Some thoughts on the tactics trainer

As I wrote recently, I've found that my tactical skills have gotten worse, so I've been spending a lot of time with the tactics trainer. I think it's a great tool, and I can tell it's helping me. But it also has flaws. Here are some of my thoughts, which I hope will be helpful to all aspiring tacticians.

First, here are the flaws in the TT:

- You only get to play out one line, and it's often not the most challenging one. Now I understand it would be hard to code multiple variations into this tool and put them into each puzzle, but it would be very helpful. It's particularly frustrating when the computer plays a relatively weak or simple line. You spend all this time calculating line A, which is the whole point of the tactic, and it plays line B instead. And maybe line B is technically best according to the engine, but it's clearly less challenging to a human player.

- Relatedly, the line you play out is often cut short. Usually this is OK, but there have been times when I wasn't sure about the evaluation past a certain point, except that it seems like I'm doing well, and would have like to play it out more, but the puzzle would simply end and tell me I solved it.

- When you get the problem wrong, it doesn't play the refutation. [However, this is also a blessing in disguise, I'll touch on it below.]

- The hints give away too much. When you ask for a hint, it simply tells you the first (or next) move. That pretty much gives away the puzzle. It would be great if the hint were more subtle, e.g. a text box popping up with some helpful info about the position. Again, I understand that would be much harder to code.

Second, here are two common complaints I see on here all the time, and my thoughts on them.

- "The rating system isn't fair." First of all, who cares? Why are you whining about a meaningless online rating in a training tool? Second, I disagree. The rating penalties for mistakes are bigger than the rewards for getting it right, and that's what people whine about. Well, guess what? That's how it is in a real chess game. You make a mistake, you lose. No partial credit. No one cares if you played brilliantly all the way up until then, or if your move "almost works". But if you make the best move... the game may go on, and you have to keep playing good moves. The positive value of one good move is smaller than the negative value of one mistake or blunder. Chess is a brutal game.

- "I suggested a move that's just as good as the given answer, and it was counted wrong." Yeah... you know what they got when they crossed a bulldog and a shitzu? This. I've been doing TT for quite a while (hovering around 2800 TT rating now), and I've had this thought myself multiple times too - but 99% of the time, that's false. There is a reason the move you suggested is wrong, and you have to find it. That's where my point above about not playing the refutation comes in.

Lastly, this is why I really like the TT.

- It really does help you get better at tactics. This may sound idiotically obvious, but whether the tactics tool is useful for improving tactics is hotly debated in these forums. I was at 2400-something TT rating a few weeks ago, and stuck at ~2800 now (up and down around that plank), and I don't think the difference is meaningless. In my daily games in these last few weeks, I've spotted several tactics and avoided several blunders that I think I would have walked into before. It really feels like it's improving my vision and thinking.

- There is a great variety of positions. I'm very impressed with how many different types of games and studies are included in TT. It trains you on all sorts of middlegame and endgame tactics.

- Maybe most importantly, it teaches you to think, and to think independently. In some ways, it's a bare-bones tool - no lessons, no explanations, no side lines given. It forces you to do all the work yourself. This is why the fact that it doesn't show you the refutation of your wrong solution is a blessing in disguise. The puzzles I'm getting now are pretty tough, and I often get them wrong on the first try. Figuring out why my try is wrong has been just as helpful as getting the problem right.

In summary, I think the tactics trainer is an excellent tool for advanced players. It's like learning to swim by being taken out into the ocean in a boat and thrown in. If you already have a good foundation of swimming skills, it may be a useful exercise. But if you don't know how to swim at all, you're mostly likely going to be feeding the fishes. For beginners and those still learning fundamentals, the TT is probably too bare-bones. For those players I'd recommend tactics books, videos, lessons, and coaching. At that level you need a lot more guidance and explanation. But once you have a strong foundation, this tool, which basically simulates a chess game where no one is there to help you except yourself, is very useful.

A great text on TT, I do a lot of tactics myself but at much lower level.

I was a bit surprised myself to see no responses. I've started only a few forum threads, but they've all gotten a lot of responses. I may have posted it at a time of low site traffic, and then it quickly got displaced by other threads.

I have a serious love/hate relationship with the TT...

I would like to see the hint process changed. Currently, with a "hint" revealing the next move it isn't very useful -- more of a "I give up" button.

I think it would be really useful if clicking 'hint' instead revealed the category (or keywords) for the puzzle. Knowing it's a mate in X moves, or "discovered pin" or something along those lines. Just a relatively small piece of information -- that would qualify much more as a hint IMO.

Also -- adding the keywords for the puzzle would be relatively easy, wouldn't it? That information already exists for each puzzle... seems to me like it would be easy to display it on command. Unless there are puzzles out there so new that no keywords have been added yet (I dont think I've noticed any like that thus far)...

Great post Chud.   Glad I'm not the only one who thinks that the solutions are cut short at times.

@FM chuddog Great quality comprehensive post as usual, I have tried to address your 1st main point with examples, if anyone understands & agrees on the need for improvement can +1 with an optional comment here: https://www.chess.com/forum/view/suggestions/faulty-daily-puzzle-2-solutions-but-can-only-accept-1

Sorry my reply cannot do justice to your post at this moment, I'm on a tight lunchbreak, I agree with you & all entirely 

The multi-variation or line (s) expressed in the 2nd paragraph as found on Chess King Apps is good idea.

I love TT. I have probably played 40 thousand puzzles in the past 3 years. But most of the things you say don't apply to me. I still play in v2. I will play v2 for as long as it is running. 

etc. ... I think suggestions like this can be sent directly to the staff.  Start at the footer of the page for the little useful links ... ? Help -> Make a Suggestion and fill out a form to suggest it!