Tactics Trainers are a Lazy Shortcut

note: for this article, "tactics trainers" refers to chess.com's tactics trainer, chesstempo.com's tactics trainers, lichess.org's tactics trainer, and any other sites that use the same method for gathering and presenting problems

It looks like a controversial statement, and it may be.  However, I honestly believe it is true on both ends.  Here is what I mean.

First, on the side of the website.  It takes a considerable amount of programming to create tactics sets on a site like chess.com or chesstempo.com.  Even after you've extracted a lot of tactical positions from real chess games, you don't have a usable set.  You also have to generate the exact solution.  This is harder than it seems, programatically, mainly because it is not always clear how many moves the solution should be.  A related problem (you'll see why) is the difficulty rating of the problem.  This is almost impossible, programatically.  These last two problems must be solved to have an effective tactical position to add to the collection, but for tens of thousands of problems, it is impractical to do by hand.

The website's solution is crowd-sourcing.  Open the problems up to the eager aspiring tacticians and let them 1) suggest when the solution should actually be and 2) "compete against" the problem to set it's difficulty level.  These sound like great solutions, and they work.  Kind of.  I actually think the first one is a great solution and have no problem with it.

However, for the second one, the difficulty level, I disagree that the solution is acceptable.  The difficultly level of problems is extremely important to an aspiring tacticians study.  A user wants (and should be able to!) to set a level of difficulty and then solve problems in that range, so it's important for the difficulty levels to be accurate.  However, the system in place (#2, above) to assign difficulty levels to problems actually only succeeds in discovering how difficult the solution is to guess the first move of.  No matter how tricky the tactic or combination, if the first move is an obvious check, the difficulty level will be set pretty low.  With thousands of players coming across a problem, a very large number will be just guessing for one reason or another.

So, ok, there are some problems where the difficulty level is out-of-whack.  Is that crippling?  Ordinarily, I'd say no, but the truth is, one of the most effective ways to study tactics is to drill easy problems rapidly.  Sadly, low rated problems (the ones you'd pick as 'easy' for that kind of drill) are the ones disproportionally affected by the bias mentioned above.  What winds up happening is a) you get a lot of slightly to a lot more complex problems with low ratings because the first move was the most obvious check in the position and b) nearly every problem you solve winds up being 'just play the most obvious check in the position to solve.'  This is NOT effective study.

This is also why tactics trainers are a lazy shortcut for the user.  It is far from optimal.  A much better way to train tactics is to train with hand-picked sets.  This is why I always recommend tactics books over tactics trainers.  Books are almost always arranged by difficulty and the problems are picked out by an author.  The author, ideally, has sound pedagogical reasons for selecting the problems that he/she has selected, such as making a complete set of all tactical ideas and progressing from easy to see motifs to disguised/combined motifs.

Sure, random positions serves a purpose (such as not having a good idea of what motifs to look for more closely simulating game conditions), but that purpose is not always one that you should be looking to achieve.  When learning tactics, you want to practice the idea first.  Compare it to solving math problems.  You should not have a lesson on one mathematical concept and then do random unrelated math problems that might or might not incorporate the idea you have learned and should be practicing.  You practice and drill the idea you are working on.  This type of control is what you are looking for.  There will be times when you will want to use random problems, also, but I submit those times will almost always be less often than most people realize.

In my view, this is another example of the "chess industry" giving people what they think they want rather than what they need.  There are excellent resources out there that are more suited to what most people need.  Often they are in the form of books of tactics (which nowadays can often be found in pgn format for download, often by the author/publisher).  Always have a plan for how you are going to learn a subject, don't be afraid to seek the advice of chess teachers who have a track record of improving players that are your age and strength (I could write another article entirely on whose opinions not to trust, but it's enough for now to just point out the ones you should trust).

Many excellent points given!

Still, as a new chess player, I like that the tacics force me to evaluate situations and improve my visualization ... well, i hope the problems are.

Then when I take the time and read my Predator at the Chess Board books, I have more a-ha moments.  (not the 'take on me' ones, but the lightbulb ones ... see music of the '80s).

I'd like it if the tactics didn't push for quick solutions as I'm not ready for fast-paced games.

YES! As a novice I'm in desperate need of tactical puzzles which are more thematic to improve my ability to recognise specific patterns with increasing complexity. I was hoping chess.com would offer us the option to limit our puzzles to just 'mate in 1's, 'smothered mates' or 'rook vs rook' tactics etc. This would really help in improving specific weaknesses systematically.

Susan Polgar's "Chess Tactics for Champions" does a good job with chapters arranged by tactical theme. I wouldn't say they are arranged in order of difficulty, but they are mostly two- and three-movers that you can solve in a few minutes.

I stopped using ChessTempo when my rating went up and the problems became too hard to be the kind of drill I wanted.

Has anyone used Chessimo for iPad ???

Please, this is so wrong; books over interactive sights, that's just laughable.  I learned more from TT in two years than from reading books for five.  Why must you bash TT?   First off, all the work folks put into learning, it certianly isn't LAZY, so wrong right here; shortcut, maybe, but tactics are an aspect that needs it's own attention, so stop being cynical and enjoy the site or else don't play TT.  Why are you writing all this crap instead of playing?  Me?  I'm playing as I write this, wouldn't waste my time any other way.

I know this sounds like I'm shilling for chess.com, but you folks are going to be amazed when you finally get over to V3 and see the new Tactics Trainer.  My favorite way to use it is to practice (in unrated mode) problems I've already missed.  I can choose to further filter my problems by selecting a particular motif.  The rest of V3 I don't know much about but the new TT rocks!

Whatever happened to different strokes for different folks?

It's fine with me if people want to use tactics trainers. I've done so myself. But I also have several tactics books I like.

Chess is a tactical game so of course tactics trainers are a shortcut.

What is V3?

Many intelligent people joke that they aren't smart, they are lazy. So they look for the most efficient means to achieve a result.

I'm all for lazy shortcuts myself -- assuming they work. Otherwise, it's like the advice from a Donner Party survivor:

Remember, never take no [shortcuts] and hurry along as fast as you can.

Words to live by.

tactic trainers in mobile it's genius. I switched from stupid games to tactic trainers to kill time. Since then I've solved a couple of thousand problems. By far more rewarding than candy crush saga. The problem with doing too much tactic training it's that I get obsessed with looking for mates when I play, so I don't focus on strategy and development enough and my game suffers. Tactic trainers are great since it enables one to do large volumes of problems in a short amount of time. The drawback is that it can be a little like playing 1 minute bullet games. Large volume practice, but low quality games. in conclusion one needs to train with other tools in addition to tactics to improve. tactics it's just a small part of playing chess.

Version 3 of chess.com.  Check out the beta version by scrolling to the bottom of the page and clicking on "Try the new Chess.com".  Don't worry, you can come back here very easily.

Never cared for the Chesstempo interface so I really can't compare them.  Of course I haven't been over there in years, so it too has probably improved.

Chesstempo is the very best as evidenced by the fact that chess.com is stealing their features for v3.

Yes. See http://chessskill.blogspot.com/2013/02/chess-tactics-training-on-ipad.html for my review of the app.

I haven't been using it much since I wrote that two years ago, but do use it from time to time. Of the five apps reviewed there, the one that I use the most in 2015 is Tactic Trainer. Tactic Trainer seems very close to the ChessTempo model and chess.com's Tactics Trainer.

Last week, I recommended ChessQuest to a young woman (recent HS grad) who has been the Idaho State Girl's Champion and wants to make the push towards A Class and beyond while in college. For serious training akin to books, but in an interactive electronic format, ChessQuest on iOS devices is the way to go.

I agree with PossibleOatmeal that books are a way to go, but not wholly the way. I find that a combination of crowd-sourced and hand-selected is best. Frankly, either/or notions didn't make sense when Plato invented Western Philosophy on that basis and they don't make sense now. One must choose a life partner, but you can drink red wine and white, and blush, and sparkling. One can use this site's tactics trainer and ChessTempo and the Encyclopedia of Chess Combinations (which has an electronic version, too) and Bruce Pandolfini's Beginning Chess (for those under 1000 Elo) and ...

I agree that hand-picked sets from a good book are the best for initial learning then using something like CT is better once you can pick your own sets.

And for premium members, let's not forget all those delightful tactics courses over at the Chess Mentor.  Gm Patrick Wolff must have at least a dozen courses on various tactical motifs and there are others as well.

To clarify, I do mention in my article that tactics trainers aren''t "useless" and do have their place, so I agree it's not a dichotomy (either/or).  The "lazy shortcut" thing refers to improving players trying to improve by doing nothing but tactics trainers and the makers of the tactics trainers trying to get jumbo-sized tactics sets by writing a script and hoping for crowd-sourced help.  It doesn't mean that you should never use the tactics trainers.

Thanks for the clarification. I agree.