I think a mistake I and, I assume, many others make is to do tsumego too quickly. It`s not enough just to see the answer. If you really want to get the benefit, I reckon you have to read it out as fully as possible, in order to understand the shape and all its quirks properly. Otherwise, you might find yourself in real games being able to see promising moves, but to be unable to make them work.
I`ve also become very interested in the endgame recently. For example, there are many, many situations where knowing the correct technique is worth an extra point or two. That doesn`t seem impressive at first, but if you were able to improve your technique to the extent of playing the right move only five times more in an average game, well, you`d be gaining anything from five to fifteen extra points cumulatively. That alone should raise you up by one rank.
Studying the endgame has made me begin to rethink what size and sente mean.
Do you, for example, tend to respond to when the opponent drops to the second line to threaten a monkey jump?
Do you tend to grab a lump of six or seven stones left in atari (or defend your own)?
Do you sometimes apply mutual damage, but find you receive more damage than you inflict?
These are typical experiences for me, and I don`t think I`m unique.
Take the monkey, for a start. The monkey jump is usually sente, but setting it up is not necessarily sente. To clarify, suppose I drop to the second line to threaten to send Mr Monkey into your territory. Well, what if you were to play elsewhere? I could then unleash my simian chum, but you could defend to cut your local loss to 8 or 9 points. If, in return, your initial tenuki were bigger than that, then you`d have done better than simply preventing the monkey jump. Indeed, even the monkey jump itself is not necessarily sente. The other side might still be able to find bigger moves than simply defending.
Again, suppose I`m threatening to capture a string of seven stones. That should be worth 14 points, but if you decide to play one of a number of second-line plays, it`s quite possible that you`d be the one gaining. It can be quite hard, even for experienced players, to believe that saving or capturing stones (in gote) is often smaller than making a hane-connect on the second line.
As for mutual damage. It`s
not a test of machismo (okay, not something I`d know a lot about but still...). I suspect a lot of players choose mutual damage for pyschological reasons, i.e., not to be dominated. But what if you threaten less damage than you receive? The key is to assess size. Sometimes just submitting is the wise option.
You could roll this kind of thinking forward to the opening. Some people seem to tenuki for the sake of playing tenuki, but surely people frequently assume a move is sente when it is not.
Let`s consider the word itself. I`d imagine a lot of people, when asked what sente means, would reply that it means "a move that needs an answer". But that`s not really true at all. 先手 simply means "the leading hand", i.e., the one who goes first. Really, taking sente only means being the first to play in a particular part of the board. If you play tenuki, then you are literally taking sente in a different area of the board.
The real question is: "what is biggest?" If you are able to get the to biggest points first throughout the game, you will win.
This could give rise to an interesting illusion. I remember a strong player (high dan) offering to play a teaching game in which he promised to play gote as often as possible but still win. I wish somebody had taken him up on that. I suppose what would have happened would be that he would answer the other`s moves, but in such a way that it`d be big enough to win. But wouldn`t that really be just sente in a party hat? I mean, he might be answering his opponent`s plays, but still getting to the more important points
first.
The proverb says "Urgent before big". Doesn`t this really mean "the biggest before big"?

An urgent move is urgent because it is in fact the biggest! Taking a 30-point move won`t exceed allowing a 16-stone group to die. Usually the problem is realising that a move is actually very large. If the opponent threatens to cripple your shape or destroy your base, it can be tempting to play a big territorial move elsewhere, but that can often be like accepting a small loan on sharkish interest rates - you get a big move in the short term, but as the game goes on you pay for it over and over again.
Go is very, very difficult. It can be so hard to know when a move is big, but on a long-term basis only.
Anyway, I wonder if revisiting fundamental concepts could be profitable for me. After all, sente and size are among the first things you learn about, but thanks to looking at the endgame afresh I`m starting to see them in a very different light.
Anyway, you never have to answer the opponent`s move. You want to play where it is biggest. If answering is biggest, then answer, if playing elsewhere is biggest, then play there instead, but no matter what, you never
have to play anywhere you don`t want to!