Struggling Badly (Queen Vs. Knight Endgame)

Hi all, I'm looking for any kind of advice that could help me solve this endgame. I'm able to pull it off some of the time, but no matter how many videos I watch, or how many times I watch them, I cannot figure out the logic behind this maneuver. For some perspective, I haven't found ANY of the other endgames to be anywhere near as challenging, and that includes Bishop+knight mate--his mate just doesn't make a shred of sense to me. It seems like I can only win by getting lucky. Please help.  

Hey WrightJustin!

The technique for checkmating against a king and knight is a bit nebulous. Im sure people have made step by step fullproof techniques for doing it, but really, you just need to follow some basic principles. 

You'll want to do two things. First, look to drive the opponent's king to the edge/corner of the board and checkmate them. Second, look for opportunities to fork the king and knight. As you drive the king closer to the edge, oftentimes, the opponent will be forced to allow a fork in order to prevent checkmate.

My only other piece of advice is to watch out for forks. On every move, ask yourself if your opponent can fork you king and queen. One simple method to avoid it is keeping them on different color squares, though you dont absolutely need to.

To practice, I'd set it up in analysis mode and just drill it against higher and higher level engines, until you can do it consistently against the top engine.

Thank you for the comprehensive advice. I did manage to "solve" it, in the sense that I can do it virtually every time from a random position. 

The issue I'm running into now is that I've started the endgame drills, the section with the dozens of king and pawn games, and these are even more difficult than the mates in the main section. I'm not sure if I'm an idiot, or if these are legitimately hard. I've only been playing chess for a few weeks, so most of the game is challenging, but I'm typically quite good at solving puzzles. These, on the other hand, are 10+ move mates, and if you make a single error you either draw or lose the advantage. 

Am I crazy?

Nah, you're not crazy.

Didn't know you had just started chess. Knowing that, I'd actually recommend you don't focus too heavily on pawn endgames yet. First, I would learn king opposition, and in particular, how to win a king and pawn vs king endgame, which looks like this (white to move):

...and defending against a king and pawn, which looks like this (black to move):

And by learn, I don't just mean know how to do it, but instead practicing this again and again until you can do it in your sleep. I like to plug it in to the analysis board and do it multiple times against the highest engine. I'd also study A/H pawn and king vs king, and how that changes the dynamics of the position. Then after that, I'd begin looking at pawn endgames.

A good book to look at for endgames is Silman's Endgame Manual. It breaks it down by level, so you don't go studying things that might not be useful to you at your rating, or that you may not be ready for.

What?! But why?! I've never ever had to play K+Q vs K+N. There's no need to master that. Maybe it's good to see it once or twice to appreciate how pieces work together, but nothing more than that, surely?!

Tactics and middlegame strategy, much better to start with that. Basic endgame study (the king and pawn endings) can come a bit later still, as you'll find that you'll reach the endgames up material. And when one side has a rook more, there's no real need to know how to promote that pawn with just the king.

It's quite similar to the K+Q vs K endgame, actually. Just be aware of possible knight forks.

My plan is to finish the endgame drills, of which there are a few hundred, and then I'm going to go all in on tactics, with a goal of 30-50 hard puzzles per day. I just happened to start on endgames first,  so I might as well finish them, but it sounds like your approach would have been better than my current plan (admittedly, the endgames are a lot of fun). Once I've managed a few thousand, my plan is to learn 3 openings for each color, in depth, and then actually start playing chess. I've decided not to play much now, as I don't want to develop bad habits. 

On another note, as a master, what do you think a reasonable 2-year goal would be? I'd like to one day hit 1800 (2000 online), but it looks like that might be a hard ceiling that requires above average ability to achieve, no matter the effort. If it helps, I'm 36 (I'm a bit old), but I have achieved an advanced level in Mandarin Chinese, which takes thousands of hours of brute force concentration and memorization, as well as consistency, so perhaps that is some indication of my memory capacity, pattern recognition, and work ethic?

Do you think a legit 1800, 2000ish online, is realistic? I see a lot of people on here from years ago with the same goal, and it seems that nearly all of them have fallen very short of that goal. Is this mostly a talent deficiency, a lack of effort, or some combo of the two? 

Thank you for your time!

I actually have studied opposition, triangulation, zugzwang, etc., and I've got the simpler drills down pretty well, but some of the more advanced king and pawn endgames are still eluding me (I solve most of them, but it takes me forever). I downloaded Silman, as you suggested, and I downloaded another popular endgame book that I intend to work through once I've finished the drills in the endgame section (there are several hundred of them). I also downloaded about a dozen highly esteemed books, so hopefully I'll be able to finish all of those. 
Oh, and is there an value in even setting the engine to anything less than maximum? I always have it maxed out, but I've noticed some people make videos and practice drills with lower settings. Is there any value in this?
Here are two of the pawn endgames I'm having serious trouble with, as an example. In the first case, I can easily do it if white kingD5 and black king to F6, but if the AI goes in a different direction, it becomes very difficult to figure out. In the second drill, I'm just not managing to get my order of operations down:

11: King And Pawn Endgame: Active King
12: King And Pawns vs King And Pawns: Active King

Not asking any of you to solve them for me, but I am curious as to what I am missing in terms of how to solve them. 

As to the question as to what a reasonable goal is, that depends on a lot of factors. How much time you're willing to spend on it, how well you train, what you do to train, how easy it comes to you, etc. etc. 

First and foremost I think you need to do what makes you happy. Chess improvement takes time. If you're not going to have a good time training, you might quit before you reach your goal. And also think about why you want to reach a certain goal. Why 1800? Will you be significantly happier than if you reach 1700? 1600? And if you do reach 1800, do you then want to quit or is there another goal then? These are questions you need to ask yourself. Chess is such a deep game, you could enjoy it for the rest of your life and never come to a point where you can't learn something new. When I was a 1900 player, I thought I understood chess really well, then I reached 2200 and I understood there's so much still that I have no clue about. This game always has new layers. 

In that sense, I couldn't be happier for you that you really enjoy the endgame drills. If it's fun to you, by all means, you should keep doing that! 

For rapid improvement at your level, you need to study tactics. You need to know which tactics there are, you need to learn the technique of spotting tactics and you would need to practice them to build up your pattern recognition. If you'd do this 20 mins per day, 5 times a week, you'd see a noticeable development in your chess. If you also take maybe 2 hours a week (2 sessions of 1 hour) to study basic middlegame strategy or maybe some basic endgame ideas, I think you'd have a workable training schedule, that is quite heavy, but not so heavy that you'll lose interest in the game hopefully. (note: there's no opening theory in here, at your level understanding basic opening principles is good enough and is more beneficial than memorizing moves).

It's everyone's guess what rating you would achieve this way in 2 years. Like I said, it'll depend on a lot of factors. If you'd train with a stronger player, you'd be able to achieve more than on your own. Also, what rating you want to improve matters. Online blitz ratings fluctuate and can rise quicker than live chess ratings. I'd think reaching a sustainable 1400-1500 playing strength in 2 years would be a realistic attainable goal. Age does not really matter for this.

I love doing the drills and solving puzzles. I almost never get tired of doing it! 

I guess my goal is to be a respectable player. Based on what I've read, and personal, though admittedly arbitrary standards, I've placed that at around 1700-1800 FIDE, which corresponds to the top 20% of chess players, roughly. I'm not arrogant enough to assume I'm that talented, but I do have an exceptionally strong work ethic, so if my natural ability doesn't stymie me, I have the wherewithal to eventually get there--this is why I was asking about realistic goals. In language learning, the obvious goal is fluency, which generally means the ability to understand native material and operate effectively in the language without too much effort. In Chess, it's not quite clear what a goal should be, but I'm setting it at the top 20% of players.

It's a competitive game, after all, right? :-D

By the way, I hope this doesn't come off as disrespectful to the players on here who aren't at that level, but in the absence of knowledge about a given player's worth ethic, talent, etc., I need something to work towards. Obviously the game is about fun, but getting stomped all the time isn't exactly fun, without regular improvement.

Soliciting help in an ongoing game is a fair play policy violation.  I hope this is just practice.

No, I was referring to one of the endgame drills in the lesson section. If you were looking at the thread, you would know this to be the case...