Before learning [X] you should at least be sub [Y]

adimare

Veteran
Joined
Oct 29, 2009
Messages
374
Location
Costa Rica
WCA ID
2011MARE02
Many people seem to think that you should reach certain speeds before learning certain algs/techniques/methods (I learned 3-style BLD from Noah's videos where he advises not to look into the method unless you're at least sub 1:30). I always ignore such suggestions because to me it makes more sense to get fast using the method you want to use (3-style) rather than getting fast using something else (M2/OP), then fundamentally changing your approach to doing what you're doing and having to get fast again with the new approach.

Does anyone think that learning something new might hurt your learning curve in the long term?
 
Not exactly speeds, but some methods or algs or something require you to have prior knowledge in how something works.

For example, before learning 1flip or Oka, it is recommended you learn and practice keyhole
 
Some people have much more natural ability/patience/dexterity/perseverence/free time etc etc than others. Starting off 'at the deep end' may just lead to failure. Starting off with something simpler is more likely to get results. Then, with the skills obtained, it's easier to decide whether to progress to a 'higher' level, and if you do, you already have some of the tools needed, and can appreciate what you need to do in order to progress.
 
No because I think that these "beginner"/"easier" methods is meant for people to understand and learn quickly so learning good methods/algs will eventually be faster in a long run.
 
Apparently this is controversial but yes, it does.

If you want to learn poker, you don't immediately jump into learning about balancing cbetting frequencies or polarising your range from the SB, nor if you learn chess, do you immediately go into learning about lategame king play.

By learning certain algorithm sets or focusing on f2l tricks at the wrong stage, you'll set yourself behind because it will create a cognitive prioritisation of what you've learned rather than what is best.

I don't think this is determined by a flat value ('don't learn this before sub n!') but more to do with factors like experience or overall efficiency or amount of bad tendencies. For example, I think it's silly to start thinking about pair selection when you're having trouble even looking ahead one pair, or having trouble solving that pair efficiently. Similarly, it's suboptimal to work on learning CPOLL when your OLLs are still at >2.5s for recog+exec, since your overall times will decrease that much through unfamiliarity that the supposed benefits (better PLLs) are totally outweighed.

There is an Expected Value for how much each separate 'thing' to learn or practise will benefit you and this changes over time. For example, the expected value of CPOLLs as a group decrease the faster the rest of your PLLs are in respect to those algorithms + your recognition. If utilising CPOLL has no cost then it is purely better since you increase the chance of skips and EPLLs (which for some are faster than the avg PLL). For myself, as someone who doesn't practise frequently, CPOLL is a detriment since my PLLs are all around the same speed, yet my recog+exec of many CPOLLs is slower than the average pure OLL. I basically don't get benefits from EPLLs and the increase in skip chance is pretty much negligible - it just increases the the variance of my times.

Of course, this is answering all in respect to getting as fast as you can be, as efficiently as possible. From a hobbyist point of view then you can learn whatever the hell you want, whenever, as long as you're enjoying yourself.
 
I always ignore such suggestions because to me it makes more sense to get fast using the method you want to use (3-style) rather than getting fast using something else (M2/OP), then fundamentally changing your approach to doing what you're doing and having to get fast again with the new approach.
This. So much this. I always ecourage everyone to switch to comms in BLD as soon as possible. It is counter-intuitive to me to practise huge amounts to be as fast as you can with suboptimal method and then switch to something different. That way you just waste time IMO and can even develop hard-to-break bad habits that you need to change to make your new more advanced method to work properly...

For example, before learning 1flip or Oka, it is recommended you learn and practice keyhole
To me Oka is just keyhole that makes more sense :D I never understood why anyone would use keyhole when there is Oka. I mean they are practically the same thing except that keyhole wastes more moves to insert the last top edge... But then again that's just me
 
...
Of course, this is answering all in respect to getting as fast as you can be, as efficiently as possible. From a hobbyist point of view then you can learn whatever the hell you want, whenever, as long as you're enjoying yourself.

I think what you said actually applies more to casual cubers than to competitive ones trying to get as fast as efficiently as possible. The slow down that happens when you learn a new alg set (for instance) is greatly diminished if you drill it down the way competitive cubers do, while I can see how a casual cuber might effectively become slower if he spends more time recognizing+executing 1 newly learned alg than he did with 2 easier algs and never practices enough to overcome this.
 
Back
Top