Stop after a mistake or keep going?

[toby1]

What's better?

On the one hand I don't want to get into the habit of learning a mistake. On the other I don't want to stop in the middle of a performance.

What's the consensus on best practice for practice?

[Bob]

If it's practicing for practicing, fix it.

If it's a performance or practicing performing, move on.  Cut the loss and just make it one mistake.

[jollisg]

In a performance situation, keep going. In a practice situation, practice the place where you mess up (that isn't really that necessary if you hitted a wrong note that you have never done before, just if it repeats).

[bronnestam]

Don't stop-and-repeat at mistakes when you practice - on the other hand, your practice should not consist of too many "playing through's". Note where you make mistakes, then work with these parts separately, hands separated, different tempos, supporting exercises and so on! Don't waste too much time on playing things you already master.

Sometimes you should make "performance runs" in your practice, when you play just like you were playing for an audience. But there is no need to do this every day.

[iansinclair]

I would add that if you are practicing practicing -- that is really working on the piece -- my own feeling is, as has been noted, to go through a coherent section once, and note where the problems are.  Then, go back and work on them -- but do NOT work on just the measure with the catastrophe in it!  Lead up to it by several measures, and continue after by several measures.  There are several reasons for this.  The most obvious is smoothness and continuity of the playing.  Somewhat less obvious is that the problem may not arise at the actual disaster.  It's not at all unusual for the problem to be something about the measure or two or three before, and, oddly sometimes even the measure or two or three after.

[nanabush]

I think the worst way to treat a mistake is if you stutter on that note three or four times then continue.  It will do that same thing during a performance.

I'd either say go back to the beginning of the section and really work that spot, or take note of it and try playing through... both of these are for different goals (one is to solidify, and the other is probably if you are running the piece the day before an exam/recital).

Just try your best not to hit the note a bunch of time as a means of 'revving up' to keep going.  So many people do that, and it's brutal hearing that in a recital because it is SO obvious to an audience when those hiccups happen.

[outin]

Note where you make mistakes, then work with these parts separately,

Am I the only one who cannot do this? If I try to make a mental note about the mistake my concentration is gone and I cannot keep up with music anymore. And if I get to the end, I usually cannot remember where the mistake was anymore...this is how bad my memory is...

[j_menz]

Am I the only one who cannot do this? If I try to make a mental note about the mistake my concentration is gone and I cannot keep up with music anymore. And if I get to the end, I usually cannot remember where the mistake was anymore...this is how bad my memory is...

Then (assuming your practicing) stop when you make it and do the analysis straight away.

[outin]

Then (assuming your practicing) stop when you make it and do the analysis straight away.

That's what I've always done, but it has it downside...Can't seem to ignore them when I should...Now I have to practice that 

[pianoplunker]

What's better?

On the one hand I don't want to get into the habit of learning a mistake. On the other I don't want to stop in the middle of a performance.

What's the consensus on best practice for practice?

Try playing the whole piece hands separate - not just the hard parts. I got that tip from someone here on PS and it seems to work to help find the issues. The "mistake" may not be where you think it is. If the easy parts have poor fingering leading up to the difficult parts well then you enter the difficult parts all screwed up and dont even know it. Not saying that is your case, but it might be worth a try.

[quantum]

You need to be able to do both, depending on context.  

If you are practice practicing, there is one thing you can do to work on both the trouble section while at the same time playing through the section.  Select a section of music starting a few bars before and a few bars after the trouble spot - or even a few beats before and after if it is a small or rather complex section.  Play that section in a fluid loop, not stopping to correct, but making note of what needs to be fix while continuing to play.  On the next iteration of the loop, make efforts to fix the trouble again without stooping.  The important focus is that you do not stop the loop for whatever reason, continue to play if you make a mistake, continue to play while you think about the fix, continue to play even if it means omitting notes so you can get through the passage, no matter what happens you must continue to play and loop that portion of music.  

The transition between the start and end loop points don't have to be theoretically perfect, you can have a weird harmonies or phrase breaks.  The point is, do not stop when you do the repeat: loop seamlessly even if the harmony, phrase, or counterpoint seems broken at the chosen points. 

[timothy42b]

If you stop when you make a mistake, you are practicing stopping and you are sure to do it in performance as well.  You may not stop completely but you will stutter.  The most naive audience can hear the stutter when they would never catch a wrong note.

So go on past it, but not too far.  Choose to stop at a logical place.

Then go back and learn that section until it is rock solid and better than the rest of the piece, so you'll never worry about stumbling there again. 

[slobone]

I prefer a third option -- don't make mistakes in the first place. I'm serious! You can play in such a way that you never put a finger on the wrong key, or even the wrong finger on the right key. Naturally I don't mean 100%, nobody's that perfect, but just get out of the habit of barrelling ahead and making mistakes.

Of course this means you have to play slowly, sometimes very slowly, and stop if you're not sure what to do next. When I'm doing this, I don't even try to play at a steady tempo. I slow down for the hard parts and speed up for the easy parts. Once I know the piece, it's usually easy enough to correct the rhythm.

[timothy42b]

Of course this means you have to play slowly, sometimes very slowly, and stop if you're not sure what to do next.

Or pick easier pieces.  One of the teachers here says if you can't read it at half speed it's too hard.

If you're not sure what to do, don't stop; keep playing the last note in rhythm at tempo.   

When I'm doing this, I don't even try to play at a steady tempo. I slow down for the hard parts and speed up for the easy parts.

I think that's dangerous for a lot of reasons.   I would never recommend disconnecting from time.  See Whiteside and Caruso. 

[toby1]

I'm going to read this at some point:

https://www.pianostreet.com/blog/piano-news/practicing-perfection-memory-and-piano-performance-5886/

Found the link on the website and looks suer relevant. Everyone's advice so far has been an interesting read. No set consensus yet. I'll also hit up a book I had sitting at home that might have some relevant input.

Cheers

Toby

[lloyd_cdb]

When first learning a piece, I assume people start with the hard sections. If you are making mistakes in that section, you need to break it down into smaller parts or accept that it's too difficult for you.

Once you've learned the hard sections: If you are making mistakes on easier sections, it depends on your mistakes. If you are making the same mistake each time when playing through medium+ sized sections, you need to break it down and work on just that smaller section. If the mistakes are in different spots each time, playing through it isn't going to be reinforcing mistakes. They are random mistakes that just mean you need to bring it up to performance quality.

So, I don't see a reason to ever stop suddenly. You either know it or you don't. If you know how to play it, it's simply just a mistake and you should be treating it as a performance mistake/recovery. If you don't know how to play it, you need to be focusing on the smallest possible way to break it down and actually learn it*.

*without sacrificing the motions you would actually use when playing through it. Sometimes people will break it down into a way that makes no sense when coupled with the section before or after. That is where proper preparation comes into play before you even touch the keyboard.

[slobone]

Or pick easier pieces.  One of the teachers here says if you can't read it at half speed it's too hard.

That would have kept me away from a lot of pieces that I ended up loving. Reading a new piece is my favorite part.

If you're not sure what to do, don't stop; keep playing the last note in rhythm at tempo.   

I think that's dangerous for a lot of reasons.   I would never recommend disconnecting from time.  See Whiteside and Caruso. 

Yes, that's what every teacher says. it took me 50 years to get up enough courage to do it the way that works for me. My problem is focusing my attention, and coordinating my mind with my fingers. If I allow myself to make mistakes, they get hard-wired in my brain.

One possible way to prevent this is just to play the whole piece really slowly. But why should I do that when I've already learned most of it correctly? That just makes practicing a chore.

Another approach is only to work on the tough parts. And of course I do spend more time there. But I also work on phrasing from a very early stage, so it's counterproductive to work on single measures. And in fact I've never had any trouble playing with a steady tempo once I've learned the notes correctly.

[slobone]

When first learning a piece, I assume people start with the hard sections. If you are making mistakes in that section, you need to break it down into smaller parts or accept that it's too difficult for you.

Once you've learned the hard sections: If you are making mistakes on easier sections, it depends on your mistakes. If you are making the same mistake each time when playing through medium+ sized sections, you need to break it down and work on just that smaller section. If the mistakes are in different spots each time, playing through it isn't going to be reinforcing mistakes. They are random mistakes that just mean you need to bring it up to performance quality.

So, I don't see a reason to ever stop suddenly. You either know it or you don't. If you know how to play it, it's simply just a mistake and you should be treating it as a performance mistake/recovery. If you don't know how to play it, you need to be focusing on the smallest possible way to break it down and actually learn it*.

*without sacrificing the motions you would actually use when playing through it. Sometimes people will break it down into a way that makes no sense when coupled with the section before or after. That is where proper preparation comes into play before you even touch the keyboard.

Excellent advice.

[g_s_223]

One of the great pleasures of being a musician, of whatever type, is ensemble playing. For a pianist, this would be as a duo partner with another instrumentalist, or in a piano trio/quartet etc. In this situation, it is absolutely essential to retain coherence of the ensemble by not stopping after minor errors. If you stop, you will frustrate everyone involved: you can usually ad-lib until you are back on track. Also, this relates to the "counting" issue: ensemble playing is a great discipline.

The practice scenario is somewhat different: in this case, just keep on slowing down until the notes are entirely correct.

[Bob]

If you do make a mistake during practicing, be careful not to swear.  Later during the performance the audience will be able to read your lips.

[nyiregyhazi]

I think that's dangerous for a lot of reasons.   I would never recommend disconnecting from time.  See Whiteside and Caruso.  

I think you're missing Whiteside's point about physical rhythm. It's about consistent flow of motion- not literal mechanical meter. If you press on into passages you're not ready for, you do more disservice to this physical flow (which is what Abbey Whiteside refers to as "rhythm"- in a completely different way to regular use) than you do by stretching the time in order to prolong ease and physical continuity. If Whiteside herself said NEVER slow for more difficult passages when practising, then she missed how destructive enforced guess work (for the sake of artificially mechanical meter) really is to the ability to make a physical rhythm that flows naturally without jolts or wild jabs.

One of the most valuable things I've learned about improving sightreading is not always to press on rigidly and fake everything or leave virtually everything out. There's an alternative where you expect yourself to stretch enough to make space for every note to be played with assurance and comfort-  even at the first reading. This is not the same thing as taking crazily fast speeds for easy bits and slowing outrageously for hard bits- but instead involves plenty of space and comfort in general, but a sense that you can always give yourself a touch of extra space when necessary. When you have both approaches, you learn way more than when you feel that constant time always takes precedence over having a clue as to what notes you are trying to play.

In general, I'd always sooner hear a student totally disconnect from time once- but ONLY once- than press into something they have no clue how to execute. Beyond that, you can better still stretch in a way that doesn't even disconnect from time altogether but merely frees it up when useful. But guesswork for the sake of maintained rhythm is simply crazy, when you're hoping to actually learn the piece at some point- rather than fake it once and never play it again. It's not effective to judge all reading from what you'd have to do if sightreading with another instrumentalist. It makes no more sense than saying that boxers should only practise against opponents who will hit back. If you only learn from make or break scenarios where you'll be getting tonnes of stuff wrong (without training from a whole range of different approaches) then you're just limiting your chance to improve.

[birba]

Of course this means you have to play slowly, sometimes very slowly, and stop if you're not sure what to do next. When I'm doing this, I don't even try to play at a steady tempo. I slow down for the hard parts and speed up for the easy parts. Once I know the piece, it's usually easy enough to correct the rhythm.

I do the same thing!

[hfmadopter]

I do the same thing!

Of course there is a distinct difference between playing a piece and a wrong note has crept in and newly learning a pieces hard parts. In the first case, go back and find the trouble on it's own, separate from a play through. In the second don't even play through yet till you think you really know the trouble spots and can run through them with two or three added measures on each side of the rough spot.

I wouldn't even worry about a single miss or landing on an incorrect note accidentally on a play through unless it starts to keep reoccurring, then it needs to be addressed.

That's My take on this issue anyway !

[timothy42b]

I think you're missing Whiteside's point about physical rhythm. It's about consistent flow of motion- not literal mechanical meter.

I don't miss her point at all.  I've snipped the rest of your post because it's one big straw man.

Please look into Carmine Caruso's "timing it in" concept and see if you think it fits with Whiteside's "rhythm."  I think they are very close to the same thing.

[lloyd_cdb]

Or pick easier pieces.  One of the teachers here says if you can't read it at half speed it's too hard.

I'm of a similar belief. Playing it slow is a relative term, and to everyone in this thread it means a different thing. When I say slow, I usually mean 2/3 - 3/4 speed. I rarely drop below that, the exclusion being when practicing 5 against 7 or something of that nature which I might be struggling with from brain issues and not technical issues. Think about the body motions that go along with throwing a ball at half speed. Do you use those same motions at full speed, just double as fast? Probably not. Piano is similar if not worse in that regard considering it takes so much more specific muscle control.

[nyiregyhazi]

I don't miss her point at all.  I've snipped the rest of your post because it's one big straw man.

Please look into Carmine Caruso's "timing it in" concept and see if you think it fits with Whiteside's "rhythm."  I think they are very close to the same thing.

My take on Whiteside's rhythm is that motion never stops dead. If you satisfy that, you can do whatever you want. Stretching the timing for a harder passage, in the learning stages, actively helps to maintain a physical rhythm that flows onward without stops. The only sin is to rush for no good reason through an easy bit- before being forced to stop dead while you then figure out the next note. Even then, a dead stop done once always beats a stab in the dark for the sake of not stopping. The key is what you do the second time, not to be forcing yourself into casual screw ups out of some pedantic idea that rhythm must always come first. Good pianists don't guess for the sake of keeping meter when learning pieces. They only do so when they have no choice but to be in a performance mode at the first play through.

There was no strawman btw. You said


"If you're not sure what to do, don't stop; keep playing the last note in rhythm at tempo. "

Its disastrous to make this anything more than an approach that is reserved for pressured performance based sight reading ie the exception and not the rule. Guesswork breeds nothing but amateurish results, if that's your norm. It's for emergencies in which you have no choice, not something to be done casually. Ironically it does more to interfere with the physical rhythm of movement (by causing wild jerks and seizures) than allowing the freedom to stretch things when they get harder. You learn more from a fluid and and natural link between the correct notes than from doing an erratic and probably incorrect movement for the sake of strict time. The first merely needs to be reorganised. The latter tells you nothing about the correct flow of physical movements that will be of use in the final product. Even if you guess correctly, notes that have been guessed at are usually approached jerkily. Even if technically in time, the movement will not have flowed naturally or effectively- rendering the physical quality useless. Guess wrong and it's all the worse.

[timothy42b]

I'm of a similar belief. Playing it slow is a relative term, and to everyone in this thread it means a different thing. When I say slow, I usually mean 2/3 - 3/4 speed. I rarely drop below that, the exclusion being when practicing 5 against 7 or something of that nature which I might be struggling with from brain issues and not technical issues. Think about the body motions that go along with throwing a ball at half speed. Do you use those same motions at full speed, just double as fast? Probably not. Piano is similar if not worse in that regard considering it takes so much more specific muscle control.

I'm not opposed to going very slowly if done steadily.  Unlike the argumentative N, I think learning is usually faster if done connected to time.  There are many ways to simplify; one can go slow, or go fast on very short sections, or do one hand, or skeletonize, and and and.  Detaching from time is one of those methods, and many people use it.  I observe that pianists in general have the worst sense of time of any instrumentalist, outside of organists, and piano students succeed in playing fluently less than any other type.  I think that is directly due to practicing outside time, but of course I can't prove it and have little incentive to argue it with N.  Think about it and reject it as you see fit. 

You may not be aware of Carmine Caruso's (1904-1987) theories.  He was a New York brass teacher of some fame.   One of his insights was the use of time as a unifying function to coordinate complex activities that had to happen simultaneously but couldn't be done consciously.  He did this through specialized exercises and lots of subdividing.  If you're not a trumpet or trombone player most of his exercises (he called them calisthenics) would be meaningless, but not the subdividing. 

Playing in strict time can produce error, and this can be good.  By error in this sense I do not mean a wrong note, and OF COURSE NOT a wrong rhythm!   Error is the raw material for technique improvement.  You alluded to the fact that we often perform motions differently at speed than we could do them slowly, and I agree.  Almost any improvement in technique has to be figured out and that requires trial and error.  (some gross postural items can be seen and directed by a good teacher, but the nuances have to be developed)  Error produces a variety of small differences, some of which work.  Repeated often enough one starts to identify the right motions.  This is why the incremental speedup practice methods can be detrimental.  They reduce error by overlearning the motions that work at slow speeds.

[nyiregyhazi]

Regarding your first paragraph, your diagnosis is correct. People stop way too often. But the idea that the answer is to take a shot in the dark is madness- except when practising performance under pressure. it's as good as saying that someone who is slow at suduko should have the odd guess in order to keep things moving.

The way to fix a point which you don't know how to connect is to EXPERIENCE the physical connection with clarity and certainty. Thinking that forcing yourself to have a go will produce technical precision in advanced repertoire is futile. Forget executing a chopin etude with such an approach.

The reason people keep stopping in the same places is that they never properly experience the missing link. All it takes is to do that once- preferably under no time pressures, so the experience of connection is clear. From then on, proper timing must be expected. The idea that having a wild guess from stage 1 is better is simply ridiculous. REPEATING wrong rhythms is bad. Feeling your way around once and then immediately adding correct timing is not.

And no, you don't need to "learn" from errors. You can play precisely from the outset. You can learn from ways that expose faults in your quality of movement that were unknown. If you have to guess to preserve the rhythm, you already know that it is improperly learned, from that itself. You cannot learn a thing from guessing at a note that you are not ready for. You can only prove that you were indeed not ready. In other words, the only thing to be learned by trying to guess a note and botching it is that you had no business hoping it might work under the needless pressure of strict time when you didn't yet know how to execute the necessary physical connection. good pianists recognise the inadequate preparation without having to keep screwing up to find out about it. They stop to learn necessary connections and only then put any pressure on the execution.

Details can be different at fast speed, but the distance between notes never changes. If you don't know how to span a distance with assurance, putting yourself under pressure (due to the false belief that bad movement in strict time always trumps a patiently controlled movement in freer time) is foolhardy, for a serious pianist, not wise. No balance pianist ever evolved by looking at things from a single dogmatic viewpoint, that is never shifted according to the necessities of the moment.

[timothy42b]

. But the idea that the answer is to take a shot in the dark is madness- except when practising performance under pressure. it's as good as saying that someone who is slow at suduko should have the odd guess in order to keep things moving.

Whose idea would that be?  NOBODY here has suggested anything of the sort, except you.

And no, you don't need to "learn" from errors.

You don't learn from mistakes, but you do learn from errors.  If your motion is not correct, you must somehow change it, try something different.  YOU DON'T TRY A DIFFERENT NOTE!  which seems to be what you're claiming.  It's much more subtle than that.  The raw material for all technique improvement is variation.  Without variation, you have no chance of ever hitting on the right combination.  And variation IS error. 

High range on brass instruments is a mystery that forever eludes many students.  How do you play up into the stratosphere like the pros do so effortlessly?  It isn't strength.  It really is effortless when you do it perfectly.  There are some basics that you have to do correctly or they'll positively prevent you from getting it.  Any good teacher helps you set up for success, avoiding the classic embouchure faults.  But once you're working correctly, you still need to get the knack of it.  You need to accidentally produce it once, then occasionally, then regularly, then consistently. 

[nyiregyhazi]

Whose idea would that be?  NOBODY here has suggested anything of the sort, except you.

You don't learn from mistakes, but you do learn from errors.  If your motion is not correct, you must somehow change it, try something different.  YOU DON'T TRY A DIFFERENT NOTE!  which seems to be what you're claiming.  It's much more subtle than that.  The raw material for all technique improvement is variation.  Without variation, you have no chance of ever hitting on the right combination.  And variation IS error.  

High range on brass instruments is a mystery that forever eludes many students.  How do you play up into the stratosphere like the pros do so effortlessly?  It isn't strength.  It really is effortless when you do it perfectly.  There are some basics that you have to do correctly or they'll positively prevent you from getting it.  Any good teacher helps you set up for success, avoiding the classic embouchure faults.  But once you're working correctly, you still need to get the knack of it.  You need to accidentally produce it once, then occasionally, then regularly, then consistently.  

I don't greatly care about brass technique. We're talking piano technique. I've never yet encountered a note that I benefited from getting wrong the first time. I've certainly wasted a whole lot of time by merely having a go though. Hitting a piano key correctly is not hard. The problem in hitting a brass note is that you don't know what you're looking for until you've experienced it. In piano, you have eyes and spatial awareness to judge whether you are on the right keys or not. Bad comparison, sorry. Anyway, the point is that you stated in plain English before that if a person doesn't know what is coming next, they should press on and have a guess- not merely in pressured performance but as the norm. This is madness, however you wish to phrase it.

You can learn from making mistakes that CANNOT REASONABLY BE ANTICIPATED and then discovering holes that YOU DIDN'T ALREADY KNOW OF. The only thing to be learned from guessing a note that you know full well that you are unprepared to play, in strict rhythm, is quite how foolish it is to have a stab in the dark rather than give a chance to get to it with a clear movement. Even if you luck on getting it right, the movements that stem from guesswork are rarely fluid or useful movements.

[timothy42b]

I've never yet encountered a note that I benefited from getting wrong the first time.

You insist on talking about wrong notes when it's been explained many times that that's not what anybody is recommending.

It is impossible to discuss anything with you.  You are impervious to any ideas but your own. 

[nyiregyhazi]

You insist on talking about wrong notes when it's been explained many times that that's not what anybody is recommending.

It is impossible to discuss anything with you.  You are impervious to any ideas but your own.  

I didn't say you recommended wrong notes. If you think that recommending that a person presses on into a guess in strict time is the answer, (if they're not sure what to next) you're not getting off on the fact that you didn't technically "recommend" wrong notes. Telling a person to walk around a crowded public place blindfolded and firing a machine gun in random directions is not technically recommending that they kill anyone rather than fire safely into the ground. The point is that they will do so if they follow the the advice, not the technicality of whether the advice explicitly requested murder. Learning pieces by playing guessed notes in strict time when unprepared for the next is dogmatic madness that totally misses the most basic common sense. It's a totally amateurish attitude that ruins things before they get off the ground and which is reserved for pressured sightreading- not serious learning of pieces.

There is more transferable context in a series of legato connections played in free time (as long as the player is capable of hearing the correct rhythm in their head) than there is in a flawlessly rhythmic execution of a phrase that contains even a single error.

[timothy42b]

Most of us are here to learn from the discussion and to gain fresh viewpoints.

You are only here to win debates.

[nyiregyhazi]

Most of us are here to learn from the discussion and to gain fresh viewpoints.

You are only here to win debates.

Your one - sided viewpoint of the whole is as old as the hills. I was brought up on it. Fresh? I can't overstate how much my sightreading improved when I stopped forgiving my perpetual sloppiness for the sake of maintained rhythm, and balanced that with work in which I expected note perfect execution, aided by rhythmic stretching whenever beneficial to that end. These days, I can sightread such repertoire as chopin mazurkas near note perfect and without pauses.

If you want to repeat the same old dogma that everyone else says over and over, fire away. I'm illustrating a fresh perspective that sees the bigger picture, without insisting some ridiculous rule that is based on the fallacy that random notes in rhythm train more about a piece than correct physical connections- that are separated from time once and ONLY once. Try opening your mind to a fresh viewpoint that involves benefits that both slobone and myself have observed- rather than repeating second hand information (that offers only a small part of the puzzle) like a broken record.

[4greatkeyboards]

I just have to remark on the high quality of all the answers I see on this question. I think they are mostly all valid.

I also note how active the response to this question is. We all are very sensitive to this topic.

For my 2 cents I offer these:

- there is phrase polishing and
- there is performance polishing.

If we are going to perform or record we do performance polishing mostly in the weeks prior to the act. In performance polishing we don't stop and correct. We strive for a wider arc. We concentrate on dynamics and nuance. We ignore single mistakes for the flukes they are, or should be. We concentrate on putting together all the many perfected phrases we have learned.

Prior to this we are in a mode where we are learning just to play the notes. That is long before performance polishing: months, likely. In this period we take the piece apart and perfect any hard phrases. We will learn the piece faster if we do not repeat any mistakes. So we play these phrases slowly and without error. For instance: play a difficult phrase over and over until you can play it 10 times without error, then move on.