Only horrible people like queen endings

Yeah, they're hard. But you might find that you can go by generalities even here. Obviously you can't calculate every possible queen check. But you can bet that if you can expose the other person's king, you will probably get at least a draw -- you often don't have to calculate that sort of thing too much. Passed pawns are extremely important here, often worth many pawns. Oftentimes you can just focus your energy on escorting one pawn in particular and your opponent can't stop it -- queens are so good at escorting. You don't necessarily have to calculate tons of variations to know that stuff is good.

Other than that, just be on the constant lookout for queen forks/checks; beware an exposed king if you're trying to win. And don't get too many pawns lost due to queen checks/forks :) Indeed a lot of it is going to be about just being very aware of all the checks and such, but as said there are at least some general ideas too. If these endings are hard for you that's fine; they're hard for pretty much everyone :)

generally no queen endings for me, but if if enuf pawns can be queened, like a ten queen game, that generally could be generally good.

I recently got into my first queen ending in a blitz game. It was 4 pawns vs 4 pawns on one side and 1 pawn vs 1 pawn on the g file, with both kings on the huge pawn island and any pawn advance results in loss of material. 

All of that plus 2 queen running around.

With 30 seconds on both our clocks, I just offered a draw and my opponent hated the ending just as much as I did. Putting in so much effort just to not blunder is really frustrating. We knew that in a blitz game it would be a premove war.

I recently played an expert at the Mechanics Institute in SF. I hung a piece on move 8 but played on hoping for a free lesson in technique, which I was happy to receive. After the game he mentioned a conversation he overheard with GM Edmar Mednis, who wrote Practical Rook Endings (a terrific book). Mednis said something like "Queen endgames are 100% calculation". My opponent agreed, saying that centralization and pawn structure are good, but there's no getting around the tactics and calculation. 

So I avoid them like the plague unless I'm sure I'm winning. Here's a good queen ending- Anand as White just works with threats all the time and converts a pawn-up ending. I remember watching this one live- for me it confirmed that Anand was outplaying everyone at the Candidates. http://www.chessgames.com/perl/chessgame?gid=1751476

Very interesting game. I've never had the guts to look at games by players of that magnitude, but I think it's worth the effort to study them after all.

uh okay.

Queens are the most powerful piece on the board, and when you end up with queens, kings and pawns it starts to get tricky for both sides, with the winning side pushing for a win and the losing side pushing for a draw. The reason for this tricky situation is due to the sheer power of the queens, against the much lesser power of kings and pawns. 

A Queen isn't always the most powerful piece on the board. And her (relative) power comes at a cost - she is vulnerable.

Ineteresting comment on structure, five. So if we take the most rudimentary position possible could we say the following:

I would think either way the passed pawn gives white winning chances -- black's queen can't stop the pawn by normal means because the white queen can escort it. I feel like white should win in those positions, given the good king shelter. Although black will naturally try to find checks right until the very end. But if white's kingside pawns were much more spread apart from white's king, that might really reduce his winning chances.

so exaggerating the f pawn thesis

Alison, all diagrams you gave are probably won because of the c-passer. Queen + passer versus queen = passer promotes (only possible with queens).

This Example by Averbakh shows the winning idea:

Very interesting diagrams.

In my experience Q & passed P endings often come down to geometry and math. Assuming White has the passer, geometry as Black's chances depend a lot on getting an avenue for an early blockade of the passed pawn and then math as to whether Black's King and pawns have move time to recover drawing chances. Black might have geometric perpetual possibilities, too.

White often needs math move time to advance the passed pawn while often also getting the King positioned to avoid possible perpetuals. Then it's geometry to try to get lines for a usually helpful King tempo check which might clinch the deal.

I didn't bother to calculate out any of the diagrams but it looks like they might all fit this curriculum. So kids, stay in school!

The Ruy Lopez Exchange.

I like playing Q + connected passed pawns vs 2 R

Good examples, five

The worst endings I find, are knight endings.

Nothing more pathetic seeing a knight (or two! or three!! or four!!!) slither around the board trying to land a check.

Sure, if white decides to isolate his queen almost entirely from the action. Making it almost as if he's a queen down :) It's interesting, I guess, what the g3 h3 structure can offer. But it doesn't seem likely that black would let it all come down to a tempo like that (e.g., only hint at making checks when white is so close to queening -- you'd think black could pressure f2 for example long before that) unless he just had no choice. But this does show the pitfalls you can run into -- it's that important to avoid any hint of a perpetual check in queen endings.