A tiny example of the value of practice

Some while ago I practiced the knight and bishop versus king ending against stockfish until I was able to mate regularly in no more than 38 moves no matter what the starting point. I could never have done this a year ago, but managed now to pull it off in a 2/1 bullet. Actually I had an extra knight which I occasionally used to correct my clumsy blunders, but I employed the knight-bishop triangle method to minimise thinking: we were both under 10 seconds during that entire ending; sometimes 1 or 2 seconds left. The 50 move rule also potentially loomed.
I guess my childhood piano teacher was right after all about the importance of serious practice ;-)

In 2+1 time control!
Excellent

I'm amazed that people can play bullet, I don't dare to play blitz

@nayf:
How long did you have to practice to reach your goal?

@bracenoodle

Cheers. I don't think blitz (far less bullet) is of much value at early levels, but as you progress in your knowledge, it does have some use to practice position recognition. You can have a decent game in a 5/3 blitz.

@cuxtraining

I'm not sure if you mean specifically learning the Knight-bishop ending or something more general. Regarding the former, last year I watched a good video on Youtube on the triangle method, and then practiced it for several days against the engine. I guess it took me about four or five days of practice, several hours a day, to really get a handle on it. I should say, the engine is hardly as obliging as my real life bullet opponent. It heads straight for the wrong corner (opposite colour of the bishop) and makes you work to dislodge it from there and set up the first triangle. I had to figure out the method for doing that myself. When you don't know what you're doing you can consume 30 or more moves just for that, and then you'll never mate the engine before the 50 move draw. But I highly recommend learning this method, as its value goes beyond this specific type of ending. You learn how to coordinate knight and bishop, whose respective movements seem at first as awkward a combination as chalk and cheese. You discover, for instance, that they are very strong when on the same colour separated by a square, because the knight then controls the opposite colour squares around your bishop's diagonals. You can hem a king in like that. After practice you start "seeing" the box corner they make. And you learn how to coordinate king and bishop, very useful for many endings and more.
Regarding other chess goals, I haven't reached any. The struggle to lift myself out of mediocrity is ongoing.......

@nayf:
Thank you for your detailed explanations. But the time it takes seems to me to be too high for the possible profit. Maybe you only have a few times this positions on the board in your whole life. I've never had it on the board yet.
But if learning the method is fun for you, of course you can do it.
I will continue to give preference to tactical and other types of training.

This is pretty incredible

I read somewhere that only around 1% of high level tournament games have ended in knight-bishop versus king. Still, if you play 300 games in a given period, that's 3 games you might have won which you may end up drawing. I suppose as well that the method might be useful to know if you are ever the one with the lone king and are trying to force a draw. I did mention I believe the method has benefits beyond that specific ending. But having now viewed your profile, I'm embarrassed to be making any recommendations. I should be humbly asking you instead what types of training you find most useful. Is there anything aside from doing tactical puzzles that you have found very useful? And by the way, don't assume that because it took me four or five days of practice to learn how to beat the engine with knight-bishop that it would take you nearly as long! Maybe it would take you one hour.

It‘s by far less than 1%... But either way you have to gather some tens of thousands of chunks. This particular endgame belongs rather to the easy stuff.

I know it's relatively easy for a good player. I, however, have learned it only this past year. Like many skills, it's easy once you master it; it seemed hard before. But I'm glad you're telling "cuxtraining" above that it's easy; it should encourage him!

I guess you're right, if the blogger here is correct that only one in 5000 games ends in NB v. K. That would be 0.02%. So I shouldn't be proud of my little bullet win as it was too easy, but I can cherish it as a statistical oddity!

www.chess.com/forum/view/endgames/endgame-percentages

This endgame technique isn't a chunk, but a method or technique (creating sequences of many moves), and considerably more complex than some well-known rook ending techniques, such as the elevator. Mastering it gives you a few hundred chunks right there.

...well... that's not a N+B ending. Not trying to detract from your achievement... it is considerable... but let's label it correctly