Puzzles/Logic puzzles/Sudoku/9 by 9/Approach to solutions/Mistakes
So you have been applying, using ink, the technique described, dotting available positions for numbers, writing in the numbers where you identify the cell's content.
Then you come to a contradiction. There is no remaining cell, or you notice that you wrote the same number in two cells in a box, row, or column.
So start again from the beginning. You may be able to see your first entries as numbers in cells with no dots, or with very few dots. When you find a number whose position is definitely correct, from the original printed numbers, mark all four corners of the box, making four edges of an octagon. This is easy enough to distinguish from a box with only a number written in it. Continue reviewing the numbers, now considering the marked boxes as if they were printed. At first, just ignore numbers you cannot yet verify. Treat those boxes as if they were empty.
This is substantially more difficult than starting with a fresh puzzle! The dots may not be very usable! However, with puzzles of up to intermediate level, it is satisfying to recover from an error.
It is possible to go through this a second time, by now filling in the corners (the four corners of the cell).
The goal, however, is to test mental acuity, not to beat one's head against the wall. From a book of Sudoku, one may see error rate as abandoned cells. Do not correct the Sudoku from the answers, there is no value in the actual completion of a puzzle and it will only hide the errors. If one doesn't want to let errors be errors, use pencil!
It is an intermediate practice to, when confirming positions to be marked as described above, verify a considered mark from the book answers. If one is concerned about error rate, it would be useful to mark the number of times one has thought at cell was done, but the answer was different.
Remember one possibility: that a puzzle from a Sudoku collection has more than one solution. Supposedly that does not happen. However, it could. A goal is to prove that the answer filled in is the only answer. This becomes especially important with Advanced solution techniques where one tries one of a pair of possibilities. If there are two solutions, one might then find one of them, it works. So there is then a technique, part of advanced techniques, that will also test the opposite choice.
Mistakes indicate some lapse of focus. This author is 71 years old, and will sometimes see a number, not as the number, but merely an unavailable cell, with a "something in it," only looking for empty cells, and will then dot or fill in the remaining cells for the same number, leading to an error if this is not then noticed. Another error is to miss a possible cell and dot only the others (or fill in a position when there is, in fact, another possibility.)
This is related to tunnel vision that commonly develops with age, so the error rate is a measure of this. It is normal, and does not mean one is becoming senile! Young people can have the same lapses. The author, however, notices that error rate correlates with his general mental state. Is he calm and relaxed? Is he upset about something? It can be extremely useful to be reminded of how our emotional reactions affect our mental performance.