I've already written a previous post on this topic. However, it's such a major issue that I decided to make another breakdown of some primary issues, in a slightly more concise post. So, do you want to be a good sightreader? If so, to understand the main foundation behind the skill, let's start with a simple indisputable truth.
The level of your sightreading (which we usually define strictly as a first ever attempt at new material) will always fall somewhere beneath what you could achieve by learning a piece over a long period. That is just obvious and reveals nothing. However, far more useful is to consider that it will ALSO fall short of your ability after "quickstudy" of a piece (which we'll define here as up to three or so short practise sessions). Your quickstudy level is typically the main yardstick in getting a realistic sense of your current potential sightreading level- especially so, if you consider sightreading to be an area of weakness.
If what you achieve in a handful of practice sessions does not reach a reasonable degree of comfortable fluency (which is commonly true of those who also struggle at sightreading), how could there be a chance of success when that little preparation is replaced by absolutely none at all? Also, consider now what traditional sightreading practice is supposed to be focused upon- ie playing brand new material only once, without ever stopping to resolve errors or uncertainties. Is there any meaningful reason to imagine that this could be an effective way to develop an initial foundation of quickstudy skills? Someone who is capable of good quickstudy will most likely do fine when given a single attempt. For others, however, it can create conditions in which sightreading is guaranteed to feel like an extremely unpleasant struggle, from which almost nothing seems to be gained. It's not surprising that they might thus feel inclined to avoid it.
Anyway, to summarise an important truth in rather blunt terms- if a pianist does not raise their capacity for quickstudy to at least the level they would hope to sightread at (and preferably a good deal higher), it follows that their sightreading skill can only be worse still! If you want the results of sightreading to even be adequate, your potential for quickstudy would first have to rise notably above that level.
At this point, I suspect that some will raise an eyebrow at the rather coldly direct language, so let's lighten the tone a little! I would certainly aim to find positives when personally addressing any individual and their skill level. However, the abstract working here is not criticising any actual person. The point is just to highlight a universal truth, for the sake of demonstrating a clear way forwards. Only when small amounts of preparation can be expected to give good results, can half-decent results start to arise in wholly unprepared scenarios. If most repertoire is exclusively based on weeks (or even months) of gradual preparation, to then attempt something only once ignores a gaping wide middle ground. It's not helpful to behave as if we have a dial with merely two settings- which move between very many hours of preparation or absolute zero. It is solely an issue of language that "sightreading" refers to a first attempt. It's easy to carry this over and imagine that these are the only circumstances in which it is possible to gain skill. However, that would be like assuming golf can only be improved by playing a round of eighteen holes. No. There are driving ranges and putting greens, before you have to test the overall product to the full. It's widely true in just about any field, that you don't best practice towards something by working exclusively within the exact narrow circumstances of the end product. That's only part of a good method.
So, if you want to be better at sightreading but are not currently good at quickstudy, there is a realistic route to real improvement. Don't try to obey the traditional sightreading advice and don't worry exactly what counts as sightreading. Develop quickstudy skills (by playing around with a wide range of moderate preparation times) and you'll have a real foundation, which you can later use towards greater focus on first time sightreading.
As an example, here's a film of myself playing an arrangement of Barber's adagio.
I'm a pretty good sightreader, but what this demonstrates is the product of some quickstudy (if you wanted to see some first time sightreading of Bach and Rachmaninoff, I included some in my other post on this topic). I found the score on the same day as the playthrough- which came after reading through two or three times (with only small amounts of practice to resolve a few errors at source). Is it perfect? No, I had a couple of slight slips in the notes (and didn't properly voice a couple of notes of an inner line). However, it was largely fairly close to what I intended, via just small amounts of time. There had been a few more errors on my first playthrough. I gained far more by going back to briefly repair them, than I would have by walking away.
It's often good to commit to finishing a piece once through, no matter how many errors. But you learn far more if you then go back and take control over those moments. It doesn't matter in the least that this is no longer strictly "sightreading". I don't memorise quickly (either in terms of developing muscle memory or a mental image of the notes) so I still have to process written notes if I want them be correct for the second or third time. Although there will be some expectations due to familiarity, to play again doesn't instantly stop using most of the same foundational skills. The reality is of another sliding scale of change. Also, we particularly need to consider anything that didn't actually work first time. When you feed your system with corrections, it greatly increases the chance of coping with similar demands, in the future. You don't learn much from something you have neither processed correctly, nor executed correctly, except when you take a little time to at least figure out how you could have.
When repeating something over and over, the benefits obviously do start to drop off. After some point, repetitions do little (if anything at all) for reading skills. Different people develop memories at different speeds, so there's no simple rule as to how much you can repeat before the main value has passed. However, what I'd say for sure is that you will never realistically lose the majority after only a single playthrough. In fact, I'd suggest that the second attempt of any passage is typically MORE valuable than the first, with regard to building your skill set- especially when you didn't get it right the first time. The sole exception would be where something was so easily within your level that there are simply no issues to go back over.
Although good sightreaders have to do some creative faking, it's important to remember that this is only the icing on the cake. To stretch the analogy a little, you could smooth over a few minor dents in the surface of a slightly imperfect cake. However, you can't expect to take a random blob of formless sponge, and then ice it over so skillfully as to give the impression of a flawless artisan wedding cake. On the same note, you can't simply fake your way around so severely as to survive without a basic ability to accurately process the primary foundations of a score. You don't get notably better at that by painstakingly spending six months learning a work, one note at a time. But neither do you get there by gritting your teeth to play to the end of a piece at any cost, and then walking away for good- regardless of how the result was. You may learn some mental resilience, but not the skills to translate a future score into a more accurate result. The best development comes when you figure something out efficiently and then walk away. If you can't do that in one session yet, do it over two or three, or lower the level of difficulty until you find music where this is possible. And keep repeating this kind of process until you expect to be able to learn pieces within a single sitting.
The time for a strictly traditional sightreading mindset is when someone can get good results from quickstudy, but has a specific struggle to avoid stopping and starting during their first attempt. That would be a sure sign that they should concentrate chiefly on a little more faking/guessing, in order to get past moments of uncertainty without stopping the flow. You can adapt the foundation of a quickstudy skill set into a sightreading mindset, by simply slightly loosening your expectations of accuracy and prioritising rhythm. Approximation can fill a few gaps. However, what you can't do is to take someone who has the skillset to process only a small minority of the information in real time, and expect a radical level of guesswork to magically fill in for the absent foundational skills. It doesn't give good results in the short term and it certainly doesn't cause any long term improvement of processing skill. Traditional sightreading ideals only complement primary background skills, while doing nothing to help generate those that are missing. These foundations are much better cemented by getting something right on a second or third or pass, than by getting something wrong on the only pass ever made.
Sightreading should not even be seen as an altogether different thing to quickstudy, but simply as the most advanced version, on the end of that scale. It's quickstudy with quite so little preparation, that there isn't actually any at all. However, if we say faking makes up ten percent of the whole here, the other ninety percent is simply seeing the information, picturing the meaning and acting correctly upon it. Would you expect to improve at maths by turning up for a test, guessing more than half of the questions and being told you got a mark of less than thirty percent? For any long term positives to follow, you couldn't simply book another test and hope resilience might eventually win out. You'd want to get the paper back, and practice going over those questions that you'd had to guess at.
Let's take the analogy a little further. Say that a child had to prepare for a speed mental arithmetic test with ten questions that allow five seconds each. First, you'd probably want to ensure that they could score 9 or 10 with thirty seconds per question. Then you could start pushing them to try answering questions in just a few seconds and see how their instincts fare. If they now got a few more wrong, it likely wouldn't ruin their existing foundations. But what it a child typically scored 4 or 5 with thirty seconds per question? Would anyone think it would be useful to now push them to answer more directly and instinctively? Would seeing them fall to a score of 1 or so achieve anything? As a one time thing you could (possibly) argue that it might give them a sense of what they will be practising towards, as a future goal. However, if they were made to take the pressured tests over and over, that would be truly deranged as a learning method. It would be more likely to muddle and weaken their foundational skills further, than to improve them. As a famous quote (wrongly attributed to Einstein) says, madness is doing the same thing over and over expecting a different result.
Unfortunately, the traditional sightreading advice of simply ploughing through brand new material is almost perfectly analogous. It forces panicked responses, due to overstretching of foundations. We sometimes need to test ourselves, and do what we can to make it through. However, an excess sense of struggle is both mentally unpleasant and confusing. Deep progress also needs underlying skills to be advanced, via (slightly) more comfortable timescales that allow for corrections. When we recognise that true sightreading is only the most extreme end of the wide spectrum in which we can practise, we are much more likely to get results.
One further thought- I think part of the problem with sightreading advice is that it was usually given in the past by teachers who had students play numerous short exercises by the likes of Czerny etc, or at least plenty of sonatinas and easy duets etc. Whatever you might think about the physicality of exercises, this certainly meant a lot of quickstudy and thus plenty of foundations to bring into sightreading. Today, it is not so common. Exam focused students often tend to learn just three pieces per year- without having plenty of easier projects to dip into on the side. Something that is very popular today is also self-learners- who often jump straight to more advanced repertoire and try to painstakingly build it by a detail at a time. Such learners may never have spent any significant time on easy pieces, never mind got used to learning them efficiently. If sightreading feels like some kind of special outside thing (that is done solely in relation to sightreading exercises), you almost certainly need to view it in a whole different light. The best sightreaders never typically had to practice it as a special project. They just learned a wide range of pieces over a wide variety of timescales.
