Is the only answer to do more tactics?

I goofed up on this position below.

Besides playing slower and giving it more time, are there specific tactics which revolve around categories like "Choose the correct rook" or "Determine the best bishop move" in positions where it is pivotal like in the position I encountered above. I hate doing tactics, not because I don't like solving them, but because they seem so random. 

If I do openings, I know there is a pattern and I go over them regularly. Is there something similar with tactics where you can categorize and go over in a similar way? For example, I might play a Sicilian Najdorf, then a QGD line is played I am familiar with. The third game is a Ruy Lopez with lines I have gone over many times. So I am set in making choice moves rather than having to use opening principles. I can just play them from the beginning of a game. Maybe the 4th and 5th games are similarly, but then they repeat and I get the Najdorf again. So, I feel a sense of growth like a seed turning into a flower when I apply my opening experience.

Tactics on the other hand are like once in a full moon, and what are the chances you will remember a random tactics puzzle you did 4 months ago? I can remember opening lines I played 4 months ago though. So, that's basically where I am. I want to improve with the tactics in positions 20+ moves into the game, but each time it seems to be something different. Tomorrow I might have to move the rook to the d file for example. It will have nothing to do with this position. I may not see this Rdb8 come up for another few months. By then, I will have forgotten it and repeat the same mistake.

So, what do you do? Is there an organized way you can remember these things? Do I make 64 folders and each time I see a tactic happen on a square, in this case b8, I make a .pgn file and put in the FEN like this one: r2r2k1/1Q1b1pp1/B3p2p/8/P2q4/8/1P3PPP/R4RK1 b - - 0 21

Then review constantly until it goes into long term memory? Or, is there another way?

I admit that this position is rather hard to see. I mean, a natural player's first instinct is to develop a rook to a centre square rather than to lump the two rooks on the a- and b- files in that position.

It's not immediately obvious that the Rdb8 move is better, but one thing is for sure - leaving the a8 rook there maintains the attack on the a6 bishop. That White queen on b7 seems rather stuck there - itself in a nearly trapped position while trying to defend the a6 bishop at the same time. Just coincidentally, in the correct sequence of moves with Rdb8, the move Rxb7 happens to protect the eventual attack on the d7 bishop by the d4 rook.

I cannot well classify tactical opportunities into the various groups, but some are more obvious than others. In this case, one has to notice that the White queen is rather tied down to the bishop and the Black rooks must not be captured freely.

It takes quite a fair bit of experience to notice such moves, and I myself did not even notice it (after looking at the position preliminarily for 15 seconds). Regardless of which, you would have taken away a useful pattern to remember in the future. 

I would probably recommend you to continue training tactics, and more importantly, do not be upset over this one oversight - this has to be counted in consideration with the fact (I'm pretty sure it's a fact, am I right?) that you have spotted and played many beautiful tactical opportunities in your games thus far.

If you have seen my post before https://www.chess.com/forum/view/game-analysis/counter-intuitive-move-2-49498953, you will notice that a similar idea comes up. In fact, I had a predecessor situation to that in https://www.chess.com/forum/view/game-analysis/counter-intuitive-move so I would say that I have not learned from my first "missed opportunity" of moving a rook to b8.

Tactics come in families, just like Model Mates do. Learn the tactical groupings, such as Pin, Fork, Double Attack, Come-Back, Skewer, Overload, Decoying, Diverting, Interference, Battery etc and review several examples of each.

Tactics often occur combined... most notoriously, Decoying is almost always associated with Diverting, and often with other elements (such as Fork or Double Attack) as well... as in the following example:

- "Tactics on the other hand are like once in a full moon"

This is simply not true. It might be that you only NOTICE the tactics once in a full moon, but that's not the same thing.

That "example" came from one your own games, correct?

(If I can remember)

Correct.

Here is another problem. I looked at that puzzle you posted and decided on Bxe5 for the first move for a different idea. I initially thought of Rd1 which Stockfish is saying is the best move so far. Then I was going to possibly take on e5. The puzzle doesn't show that black can't castle. That is hidden in the FEN. So, I don't think this helps at all.

I reject the better move/s and then end up trying to work out inferior lines like Bxe5, Qxe5, Qxa7.

I'm not sure what difference it would make even if Black COULD castle... he can hardly reply to 1. Bxe5 by castling (it loses his Queen), nor can he reply to 2. Qxd8+ by castling (it's illegal to castle while in check). So the point seems to be moot.

If you didn't spot the decoying/diverting - decoying/diverting - fork combination, then that indicates that you need more tactical drills.

My main point... which you didn't address... is that tactics come in related families, just like openings do.

Another example:

Checkmate patterns base on an advanced Pawn on f6 vs a castled opponent are referred to as Lolli mates (named after a chess writer). Here is a family of Lolli mates, getting progressively more complicated:

... and to demonstrate that these tactics really do come up in actual play, a game from an over-the-board tournament, in Round 2 of the Moncton Easter Open 1983:

Posting a bunch of puzzles that I don't understand is not going to help. I am still trying to understand the first. 

If black can't castle, then Rd1 is the best. If black can castle, then Re4 is the best. Bxe5 is not the best move in either case. Why make a puzzle like that with Bxe5 as the answer when there are two better moves depending on castling rights?

I think you put the position into your computer incorrectly. There IS no Re4. White's only Rook is on a1. It cannot reach e4 from that position.

Why would Rd1 be better than Bxe5, which wins a piece and skewers the Black Queen to the Rook in the corner?

I have no idea what lines you are looking at.

I meant Ne4

Those 2 moves are better than Bxe5. 

The answer isn't "do more tactics," or studying "which rook to move." The answer is that in complicated positions, with a lot of pieces in contact, you have to calculate. Principles and maxims are of no use. It comes down to "I go here, he goes there . . . " You just have to grind it out, Look at the position and figure out why Rdb8 is better. You can only do that by calculating and visualizing. 

According to the computer, yes. But that requires looking ahead nearly 20 ply (10 moves). An easy task for the computer, but can YOU calculate that far ahead? The line I gave (Bxe5) requires looking ahead only 3 moves, and gives you an endgame a whole Knight plus a Pawn ahead. Should be pretty fool-proof.

But I'm glad you brought this up. Is your approach to tactics REALLY to get the computer to tell you what the best move is, then memorize that? If so, then we've found the problem.

There are billions of possible chess positions. You CANNOT learn tactics by trying to memorize the best move in each of those billions of positions. What you need to do is to develop your ability to FIND the best move... or at least, find a good move that gives you a winning position that you can understand and play correctly.

So using the computer to find the best move is a blind alley. It only solves that one position, out of billions. That's basically worthless to you. What you need is a method of FINDING the best move, and a computer cannot offer you that.

Learning the families of tactical elements, and the typical ways to combine them (ever wonder why it's called a "combination"?) will give you the tools you need to improve your tactical vision.

"But I'm glad you brought this up. Is your approach to tactics REALLY to get the computer to tell you what the best move is, then memorize that? If so, then we've found the problem."

Is your approach to tactics what you did? Perhaps we found the problem. I found Rd1, but your puzzle said I was wrong. So then I thought of other lines which didn't work.

If my guess WITHOUT computer was Rd1, then I think I should use Rd1. I didn't see Re4. I don't see this as a problem. I think Rd1 makes more sense in the position than Bxe5.

After that, we can further use more common sense. You don't have to calculate it all, and you don't need a computer to tell you white's moves. 

Chess isn't 99% tactics, it's 99% calculation, or calculation + visualization. Learning tactical patterns is very important, but you need to calculate to see if a particular pattern works in this exact position. Work on solving harder problems, and solve them without moving the pieces. Write down your variations and don't quit on a problem until you've given it at least 10 minutes. Your calculation muscles will grow rapidly