. OK sure but remember I didn't look at 20 to 10 secs specifically, I looked at 100 to 10 secs.

Well I don't really disagree with any of that ... or more precisely, I'll take your word for it because I actually don't know. All I've been saying is that if we look at 100 to 10 secs, TPS is a large part of that overall (definitely the largest part ) but that efficiency and pausing are also significant.

It's also worth noting that the reason for this is that TPS is far easier to improve on than anything else, especially for a beginner. It's the only thing that gets better with mindless practice. Lookahead and efficiency need you to actively concentrate to improve, but just mindlessly scrambling a cube helps with TPS.

If anything I think that supports my point. Maybe TPS is the biggest factor in improvement precisely because it is largely mindless -- although I would challenge whether finger tricks, which can really help TPS, are mindless? They seem to take strong effort at first.

On top of that, having a higher TPS actually makes it harder to improve at the other things. The faster you turn, the harder it becomes to track pieces and see the next case. Honestly, I've cut my times by 20% just by turning slower.

That may indeed be the case at some point in your 100 sec to 10 sec journey. But my observation is regarding the overall improvement over the entire period.

I'll try take you up on that offer ... Let me try make an analogy. A bicycle can cover 10 km in 20 min (30 km/h), and a car can cover 5 km in 2.5 min (120 km/h). The halving the distance for the car contributes a 2x factor, whereas the speed contributes a 4x factor. At first glance this seems fine - you would conclude by your argument that the car being faster is more important than the distance in determining travel time. If you think about it, that's obviously not a valid conclusion. I could increase the distance for the car, to say 200 km, but leave the distance for the bicycle the same. Now the bicycle will reach it's destination long before the car, so it appears that the distance is more important in this case.

Um yes, because it actually is. If you are trying to answer the question "Why does the car take longer than the bike to reach its destination in this case?" then the answer is VERY much, "Because the car had to travel 200 km and the bike only traveled 10km." Am I missing something?

But actually I love this analogy! It can be used to illustrate exactly what I'm saying. Presenting ...

RUBIK'S ROAD RACE
Rules: a 100km race, use any vehicle you want, if you can find a shortcut you are welcome to take it, no maps allowed.

Novice racer: uses a car with a top speed of 120 km/h, takes no shortcuts and hence travels the full 100km (in 50 mins), but spends half the time stopped by the road working out where to go (also 50 mins). Final time: 100 minutes.

Expert racer: uses a car that is 2.8 times faster (336 km/h), takes a shortcut which means he only actually travels 50km, and spends only 10% of the time stopped by the road working out where to go. Final time: 10 minutes.

Sound familiar? I hope so. 'Speed' here is TPS, 'distance' is move efficiency, and 'stopped by the road' is pause time. And it perfectly matches my example, using the same contributing factors. And as per my example, the most important is ... speed, aka TPS.

Obviously there's a tradeoff. Which is why, as has already been mentioned a few times in this thread, most people use algs that are a couple moves longer than optimal, because they flow more smoothly and can be executed faster. This is something that pretty much every fast cuber is well aware of.

100% agree in fact I mentioned it myself previously in this thread

 

[GenTheSnail]

Hi Alpha! My comments in bold in this quote ...

From what I've seen, the typical experience is this:

1. Learn to solve the cube, get to 60 seconds or so with a beginner method and a little practice. So TPS + lower pausing get you here?
   
2. Learn F2L and maybe 4 look last layer. OK that's move efficiency.
   
3. Start trying to solve faster. Times drop quickly with practice and improving fingertricks and TPS alone. This gets most people to 25-35 seconds, but then they plateau and stop getting faster. OK so you're saying efficiency + TPS take you from 60-30.
   
4. Learn better methods - 3 look or even 2 look last layer. This can get people to 20 seconds. Many people get stuck here trying to break the 20 second barrier. So then move efficiency takes you from 30-20.
   
5. Learn to transition between steps, and recognise cases without having to pause. To do this, you actually have to slow down your turn speed. This is where I am at the moment, and it seems to get people to around 12-15 seconds. So pause reduction takes you from 20-12.
   
6. Learn to predict further and further in inspection and see two or more F2L cases ahead so you absolutely minimise pause in the solve. This seems to be the key difference I've noticed between 11-12 second solvers and 8-9 second solvers. More pause reduction.

I don't disagree that TPS is important - its a large part of what separates a 100s solver from a 30s solver, but the key difference between a 20s solver and a 10s solver is lookahead, minimal pauses, and efficiency, and TPS only plays a small role. OK sure but remember I didn't look at 20 to 10 secs specifically, I looked at 100 to 10 secs

Well I don't really disagree with any of that ... or more precisely, I'll take your word for it because I actually don't know. All I've been saying is that if we look at 100 to 10 secs, TPS is a large part of that overall (definitely the largest part ) but that efficiency and pausing are also significant.

If anything I think that supports my point. Maybe TPS is the biggest factor in improvement precisely because it is largely mindless -- although I would challenge whether finger tricks, which can really help TPS, are mindless? They seem to take strong effort at first.

That may indeed be the case at some point in your 100 sec to 10 sec journey. But my observation is regarding the overall improvement over the entire period.

Um yes, because it actually is. If you are trying to answer the question "Why does the car take longer than the car to reach its destination in this case?" then the answer is VERY much, "Because the car had to travel 200 km and the bike only traveled 10km." Am I missing something?

But actually I love this analogy! It can be used to illustrate exactly what I'm saying. Presenting ...

RUBIK'S ROAD RACE
Rules: a 100km race, use any vehicle you want, if you can find a shortcut you are welcome to take it, no maps allowed.

Novice racer: uses a car with a top speed of 120 km/h, takes no shortcuts and hence travels the full 100km (in 50 mins), but spends half the time stopped by the road working out where to go (also 50 mins). Final time: 100 minutes.

Expert racer: uses a car that is 2.8 times faster (336 km/h), takes a shortcut which means he only actually travels 50km, and spends only 10% of the time stopped by the road working out where to go. Final time: 10 minutes.

Sound familiar? I hope so. 'Speed' here is TPS, 'distance' is move efficiency, and 'stopped by the road' is pause time. And it perfectly matches my example, using the same contributing factors. And as per my example, the most important is ... speed, aka TPS.

100% agree in fact I mentioned it myself previously in this thread

So, one thing I don't think you get is that when TPS increases, your lookahead goes down and your pauses go up.

A quote from AlphaSheep from June talking about data tracking:

A few months ago I decided to focus purely on TPS. I got some awesome singles, and some really good Ao5s and it certainly felt like I was getting faster. But then when I checked my progress graph, my big averages were steadily getting worse. It made me look at my solves more critically, and I realised my lookahead had been suffering. Once I started working on that, the big averages started falling again.

I don't think you can really compare 100sec -> 10sec.

As alphasheep said,

From what I've seen, the typical experience is this:

1. Learn to solve the cube, get to 60 seconds or so with a beginner method and a little practice. TPS
2. Learn F2L and maybe 4 look last layer. Efficiency, a little TPS
3. Start trying to solve faster. Times drop quickly with practice and improving fingertricks and TPS alone. This gets most people to 25-35 seconds, but then they plateau and stop getting faster. Thorough F2L understanding TPS. This is when TPS stops helping as much
4. Learn better methods - 3 look or even 2 look last layer. This can get people to 20 seconds. Many people get stuck here trying to break the 20 second barrier. Efficiency, some basic F2L tricks, and yes, some TPS improvement
5. Learn to transition between steps, and recognise cases without having to pause. To do this, you actually have to slow down your turn speed. This is where I am at the moment, and it seems to get people to around 12-15 seconds. Pause reduction, control you TPS, dont spam it.
6. Learn to predict further and further in inspection and see two or more F2L cases ahead so you absolutely minimise pause in the solve. This seems to be the key difference I've noticed between 11-12 second solvers and 8-9 second solvers. Pause reduction, efficiency

I've met so many people with TPS far higher than mine stuck at step 3 because the think that TPS is everything.

I average high 16 right now, but a couple months ago, when I was still sup-17, I decided to do a slow solve. It was completely random, I just felt like doing a slow solve. I did it and I got a 15 with barely ~3 TPS.
The thing that dropped 2 seconds, was a decrease in TPS and pauses, and an increase in efficiency.

 

[cubo]

Yes your reasoning is clear, but I don't agree with it.

That feels like you're saying "It's clear that 1+1=2. But I don't agree with it." ??? When it comes to the maths, which bit don't you agree with? Please please point out my error if you see one. I've pointed out yours in detail above.

You've now created three hypothetical solvers who have regressed on each of the three variables alone. Except that your pausing solver hasn't fully regressed because he still pauses 41 seconds less than the 100 second solver.

Exactly, I created 3 hypotheticals to illustrate the way the 3 factors interletate to each other to create, in combination, a x10 time savings, because it was clear from your 48 second-pausing-gain claim (now changed to 41 seconds) that you weren't grasping it. But now we're going backwards ... you've just repeated your last claim. Let me put it this way: the 41 seconds you talk about is as much about TPS and efficiency as pausing, because the improved pausing rate of 10% is relative to the total time, which has also been lowered (in fact, more so) by TPS and efficiency. Pausing has absolutely NOT NOT NOT saved all that time itself. And when you look at the 3 factors in their interaction, as I have done multiple times now, it is clear which one is most significant: TPS!

If you really think pausing saved 41 seconds, let me ask you this: how much do you think TPS and efficiency saved on the way from 100 secs to 10 secs? Please show me, using the parameters set out (100 secs, 100 moves, 50% pausing ... to 10 secs, 50 moves, 10% pausing). I think you'll find your way just doesn't add up.

I think you should go back and read the analogies about speed and distance above.

My post above this one addresses those and the simple mistake alpha has made.

I'm curious about your experience. How long have you been cubing and how has your own progression been so far?

6 months, not regularly because I've got kids and a job, current pb 45 secs, av 60 secs. I consider myself a noob. My progress to date has been a mixture of improving TPS (practice, practice, finger tricks and choosing my algs wisely), efficiency (currently transitioning from keyhole to F2L) and reducing pausing - so all 3 I'm a long way from 10 seconds.

 

[genericcuber666]

Nobody really pauses who bring that up even when we were able averaging 40-70 we only paused for max 10 seconds after that we all learnt to look ahead making pauses irrelevant except at all if that's what you mean but I don't see sub 10 people pausing

 

[sotolf2]

That feels like you're saying "It's clear that 1+1=2. But I don't agree with it." ??? When it comes to the maths, which bit don't you agree with? Please please point out my error if you see one. I've pointed out yours in detail above.

I think what he, me and others don't agree with is that TPS will get you anywhere without good lookahead, if you don't have good lookahead you have to do random moves to keep your TPS up, which will make your lookahead even worse. I find it to be nonsense to say that TPS is more important than lookahead because it's not something you can meanfully obtain without good lookahead.

I have seen people do slow solves that are a lot faster than my TPS spamming tries, but I haven't seen someone with bad lookahead spamming moves at random managing to beat me.

And how can you train lookahead by spamming TPS? it can't be done, it has to be done slower, so go slower to train, and first go faster when you actually know where the pieces are, it's not like you start driving training on the highway in 120 km/h you start slower until you know what you're doing, and first then when you know what to look for and are more secure in what you're doing you're finally doing speed, and by that time the speed is better since you've slowly built it up together with your lookahead ability. To be honest I don't get what's hard to understand about that

 

[mark49152]

Exactly, I created 3 hypotheticals to illustrate the way the 3 factors interletate to each other to create, in combination, a x10 time savings, because it was clear from your 48 second-pausing-gain claim (now changed to 41 seconds) that you weren't grasping it. But now we're going backwards ... you've just repeated your last claim. Let me put it this way: the 41 seconds you talk about is as much about TPS and efficiency as pausing, because the improved pausing rate of 10% is relative to the total time, which has also been lowered (in fact, more so) by TPS and efficiency. Pausing has absolutely NOT NOT NOT saved all that time itself. And when you look at the 3 factors in their interaction, as I have done multiple times now, it is clear which one is most significant: TPS!

Your 100 second example solver paused for 50 seconds. Your 10 second solver paused for 1 second. You then created another hypothetical solver that had the TPS and efficiency of a 10 second solver and claimed he made "no improvement" in pauses relative to the 100 second solver because he paused for 9 seconds which is 50% of his solve. This is what I disagree with. The measure of pauses is seconds, not a percentage. Your assumption that the time you spend searching or thinking is somehow a function of your TPS and move efficiency is bogus. If he had made genuinely no improvement in pauses, he would have turned for 9 seconds, paused for 50 seconds, total 59, and your case falls apart.

The reason I asked earlier where you were going with this is because I don't understand the purpose or value of your argument. You seem determined to prove and hammer home this TPS-is-most-important point, which seems a useless conclusion derived from over-simplification and dubious assumptions. It's an interesting discussion, which I enjoyed at first, and there has been some good info and advice posted by others, but why is this TPS point so important? As has been said many times, it's really not as simple as that.

 

[AlphaSheep]

Um yes, because it actually is. If you are trying to answer the question "Why does the car take longer than the bike to reach its destination in this case?" then the answer is VERY much, "Because the car had to travel 200 km and the bike only traveled 10km." Am I missing something?

But actually I love this analogy! It can be used to illustrate exactly what I'm saying. Presenting ...

RUBIK'S ROAD RACE
Rules: a 100km race, use any vehicle you want, if you can find a shortcut you are welcome to take it, no maps allowed.

Novice racer: uses a car with a top speed of 120 km/h, takes no shortcuts and hence travels the full 100km (in 50 mins), but spends half the time stopped by the road working out where to go (also 50 mins). Final time: 100 minutes.

Expert racer: uses a car that is 2.8 times faster (336 km/h), takes a shortcut which means he only actually travels 50km, and spends only 10% of the time stopped by the road working out where to go. Final time: 10 minutes.

Sound familiar? I hope so. 'Speed' here is TPS, 'distance' is move efficiency, and 'stopped by the road' is pause time. And it perfectly matches my example, using the same contributing factors. And as per my example, the most important is ... speed, aka TPS.

I think you've missed the point. All you've shown is that speed contributes the most in this specific case, not that it's the most important, but if you change the circumstances you get a different results.

I know you're considering a 100 second solver but what happens if you consider an 80 or 60 second solver?

Also I'm still not convinced the numbers for percentage time paused and total move count. Would you (or any other person in 90 seconds or slower) be willing to record a video of 5 solves? I'd like to try count the moves and pauses.

 

[JanW]

Also I'm still not convinced the numbers for percentage time paused and total move count. Would you (or any other person in 90 seconds or slower) be willing to record a video of 5 solves? I'd like to try count the moves and pauses.

When Marcel started the thread currently known as Older cuber discussions back in 2012, he posted a 90 seconds solve which seems to be about 100 moves. The video is still up with a link in the OP of that thread. I just watched it and tried to count the moves by clicking the calculator to increase by +1 whenever he turned. It was actually hard to keep up at times, because his TPS while doing algs was pretty fast. The majority of the time is spent rotating the cube and looking for pieces.

If you do an analysis to count moves and pauses, then I think unnecessary moves, such as U-U-U2 while looking for the next F2L pair, should count as pausing as well.