Chesstempo Progress

@jimj12 You definitely want to solve it all the way through.

Most tactics up to a certain point are completely trivial if you're capable of some deductive logic. I mean if you're in a position where you, crucially, know there is a tactic and you have the option of Qxh7+ or a bunch of normal moves then the first move is probably Qxh7+. You can make that move, see the result and after the move you'll probably be able to suss out the idea (if the computer doesn't immediately give away its queen ending the problem because it sees that as better than letting you find the super tricky mate in 7 Qxh7+ just started). You might've increased your tactical score a bit but you've certainly learned nothing doing it this way.

This is why in many ways I think books are better than online tactics trainers, particularly for those that don't have the discipline to not solve by the process of elimination. When you have to write out the line and all major defensive variations before seeing the answer (and generally doing these problems in batches of say 6 before checking any of them) then you have to really see everything before making the move.

Working everything out helps build the intuitive understanding that eventually let's players start to recognize tactics in a fraction of a second that lesser skilled players wouldn't find even if you gave them hours and especially not if you didn't tell them there was a crushing win available to them. It also improves your visualization and board vision, which are also crucial.

I actually suspected I wasn't learning much from Lichess puzzles by the fact that I enjoyed doing them (your writing out the lines method seems tedious, but much more effective)

I realise now that I often solve these puzzles via eliminating obvious bad moves, which as you say leads to me not learning much.

And I have thought about how much of a difference knowing that there is a tactic available makes. In real games, this is not the case and it ends up being a gamble for me to spend time looking for them.
Maybe there could be use in having some 'non-puzzles' mixed with puzzles - ones with no clear tactic?

Are there any books you recommend?

Definitely agree on the value of mixing in non-puzzles. ChessTempo does this a bit where there will there are some really enticing ideas, but it turns out the answer is one of the plain moves. Or having you solve problems where the opponent just sacrificed a piece and you have to decide whether or not to take it. But most of these problems seem to have fairly high ratings which ruins their utility for lower rated players since they won't get them very often.

A book I've always enjoyed is an oldie - Fred Reinfeld's 1001 Winning Chess Sacrifices and Combinations. There's a newer edition available that has the problems in algebraic. Just a really great selection of tactical problems broken down by motif.

@Rembrandt16,

I can speak from my own experience that going all the way through puzzles is the fastest improving method by far. I used to give myself like 1 or 2 minutes at most when I first started doing tactics and I was not improving at all that way. Then here on lichess and CT I started not limiting myself and trying until I either solve the puzzle or just give up after I get tired. I remember thinking even half an hour on a single puzzle. But with this discipline I had an amazing improvement to my tactics rating. I first broke 1900 which looked impossible to me before but then 2000 came, then 2100 and then 2200 and then even 2300. And I hit my peak rating 2360 here on lichess a month or so ago. Right now I fluctuate between 2100-2300 depending on the time I spend on the puzzles. And when I bullet through the puzzles I get as low as 1700-1800.

Similarly, with the same approach, I consistently improved on CT for a six-month period from 1500 to 1950+ and I broke 2000 for the first time just a few weeks ago.

My current approach though is giving myself a 10 minute limit on lichess and 15 on CT. I could say this is better than not having a limit at all because having a limit kinda force you to not get lazy and focus on the puzzle with your all attention. And after you improve, most of the time you don't even need that much time anyway unless you get a very tricky or subtle puzzle.

@OhNoMyPants

Thank you very much for the advice. I think I will switch to ChessTempo for puzzles

I read a passage recently here:
books.google.com.au/books?id=nSVBCwAAQBAJ&pg=PA234

I'm not sure if you'll be able to see it but I think it is relevant to the way that Lichess generates puzzles

I recommend this book http://www.goodreads.com/book/show/1875671.Practical_Chess_Exercises

It really forces you to use your brain and explore all the options. You have no idea if there is a tactic on the board or if the best move is a positional one. I haven't done a lot of the problems yet but so far I find it very instructional and the reviews are great.

Thank you guys for the advices and such. I will do it all! :D I will improve. I believe.

Have fun!